Transcripts
1. THIS IS AFTER EFFECTS!: Hey, guys. I'm Set. Whether you use AI
or manual editing, you're probably wondering
why everyone is talking about Afterefet. Well,
let me just show you. For example, we can
perflytrack something that moves and keep it aperfly
stabilized or dc. Or you can create motion
graphics animation like this. And maybe you need some text
that is pushy and pulsy. And you know what? Let's make it three D and give
it some lighting. What else can you do? Well, you can change the background, you can move it or you can remove the foreground
and change the penguin. You can perfly speed
Bram gesters to music. You can create CR environments
to contain odio reactors. Or you can create
three D movement from a free spring like that. And finally, my
favorite ich art. If you're ready to
start, let me know.
2. Download The Resources!: All right, so before you
start doing anything, you need to download the
resources of the course. So you need to go
to the landing page of the course and
then scroll down, go to Projects and Resources here and click on
this Dynamic Link, first of all, and it will lead you to a Google
Drive folder, so you need to simply
download this file, right? After you have downloaded
the zip fully, simply right click Extract here, and it'll ask you a password. The password is, let
me show you Leviathan. One, two, three, dot, dot, dot. This is the password. All right, Leviathan, one, two, three, dot dot dot. Alright. And then
simply press Enter. Wait a couple seconds. And, you have the full files. Alright. So at that
point, you have all the files extracted for you. Alright? That's it.
Now we're ready.
3. Optimize The Software: All right. So before we
even start doing anything, we need to make sure
that After fax is optimized to be a little
more stable and fast. So let's open the program completely until you
reach the screen, and then you go to file. Well, actually,
edit preferences, and then go to memory
and performance. Wait a couple seconds
and then disable this. Disable this option here, enable Multi frame that has to be off in case you
have it, disable it. Then this number
here needs to be between one quarter and half
of your taller RAM, right? So that's the sweet spot. So in this case, as
you can see, I have it on 50% of my RAM. In your case, maybe you can
afford to go to one quarter. Make sure that it's 25-50% of
your total RAM. That's it. Then go here on scripting and enable this guy allow
script. That has to be on. Then go to AutoSave
and enable put this number to four or five and then increase this
number at least to 50. That's my recommendation because you never know when
after facts will crash, and you may lose some
precious minutes of work. So the higher this number is, and the lower this
is, the better. So these are my settings. Just go around that
area. All right. Okay. Then go to previews here and check design
fundamentally check. This has to be on. Then go to general and
so check this on here. This has to be on
this has to be off. So enable home screen
has to be off. Finally, go to media
and cache here and this number has
to be in at least 30. That's my recommendation, and then the highest, the better. So fundamentally, this is a space in the SSD
where Aftereffects will draw garbage data to be faster when
you ran the stuff. Just change these folders to the custom folder
that you choose. Don't use external
SSDs for this, has to be internal, so
it's fastest as possible. So change that folder
here, here and here. Once you have done
all these things, you can finally
close and remember to purge the cache
data every so days. You know, maybe
three or four days, once a week, it depends on
how much you use A refract. So to do that, go to edit, purge, and then all
memory and cache. So let's see how much data. Okay, in my case, as you can
see, I have 64 gigabytes, so just clean it completely. So remember to purge that data often because if
that space runs out, Afterefax stops
working instantly. So the larger the folder is, the better it is
fundamentally, right? So remember to do it. And
finally, you know what? Just create the empty project. So just randomly, you know, click a new composition. Just don't worry
for now. Confirm. Like this and go immediately
here, composition, preview, and uncheck this
case frames has to be off. Remember this, when
you work on a project, you may accidentally close the windows like this and you may get you know, feel lost. Oh, wait, wait. What happened? Right? So if this
happens, don't worry. Just go to Window. Workspace and reset. Boom. So that brings you
back to square one, alright? And always remember that the final result is
always going to be only displayed on this window
called composition. Here, right? So if you happen to
be here here here, it will never display what is happening truly on the
composition, alright? Go here, composition. And that's it. You are free to move around the windows like
this as you like. You can save your
custom presets here, workspaces, where you want. But remember that that's
how to go back to square one in case you kind of close accidentally windows. Once you have done
all these things, for the most part, after effects will be fully optimized
on your side. So we can call this
a win already. Wow.
4. Full UI Tour: Let's make a little tour of
the entire interface, right? So when you make a new project, remember to just press Command
N or Control N or just click here New and select the
resolution of the project. So even if you foot is four K, I recommend actually working on a ten ADP composition
because that way, you're going to make
everything all faster. Let's just choose
standard ten ADP. Also, 80% of the screens of
the world is still ten ADP, so most people will
never appreciate four K. But the
way, that's fine. Recall your footage is four K, but then go to ten ADP
when you are here. Then let's just choose 24
frame rate per second. To have a cinematic feel or you can go double
in case you make it more gaming genre so advance, if you need fluidity,
just go to 48 or 60, 50 where we want.
That's fully custom. So let's go to 24,
and that's it. And finally, choose a
length for the timeline. So let's give it 1 minute
and that should be it. Alright, so let's go and order. Here is where you drag in all the material,
so it can be anything. Images, videos, wherever you desire is going to be always
dragged in here first. Or you can also drag it straight down here in the timeline
area here, right? So here we have the time cursor, so we can move around to see
all the different parts. Obviously, we can zoom in zoom
out like this, like this, using the wheel of the mouse, or you can just use
this nice button here. Or if you have a touch pad, just do the pinch and
we'll also do that. Finally, here on the right, we have all these
nice drawers that we can click and open only
when we need them. You know, so I can open this
and then I'll re close it. Just click, click, you know, we'll open and reclose
only when needed. And all these things
can be rearranged to your custom liking, so you can do this, right? So there's full
freedom about that. It's fully modular.
At the center, always going to see
the final result, what we're doing, right? So here is what we're going to look at most
of the time, usually. And finally, here at the time
we have some extra tools, I'm going to show you
a full exploration of all these things. Later,
don't worry about that. But for now, just make
sure about two things. First, turn this on. Snapping has to be
on along with this. So on on on here. And finally, if your computer
is not the most powerful, maybe you can go
to lower quality. So here, maybe let's go
to half a little bit. In more drastic situations,
you should go even lower. But for now, let's
go to one quarter. And yeah, so that's
fundamentally the full UI exploration
just for now. So now let's go a little deeper.
5. All The Types Of Layers (in short): To Alright, so now let's do a full exploration of all the main type of layers
that we're going to use in Arp fax and also the main properties
that you can change, right? When you work with layers,
you have two options. You either drag in pre
existing stuff, right? And that's already going to
be one type of layer, right? You're going to have
images, videos, audio, sometimes or Photoshop files, blender files, three D models. Whatever of those files
can be dragged in here, straight into the timeline, and then you can work
with those, right? But sometimes you can generate fundamentally layers
from scratch, you know, in rpafax natively. Right. So to do that, you need to right click here. And then, and you have
the first option, text. All right, so fundamentally,
just type something, for example, hello
and click elsewhere. Every single layer has some basic properties
that you can change, and those are accessible by pressing the initial
of that property. So if we press R, we have the rotation,
and we can change it. Right? Then if we press P, we have the X and Y, obviously, and then S for scale, right? Very simple. All these things can also be done in
case with a mouse, but I don't recommend
that it's very risky. So only very extreme cases,
you may need to do it. So I never use this. I think it's more precise
if you do it this way. Also, when you change a value
from here with a mouse, you have three
levels of precision. You can just drag with
a mouse like this. Or if you press control, it's going to do a fine tuning, so it's going to
increase the values way more slowly, you know, And whereas if you need to change the values
way more violently, fundamentally, just use shift. Right? So remember all these
three levels of precision. I'm going to use those.
I can grant you. And then we have
the anchor point. So the anchor point is never changed this way from
here because as you can see, kind of changes the
position of the layer. It's very annoying. But there's a trick,
very good trick. You go here, how this tool. And you can change the
central gravity of the layer, just like that without
making it move, right? So for example, if this
is going to be a ticker, we're going to place it
somewhere here, probably. So that way when it rotates, it rotates from there, right? Or if we need to center
this text fully, we can place it
exactly there, right? So now the central gravity
of this layer is there, and consequently in case
we rotate it, right? Woof makes more sense, right? So remember all this,
very important. And sometimes to reposition
the anchor point, you need to actually bypass this snapping behavior, right? So to do that, the trick
is to hold control, and you're going to bypass
the snapping only temporally, because most of the
time, in my experience, you're going to need this on
along with this and this. So that's the trick.
Final property that you can change on a
layer is the transparency. So if you press T, we have the opacity here, and obviously, this allows us to maybe overlay some layers or make a layer
semi visible on a scene. There are many uses for this. All right, so the second
type of layer is the solid. So you're rightly here, and
we have the solid here. Which is fundamentally
a basic shape that can become
something else, right? So that's what its main
purpose, you know, so we're going to
convert it sometimes to something that is completely
different at times, right? So that's its main feature. And Okay, let's call it, for example, basic material. Let's call it actually
because we still don't know what it's gonna be. So, alright. And we can make it
become a square by default or fill the
screen completely, and we give it a default color. Alright. So again, so this has, again, those basic
properties, right? Those are always available on most of the layers
like this, right? And then you can design it by using some
effects, you know, or you convert it to
something that is completely different using a specific
generation effect, right? So this can become, for
example, a particle system, or it can become a texter or a glitch source for
something else. There are so many purposes
for the solid layer, but this is the main
core functionality. This is going to most likely
become something else. It's very rare that you're
going to use the sold just for its bare initial shape.
That's the point, you know. But sometimes actually
serves that purpose, you know, so for
example, you can do it as a background color, right? So you can right click
on the layer, transform, fill complete the
frame like this, fit to comp. It's right? Now it's going to
be just simply as a serve as a background color, so you place it
actually below, right? So you know that this now the text is above
the solid, right? Remember this. So the order of the layers matters is
kind of a hierarchy. So it goes from the
bottom to the top. And obviously, the
reason why we don't see the text is white on white. So when you change the
color of the text, so you select the
text and you go to character and just change the color, you know,
maybe to black. Right. So now we start to
see the text on the back. If we place the back
on top of the text, obviously, we don't
see it anymore. So you can see that the
order matters a lot. Then we have shape layers. So shape layers instead, are a more native way
to control shapes. So it means that we
don't lose quality. First of all, in case we do very high Zoom amounts
compared to the solid, and we have special facts that are integrated and only
exclusive from the shape layer. Let me show you. For
example, we start to create it,
right? Shape layer. And as you can see, by default, we don't have any shape because
this already by default, expects us to draw
a custom shape, so it can be a rectangular
shape or a circle, wherever, or a custom shape, you know, like this, right? And then you close it, okay? And the shape is always going to be made of two components, the stroke here, and the
fill. So let me show you. First of all, we
don't see what is happening because we are
white on white, right? So why not disabling all the
layers with the I, okay? So now we start to see
the Acer shape layer. But you see what it
did? I had you press the I twice, right?
It's very annoying. Why not pressing
this button here instead This does
the same thing. So place any layer
instantly in solo. So remember these two buttons, the solo versus the I. Okay, so the shape layer, as you can see, has now
it's completely white. We can still change it.
And let me show you we can tweak the thickness of the external line
fundamentally the stroke. We can disable the
internal color like this. And obviously we now
have a empty shape. So we can customize this
fully at any moment. There's nothing destructive
in all this, you know. So you can tweak the shape, and then the main feature of the shape layers specialized
for shape manipulation. So you can then morph this
natively into something else. I'm going to show you a
layer it's very cool, and you can make very cool
things. That's the point. In short, remember,
the main difference between solid and shape
layer is that the solid is gonna most likely be used to convert
it to something else, you know, like
particles or a texture, usually, very rarely, you're going to use it just
for its bare shape, even though there is a way. I'm going to show
you even dying case. In all your DO cases,
where, for example, you need full control on the shape itself,
use a shape layer. Then let's continue
our exploration. So let's go back to here
new and we have the camera. So the camera is
fundamentally a way to control your point
of view into a scene. So for now, let me just
show you very quickly. I'm not going to
explain just for now. We're going to go deeper layer. So it just looks like this. I just layer it up.
Then I you change the perspective just by changing the position of the camera
into the scene, right? As you can see, right now,
it doesn't change anything because the layers need to
be three D for this to work. So as long as we check
this on this cube here, and we change the Z
axis of the camera. Now, whoo. You see
what is happening? We start to change
the perspective. But for now, don't
worry about this. I'm going to show
you a layer this on a dedicated chapter.
Don't worry about this. Then we have also
the control layers. So a controller is
fundamentally a null object. So you create a null
object like this, and you call it
usually controller like this or something else. The null object is
fundamentally ghost layer. Doesn't do anything.
It's just this guy here. And what it does is
that usually it just places controllers
on this layer. And then many
layers are going to refer to the controllers
on this layer, you know? So that's the point. It's going to act
like the center of control of many things
all at once, right? And one of the simplest way
that this is used usually is you just select all the
slave layers, right? And you parent those to
the controller layer. So for example,
this background a little smaller like this, right. Okay, so we have our tax, the shape, and the second
shape there, right? So let's say that
we want a way to control the position of all
these at once very easily. So you can simply select all these three layers and
then create a cable from here, parent, create a cable to
the controller simply. Boom. Alright, now they are
connected to the controller, which means that as
soon as we change any of the transform properties
of the no object. You see what is
happening? Now they are all following this. Now there's a quick way to control all them at once, right? So that's a fundamental
principle of parenting as well in
the null object, right? There is one more type of
layer that you may need, which is the adjustment layer. So this guy has a unique
feature, fundamentally, whatever effect you apply on this layer will automatically be applied to all the
layers beneath it. So it means that we place
the adjustment layer conveniently here
at the top for now, and we call it post, right? And we go to one effect
called simply glow here. We drag this guy
into the post layer. Let me show you. Woof. You
see what is happening? Now, all the layers at once are being subject to
the globe at once, so we can decrease the
threshold to detect all the colors
more like this and then spread the radius.
You see what is happening? Just like this to all
three layers at once, because the post layer now
is above earthing, right? Which means that now to exclude some layers from
the adjustment in layer, we just need to change
the position of the layers here in
the hierarchy, right? So if we want to exclude the text from this because
CNT kind of burns the colors, right, let's just put it above the post layer
here. Poof, right? Now it's excluded.
It's completely above, so it's fine, right? So, so that's the
fundamental principle. And you can identify a adjustment layer by
this icon here, you know? So little curiosity. In theory, you can convert anything to an adjustment layer
using this icon here, but don't do it. That's my recommendation.
Use always a native adjustment
layer because that is going to behave
always in a consistent way. Alright, so, and finally,
we have the light, which is fundamentally
a layer that you're going to use way less
often because it goes usually in combination
with a three D model that you drag in or with a camera
or all at the same time. So camera plus light plus
three D models, you know. So it's kind of a more
niche case and finally, same thing for here, the Cinema
four D or PhotoshotFles. These you're probably
never going to use except Photosht files. You're going to drag those in, usually, you know, or
Illustrator files. Okay? And before
closing this chapter, let's just go and
do a little recap. So these are most of the
type of layers that we saw. And remember that by
expanding each layer, you can see their full
exclusive properties. So some of these
properties are common, so like the transform, but some layers have
exclusive properties to their type specifically. So for example, the text layer has this guy here,
the animate, right? We can explore it
later, the worry here, which with all many things also here has the
text section here. That's completely exclusive
to this type of layer with the source text property
and these things here. Then we have the shape layer, which has the contents instead
available like this here, which divides
fundamentally the control of the fill versus stroke and also has a mini effect section exclusive just for
it, you know, here. And the other layers
are a little simpler, so the solid kind
of a simple layer, yeah, has the effects versus
transform. That's it. And then object even simple doesn't have
anything, transform, right. So remember to
explore these things and get yourself a
little comfortable with these principle first before going over the next
sections, alright?
6. Blending Modes & Track Matte: Every single layer
can interact with all the layers in different
ways from the metric. You know, there's one way which is to use the blending modes. Right. So for example, as you can see this tax here, right now, it kind of
disappears, right? So just put this tax on purpose, first of all, completely
above like this. Alright. Okay. So the blending modes
is a cool way to kind of make layers interact color wise or using
the brightness. So this is kind of an
experimental thing, but it can be very fun. For example, if you use the
exclusion here on the text, you know, right now,
it kind of disappears. So if the text was
actually white on white, let me change it
back to full white. Let's go here. Okay? Alright. Okay? So as you can see, now appeared again, even
though it's white on white, because now we're using the
exclusion blending mode, which kind of subtracts
fundamentally, you know, the color of this inch to the
background, right? And that's just an
simple example on how you can make layers interact
using the blending mode. Remember this, you
know, so this will now interact where is below it. Same thing for trach mate, which is fundamentally a
way to tell a layer to exist only where the opacity
of another layer is, or where the luminance
of another layer is. The luminance is just I don't know what you
say brightness. Let's say, you know,
let me show you a simple way to
explain the Shakmt. So, for example, we have a
second background layer, so just duplicate
this guy, okay? And we make it a
different color. Okay, so let's make it maybe
just something like this. Okay. All right. And then we make it bigger. So just increase,
increase, increase. Alright, and we put
it somewhere here. Alright. Okay. So let's say that we want to tell
this custom shape, let me just isolate these two. So right now, I'm talking
about this guy and this guy. Okay, so let's say that we want this custom shape to exist
only where this layer is. Right. So do that, you can do it with
blending mode sometimes, but that sounds more like a jump for the track matte.
So let me show you. You just tell shape
layer one here, which we're going to call
like crown one, okay? You just connect it
to the back layer, like this. Boom, like this. Alright. Okay. You see
what is happening? Now we start to have
the crown visible only where the intersection with the er layer is
fundamentally, right. So it means that now if I move the crown layer
around Poof, right? That's the point. And there's also obviously this kind
of a default thing. You know, by default, kind of
hides the modulator layer, but maybe you want to
keep this visible. So let's keep it visible,
actually, like this. Alright. So now, when
we move the crown, we actually see it
appear, you know, as it intersects the background layer
like this, you know? That's the point. And you
can also do the opposite. You can move the back
che layer like this, and it will make that way progressively reveal.
The other layer. That's it. You know,
so you can make layers interact in very fun ways. And you can also invert
the behavior like this. So it's going to now disappear only where the intersection between the two layers
happens, right? So hide now back to you, right? In that way, we can
have this effect. We can make the layer
disappear, right? And these principles are
extremely important, and we're going to use those
for er. That's the point. So make sure that you make some practice with these h.
7. Motion Graphics (Basic): All right, so in this section, we're going to maximize
your knowledge of motion. By motion, I mean not necessarily
moving a layer around. So let's make a empty
shape like this shape. Let's make 512, five, 12, so it's going to be square. Okay? So by motion doesn't mean moving
this around. That's it. I'm going to show
you a technique that you're going
to be able to use for any property you
want in general, right? So it means that you can
engage also rotation. Or what else? Scale, position. We
already know it. Or weirder things, for example, opacity and effects properties. So effects, for example, have sometimes
percentages like this. So this technique I'm going
to show you is going to be valid through anything you
want, not just position. In this case, we're
going to use it for simplicity on a jump animation, but remember this concept. You just need to learn
the technique at its core and then use it for wherever you
want in your projects, alright? And we're
going to go in order. I'm going to show
you the basic method first and then the
advanced method, alright? It doesn't mean that
the basic is bad. Actually, you should
know it as well because sometimes
it's convenient to convert to the advanced one. And right, so let's go. Let's make this square jump. So we're going to engage for
sure, the scaled property. In the rotation and
position at the same time. This is a very classical
type of need in. In most cases where you
need to move stomping, you're going to engage
these three things. And again, remember
that you can use this technique for
whatever other property. Alright? So let's go. Let's make this square,
first of all, appear. So let's choose
our initial value. So let's say that it's
going to be not that big. Like this, alright. Okay. Even before that, choose the anchor point. The anchor point is
this thing here, the middle of the object. You can change it
using this icon here. Make sure that these three
buttons are enabled. And that way, you can snap the anchor point to any
point of the object. That will become the center
of gravity of that element. That way, it will scale
up from that spot. You know, sometimes
you need that. In this case, let's keep it at the center, so
let's keep it here. Okay. So, let me make
a little bigger. So, let's say that it
goes from 0% to one, two, three, four, five,
to that size, okay? That's how it appears. Okay. And then we wait for, let's say, about 20 frames. And here, we're going
to start the movement. So let's make it jump
first. So for sure, we want to make
it go vertically. So let's go one, two,
three, four, five, six, go higher, jump around here, okay? And then return. So say about 20 frames later, copy this keyframe, paste. Okay? So let's see what
is happening so far. So it pops in, jumps and then returns here. Alright. Okay,
let's also make it rotate clockwise while it jumps. So we go back to this
keyframe, right? While such you jump. Let's place the rotation here, okay, and then at the end of it. So when it lands here, let's make it rotate 360. So this first number here
means one full resolution. So if we increase it
by one, it will do one clockwise, full 360. Boom. Wow. You see what's happening? Alright, so let's start
you look this animation. Let's see what is
happening so far. Okay, boom, boom. Alright, that's a
masterpiece. I'm kidding. Alright, so and then we
start you enable motion. Motion blur is this
jack box here. Alright. And let's see. Okay, it looks terrible.
Alright, let's see how we can make this
look a little better, right? So first of all, we need
to use busier curves. So after you have chosen just the bare coordinates of each property during
your animation, you need to use
most likely curves. So to do that,
select everything. And then press F Fini like this. Or if it doesn't work, you can right click on
all the keyframes, interpolation, and then
change linear to busier. Boom. Alright, that's
slightly better. Let's also make it pop out
at the end of the animation. So we wait slightly here and then copy and paste
the scale, right? And also we flip it
because IIS is going to just go to zero instantly. So we right click on
these two keyframes, Keyframe assistant.
Time reverse keyframes. Boom. Alright, so that kind
of flips those two values, so it will automatically
close back to zero. So let's see what is happening. So it pops in. Oh, yeah, I
think I forgot something. Let's We start this with F nine. Okay, let's see.
Boom. Boom, boom. Alright. Okay. So we should
have a little cycle. Alright. Let's see how we can
improve the animation. So first of all, it's
still you linear, right. So we need to actually
make it effect, like, boom, boom, right? So let's tweak the
rotation, first of all, so let's select
this key from here, the rotation, and we go
to the graph with us. Okay. And let's make a
lot most ways. All right. As you can see it's still
linear, so let's make it. For example, let's push. When you open this
for the first time, the graph editor, make sure that you double check
these settings. Click here on the I and
copy my settings here. Okay. Then go here. Make sure that it shows
only the value graph, not the speed graph. Avoid this like anything else. So enable only this
value graph. Okay. Okay, that's just copy my settings here,
what you see here. Okay? So, this way, we can make exponential curve. So for example, we
can kind of make it rotate a little more at the
beginning of the jump, okay? And then we kind of make it accelerate slightly
around the end. So we start to obtain
this, boom, boom, okay? Or just simply this just Okay. Okay. And then let's do the
kind of the same thing on the on the jump on the Y on
the vertical movement, right? So we squeeze this value down, so we'll jump higher, right? A little more violently
at the beginning, boom and then we can do an exponential decrease
exponential, line. Okay, let's see. Boom. Boom. Okay, it's still not that already starts to feel
better as an animation, but still not as realistic
as you can see, you know? So let's see what we can do. For example, we can add a
slide bounce at the end of it. So here, at the very end, we can actually go
maybe two frames later. We copy this keyframe, right? The final landing white
position, copy paste. And we go to the
previous keyframe, this, and actually we push this slightly lower. You
know, slightly lower. That way, it will
kind of bounce. Let me see. Boom, boom. You see what happens?
Can I add that detail, boom, right? So, very important. Okay, but still not as
realistic as we want. So why not try to smooth these
three key rams even more. So we can try to right click
right click on all three. Okay? Interpolation. And
we use continuous bezier. Continuous bezier,
as the name says, we'll try to make a
continuous change instead of making a
manual busier, you see? Alright. Even better, okay? But
still not the best. And this is kind of the limit of the first method of
animation of the properties. So that's the thing. You know, it kind of feels
still kind of sequential. That's kind of a hard limitation
of this first method. It's not your fault.
It's not my fault. It's just the way it is, right? So sometimes this is well enough to make any project to just solve your
animation needs, right? But if you need superior
smoothness in your animation, we need to escalate
this situation and go to the advanced approach, right? So, be careful because
we're going to go slightly more complex
now, alright? Let's go.
8. Motion (Advanced): Alright, so first of all, now let's duplicate this
layer here and let's call the old one just
backup or just old. Okay? Let's make it
slightly transparent, like this so we're gonna keep it there just as
a comparison later. So for 'let's disable it. Alright, so now we
have a new one. New. And so let me show
you the principle. So to achieve true
smoothness in animations, you need to not animate
the element directly. That's the first thing,
because the core limitation of Afrofax in actually any animation program is that you only have one
lane per property, which is still good. You know, you can make pretty
much all the stuff that, but you are limited with a sequential sequence
of keyframe, you know, so you
can do only A, B, CD like this, only from
the left to the right, right, and only in one lane. But there's a way to override this limitation by fundamentally having multiple streams of keyframes on the same
property at the same time. So that way, they
average each other, and that's how you reach a
more realistic smoothness. That's the principle
we're going to use. Alright, let's put
this into practice. First of all, let's delete all this animation. We
don't need that anymore. Let's keep it down all
the as a reference, so let's delete this, delete everything like this. So not animate the
element directly. Instead, we're going to
create a controller. So right click New no object. So the null object
is the purpose of this is to just
control other layers. That's what the
null object does. It looks like this, a
simple square like this. So it doesn't have any meaning. It's just a simple
controller, right? So first of all, let's
call it Control one, okay? And let's also calibrate
it calibrate it, right? So there is a way to fully synchronize the
anchor point and the position completely to our
actual object here, right? So the trick is to simply
use this cable here, the parent here to the
actual object like this. But before doing this, hold shift the whole
time, like this. Alright? So cable, shift, and then
release the mouse, Okay? And finally, release shift.
Okay, you see what happened? Now the square is fully synchronized with
our object here. At that point, you can
disengage the parenting here. And instead, parent our actual
object to the controller. So in this time without shift, so simply the cable. No shift, Ls, boom. Alright? So this is called calibration. So now the controller is fully
calibrated to the object. And let me show you
what is happening. Now, if I rotate the controller Poof. You
see what is happening? I actually moving
exactly the square as if we were just controlling
the square natively. Alright? That's the reason why we did this. Do this always. Let's start you animate now, simply the properties
on the controller. So rotation, Y, scale. Alright. So for now, well, you
already know this. So simply go here, zero, 32. So we wait 20 frames here,
we can search you move. And that's now the
first difference. Instead of doing the
entire animation, on the position on
the controller, we only do one movement
per controller, right? So here, it means
that we're going to do only the upward
movement, right? So we go here, and we do
about 20:20 frames, yeah. And we move those up. That's only one moment. Alright? One moment per controller. Okay, so goes up. Let's do the same
with a rotation. So rotation goes from
this to four rotation. The rotation doesn't need too
much too many controllers. So let's do a single
controller in this case, yeah. Okay, so it goes
there. As you can see, so far, it's only half of the
vertical movement, right? Okay? And then let's start to enable the Bzire curve
on all these key frames. So have nine. Then we
select the position. Let's go to the graph, and we push like this. So we go up. Fast like this. And
then same thing here. On the rotation, we just do similar curve as
earlier like this. The rotation doesn't change, really, like this. Okay? The trick is just in the y. Okay? At this point, we have only half
of the movement. You're probably wondering why? Well, simply we need to do
one moment per controller. So at this point, we go back at the beginning
of this moment. So here, and we duplicate
the controller. So now we have controller
number two, right? And let's pull up the
keyframes, pressing, as you can see that's the
shortcut and we need to clean the keyframes that we don't need on
the controller two. So these have been used, so we don't need those anymore. Rotation, also, we don't need that, and the position as well. So actually let's delete
everything like this, okay? And before doing anything else, we need to parent Control one to control
two simply like this. Boom. Alright? So as
you can see, so far, nothing changed, but there is simply one thing
that happened now. Let me show you
see what happens? Now we can still keep controlling the square from
controller two as well. So this is the
fundamental principle. We have a chain of controllers, and each does one moment, only one, you know, one movement
per property, you know. Okay, so let's go here. And as you can guess,
this is going to do the pullback movement. So we enable y
position here, right? And then it's going
to push the square down in place around
here. Okay, let's see. Let's pull up the old let's
see where it actually lands. Okay, so it's actually
there, right? So let's pull this down here. Okay. Oh, sorry. Let's go here. Okay, that's the
actual position, so let's just override it here. Okay. To use the snapping
in Affaxino hold shift. So normally it goes
free hand mode, but if you hold shift, it will
snap to the closest event. So remember this. Okay, so let's disable the
old one for now. Okay. So that's
what is happening. While the Control one is pulling is pushing
the square up, at the same time,
the control two is pulling it back down
in place, you know? And fundamentally, these
opposing forces are averaging each other
in a natural way. So final strokes
youuld now enable the zero curve also on
the pull back, right? Let's go to the graph, and
we need to do the opposite. So here on the push up
moment we did this curve, so kind of violently
push it up, right? Boom on the pullback Animation, we actually need to do a
exponential curve like this. So I will actually wait some time and then
pull it back in place. Alright. And what else? Let's check this out for now. Let's see what is
happening. Let's also close it back at the end. Finally, we can
close the square, simply go back to 0%.
Let me check that. Oh, yeah, obviously, we
need to close it from the layer that is centered so we can close it from the
native one, actually. Okay, I think we can also
do from the Control one. Yeah. Control one is calibrated. I just forgot to
calibrate Control two. So if you need the scale to
work from Controller two, we would have had to
calibrate it as well, as you can see, because
the anchor point is not calibrated.
But it's fine. Just go back to a
previous controller that is fully calibrated. In this case, Control one. And then you can
close it from here. Like this right
click, flip Boom. Okay. And Okay, let's see
what's happening so far. Boom, boom, boom. Woof. Okay?
And it's still not over. There is still one more
thing that we can do. In Fairfax, there is a
area for the Asi Shu enable additional features on a property that Asi Shu
use last keyframes, but still trigger animation. It's called descripting
or expressions. In other words, I'm going
to here from now on, I'm going to give you some of my personal expressions that
I created over the years. And I'm going to show
you how you install those and how to
use them, alright? So they will appear here in the effects section and
just search keywords. So the first one is
called overshoot. Okay, just type overshoot here. Overshoot is the
property of element when kind of stops
in real life, right? And instead of
stopping drastically, it will actually
have a slight bounce back kind of recoil, you know, like, boom, right? Boom. That's what
overshoot does on a property will make it aware of the momentum that accumulates. Very important stuff, right? To install this, you
need to simply select a property that has key
frames. That's the trick. So in this case, this guy
here and then go into animation and then save Okay? And then you will automatically
make this folder pop up. As you can see, it's
called user presets. This is where all the scripts, all the scripts that I will
send you will go, too. So you need to get the file. You see it right now
from the resources of the squares and paste it
into this folder here, right? Let me show
you where it is. Overshoot here.
Alright? So paste that introduce folder here, user presets, and then
just reboot Afterefax. After you do that, when
you search overshoot, you will see this file here. Alright. So, okay, that's said. Now that we have this,
we can simply select the property that we need to make it momentum
aware, right. So in this case, Y position and double
click on overshoot. Ff, the property changed color. And that's why now
has the bounces Poof. You see what is
happening? Boom, boom. All right. Okay. And this, let me show you
what is happening. We can see this behavior
graphically by going to the graph editor,
select y position. And when you double
click it, you will see this button here. The As you see what
the expression is doing after the keyframes
is done working, right? And obviously,
this is responsive to how much momentum
you accumulate. So obviously, if you're
writing very slow, there will be no bounds and
the other way around, right? So, now you're very
free to do this. Okay, and you can also tweak
the exact type of bounce. So when you double click a property that
has an expression, you can see its
source code here. In this case, we have a couple two numbers
that we can edit. We have frequency and decay. So frequency is the
tension of the spring. So ten means ten times a second. So let's actually make it a
little slower, so maybe four. And then decay,
the higher it is, it will bounce for
the least time. So let's make it
bounce last, maybe 20. And now let's see
what is happening. Okay, let's see. Boom. Alright, let's see
what's happening. So that's fundamentally
now a single bounce. Okay. And then let's add the same expression
to the rotation, right? So let's right click on Y
position, copy expression only. And then let's go
to the property that has the rotation animation. So let's see. Let's see
what it is. Here is. Okay, so select
rotation and paste. Okay, now is FV is
also rotation aware. The momentum is also applied to the rotation. F. Simply at that. Boom. Alright. You see how more smooth this
type of animation is. So let's actually
compare it with the previous method.
So now we do count. All right. Yeah. Let's pick
it places side by side. Let's go here. Okay? See how more stiff
the first method is compared to the
second one here. All right. So that's
my precedent for you. So this solution is
very important to achieve ultimate smoothness in any project you desire, alright? So the core principle is this. When you need simple animation, the first method is fine. If you need ultimate smoothness, you need to use multiple
control layers, one for each movement, right? And each one needs
to be calibrated. And then, make sure that
you average the values simultaneously so
that they will kind of achieve this
level smoothness. Seems like that. And yeah, just play around with
this and have fun. Alright, and finally,
let me show you a practical example of an application of this
technique, right? So, this is a
recent thing that I did using this technique on
D objects. Let me show you. All right, so this was
for my new sound logo. As you can see, I really
use this technique. So I switch objects left,
right, left, right. And to achieve that smoothness, I really use this technique. Let's go here, key frames. Let me show you a little bit of the real projects behind it. So let me go here. All
right, here we are. Let's go. Let's go
into the main scene. Alright. So as you can
see, I switch objects. And you see what
is happening here? I have all the controllers. Right? And it's really the same technique
I just showed you. So, fundamentally,
I go left, right. And for each object, there are at least two n objects that counterbalance the forces. So there's one that pushes the object to the right,
and at the same time, there's another one that already pulls it back to
the left, you know, and same thing for the rotation, you know, and then obviously there's other
thing going on, right? But the core principle is this. So just to show you that
with this technique, you can then use it and repurposes infinite in
infinite ways, right? So as long as you learn
the core of this, then you can do
whatever you want in real projects, alright?
9. Morphing (Basic Method): All right, so when it comes to morphing a element into another, there are different
ways to do that, right? And which technique you
need to use depends on how complex the morphing
they need is, right? So let's go for the first
case, the simplest one. So if the shape you need to
morph to is pretty simple, kind of geometrical,
you can do it internally into the element
itself using masks. So we have the the project
that we create so far, right? So let's just select
our shape here and we're going to double click on the rectangle tool up here. That way, it will automatically
create a so called mask, which fundamentally is the area that allows us to see only
one part of the image. So right now, we have
this thing here. Let me show you happens.
You see what happens? I can also edit specifically
one of these points where the shift clicking on one of the edges so you
see what is happening. So now, for example, it's kind of like a boomerang, right? Right? And this can
be seen from here. In case you close it, just press or double M c, and you can feather
it like this, you know, we can expand it. Am I? So, but one
of the features is to morph the shape
dynamically here. So we animate this
property path. So we turn one keyframe on here. So now we have this
new shape here. Actually, you know what?
Let's go back in time. Okay? And we actually place one
keyframe with this shape. So the standard
rectal like this. Boom. Alright.
Now, let's pull up all the key frames
so we can start to go exactly before the jump. So we go here. And the first shape
is going to be here. Let's color it differently. Let's go here. Label yellow. Alright. Okay. So
then let's go one, two, three, four,
frames later, right? And here, we're going to
just draw our costume shape. So we go here and just
click on one of the edges. And for example, we morph it
into a boomerang like this. Okay. And then we go back
just before the landing. So we go the landing
is here, okay? So we can simply
cut and paste these two and reverse them
as we already know. So we go here, reverse, okay? That way it will morph into a boomerang and then
morph back into a square just before
the landing, right? So let's just compress
these things. And that is already
the first way to morph shape into
another dynamically. Woof. You know, Woom Woom, wom. Alright? And you can
choose obviously how fast the morphing happens just by doing this,
this, and this. Let's go here. Now it's
more gradual, right? And you can use
Bessie as well to make it even
smoother. Let's see. All right. And that's
the first case.
10. Morphing (Allaround Method): If we need to morph to
a more complex shape, like for example, this
nice alarm clock, let me show you here. Obviously, we cannot
recreate manually a mask that perfectly recreates
this guy here, right? So we need some kind of
more automatic way to morph from shape A to shape
B more easily, right? So I'm going to show you
an easy way to do that. So let's go here to
our main compensation. Let's delete the first
case like this, right. So it's just a simple
square, alright? So first of all, we need
to prepare the image. So we right click on our image
that we need to morph to. So the clock this. Let's drag it here. Okay. And first of all,
Gase has background, so we can remove it with
extract, like this. So in case we can remove easily the white background like
this, right, like this. And let's also filter just the
things that we don't need. So we use the pen tool,
simply remove maybe these lines here and the floor. Alright. Simply click, click. Remember to close
the shape, and boom. Alright. Okay, so, and finally, let's reverse the color. So maybe just extract. Sorry, we need to use fill, and then just white. Okay. All right,
all these things. And finally, applying nice glow. So we maximize the aesthetics of the object, first of all. Always important. Alright.
Okay, now it's ready. And so how do we morph from
shape A to the clock, right? So the first thing
that we need to do is to have a way to perfectly synchronize the clock
with a shape so we can make them do exactly
the same movement, right? So one way to do that is to
start to use precompositions. So we can simply pack the
shape into a kind of a folder. So it's called precomp. So we right click on shape. Precompose. And then just
select the first option. For now, don't worry
about the second option. Select leave here. And then let's call
this shape one. Alright, and just confirm. Okay. So as you can see, nothing changed, still
perfectly working. Everything is fine.
The only thing I changed is that now we
have this icon here, which means that we can
double click on shape one, and it will lead us to a dedicated timeline
for this, you know, where we just see
the bare content, you know, like a steel
frame, you know? So, fundamentally, when you do a precomposition
in leave mode, it just serves as a
place holder element, which means that now, if we duplicate this guy, Okay. And then we also right
click on this new copy, reveal reveal layer source. It will automatically select
it here at this level, and then we duplicate it as well here and now it
becomes shape, okay. So make sure that
the copy here is selected and shape two
is selected also here. Then you hold option
if you are on Mac or Alt and then drag slowly, shape you into the copy here. And you would see
the confirmation by the fact that
the name changes. So now what happened is
that, let me show you. For now, everything
is fine again, okay. But actually, now we can
change the content of this shape independently
from shape one, right? So, in other words, now we
can just cut the clock here. Okay, and we paste
it inside shape too. So we let this paste, our new shape, so
we just center it. So right click Transform
fit to width here. Okay, and then manually
tweak it to this okay until it's fully
centered as much as possible, maybe a little bigger. Okay. When you move values with mouse, you have three
levels of precision. When you use just the
mouse, it's just level one. When you hold control, it will actually move kind
of a fine tuning mode, you know, more
precisely like this. Or if you need to increase the numbers significantly,
hold shift. Okay. Alright. So that's it. Okay, that seems fine. So let's go outside of here. To move between the
coosons as you can see, I'm using this mean technique. I'm pressing Tab to do that. So when you press left,
it will automatically go to the surface
of the iceberg. This is especially important in real projects because
you cannot rely on these tabs here because they
will start to accumulate to maybe lose the
point where you are. So if you do that, you can
orient yourself really easily. Okay, that's it. Let me
show you what is happening. Woof. You see what is
happening? It's really working. We really have just
the clock doing exactly the same
animation as the shape. Alright? And that's where we're gonna now we're ready
for the morphing, right? Because now they're doing
exactly the same movement. Okay? Maybe we can make
the clock a little bigger. So let's see let's make it. Let's make the square
slightly transparent. Yeah, maybe we can make it
a little bigger, right? So let's go back inside and we make it a
little bigger like this. So now there's an issue. If I go outside of the boundaries
of this square, it will crop the clock, right? But there is a way to preserve
the information that goes outside by enabling
this button here, this button here on
shape, let me show you. And obviously, we
need to disable the mask because we
don't need it on shape. So right, so that way, we preserve the
full information. Okay. And if you wonder if there's a way to tweak
the size of an element, while watching how it looks in the final result,
yes, there is a way. Sure, you lock the
composition view here, and then you double
click on Shape J and then we can put these tabs
side by side like this. So here we have the
final result on left, and here we are just in
the local timeline here. So that way, we can resize
it while being able to assess how it looks like
in the final result, right? So we can easily choose a more profit size or maybe
something around this. Okay. All right. So finally, then just
delete this and re unlock. Okay. Let's go back to the
surface. Okay, that's it. The size is better. So let's
go back to 100%. All right. And now we're ready
for the morphing. So for the morphing, we need to use an
adjustment layer. So let's compress
all this stuff. We right click here
New Adjustment layer. Okay? And we're going
to call it morpher. And here we just need
to use two effects. One is a Gaussian
blur. Double click. And then a median
is called median, like this. Double click. Alright, now we're going
to turn on two keyframes, one on the blurriness amount,
and one on the radius. Okay. Let's pull
those out like this. Make sure that the adjustment
layer is above everything. The adjustment layer
has this you know, I already told you, so it will apply the effects that it has on everything
that is below it, so to these guys, right? Okay, that's why it's here. Alright, so let's search you
put these two keyframes in sync with the
beginning of the jump. So here. Right? Boom. Okay? So obviously, we need to fade out from shape one
and at the same time, fade in to shape two, right? So let's also do the fade. So we go to shape one, and we do a simple one, two, three, four, fade out. Boom. And then we copy
these two keyframes, we paste to shape two, and we reverse those, right? So fade shape two will fade in, so they will cross fade, right? One goes out, one goes in. Let me show you
what is happening. Boom, right? Boom. And also, when we return, we
return to the square. So let's just copy this again. Yeah, actually, well, at this
point, we can just do this. So this can go here. So it fades back out, and this is a fade in, so we can use it to
make it fade back in. Boom, you know, so
as you can see, when you make a project, you can recycle more and
more stuff as you go. Alright? So let's see
what is happening so far, fades out, fades back to square. Alright? Okay, so far so good. Then let's you make the
actual morphing, right? So let's go to the morpher. And the way it works is that
you increase the blurriness. Like this around go kind of in the middle of the
cross rate, so around here. Okay? And then increase the blurriness to
a certain amount. Let's say, let's go
with maybe maybe 20, something like this
Through enough, okay? And then median, increase it. Remember to enable this button
here, operate. Boom. Okay? And then increase slowly. You see what is happening?
Now it's kind of morphing the two shapes. It's kind of merging the
edges of the shapes. That's the cold thing about this techniqu just
around here, okay? And then just
return, choose zero. So copy this paste, right? So kind of does a more gradual morphing between the two shapes
with this technique. You know, of. And then just use bezier always
in these cases, then copy all these six and repeat just when we return
to the normal shape. So we go here. Boom, boom, you know? Let's make a shift. Slightly. Let's see. Boom, boom. Alright. Okay, let's see
what we have so far. Oof. Alright. Okay, fair enough. And if you really want to
add even more smoothness, we can enable a motion
blur on top of this. You're probably
wondering why because we already have a
motion blur here. So the point is
that these motion blurs only work when we animate the transform
properties, right? So but here, there's a way to trigger motion blur brutally, even though we're not physically animating the
transform properties. Right. So it means that
these values here, in theory, are creating additional movement that may
be smoothed by motion blur. So why not adding an additional adjustment layer
with the motion blur Okay. And we disable the default
motion blur for now, so we disable these guys. So we need to use two
motion bs at once, right? We use a single one at the top of everything.
I was more efficient. So we go here and we search. There are two motion
burrs in our ffex. There's force motion blur
or pixel motion blur. Pixel is higher
quality but heavier. Force is the opposite, so it looks slightly
worse but more efficient. So let's go with this for now. Okay? And to see what it's doing automatically it's
detecting the moment. Let's increase these
maybe 30 tube. Okay, let's see. Let's
give sag a couple seconds. So you can see it's
heavier Now's render Okay, in case you can
decrease the quality here to have faster review. Okay? Let's see what's
happening. Woof. You see what's
happening. So now we start to have a very
decent morphing between two shapes, you know? And that's now a more
advanced approach, right? So when you go
from initial shape to a more complex shape, this is the way to
morph to where we want. Finally, the cool
thing is that now, thanks to the precompositions, we can at any moment, change the shapes to
wherever we want. So now you're not forced to shape one should
be just a square. You can do whatever
you want. It can be even a second image. Hey, for example, we have
this Xbox controller, so we go inside, shape one with lead or with disable the old
shape, drag this in. Sanford. I was always a little here, maybe
a little bigger. Okay, and now it's already
working. Let me show you poof. Boom. So this will automatically
morph into shape you. Just like that. Woof.
You see? That's it. Works like magic. Alright?
So yeah, have fun.
11. Morphing (Advanced Method): Oh. Alright, so in case you wonder if there
is an ultimate version of this technique,
the answer is yes. This version doesn't require a motion blur or a morpher so we can
start to disable these. The disadvantage is that I slide more complex, even more complex. But if you're ready, I'm
going to show you this. Okay, so let's go to here. Okay? We don't need shape
anymore for this version. So we just copy the alarm clock here
and we delete shaped. We don't need it
anymore. And we just paste it natively
inside shape one. So just paste it here for now. And let's make the alarm
clock a little smaller. Alright. Let's kind of equalize the size of the two elements
a little bit. Alright. Okay. That's fine. Maybe the controller
can be a little lower. Like this Alright. Okay. So this will allow
us to natively morph between two images in the most close way
as possible, right? So fundamentally is
not compromises. Alright. Okay, and
so now we are here, and the first thing
you need to do is to normalize the scale
of the elements. As you can see the
scale properties right now, it's different here. So once you find a way
to make the scale, we become 100 for both, right? So to do this, simply precompose these two
layers separately. So control shift sim. And we call this controller. In this case, let's use
the second option, move. So confirm. Okay. And now the scale becomes 100.
Same thing for this. Here and confirm. Okay, 100 100. So now they are normalized. That's how it set. Okay, so first thing is to now
gather a mask that perfectly replicates the
outline of this first shape. Alright. So we select
the first shape, and we go to layer auto trace. Okay, make sure that the
background is removed. In this case, is ready removed. So we can simply skip
to this step here, and we need to detect the Alpha. The Alpha is the transparency
of a layer, right? So since we know that
there's no background, they will automatically,
right now, detect the outlines, make
sure that preview is enabled, and copy my settings. Usually these
settings are pretty good you ready recognize this. Okay? And then just confirm, make sure that this is disabled, and then copy my
settings and confirm. Alright, so now we
have a perfect mask that replicate the outlines
of this layer, alright? Okay, and that's
step one. Let's do the same for the
clock. So we go here. Know, simply. So now
here's an issue. The clock is a
little more complex. You know, you can see has many holes everywhere, you know, so there's no way we can summarize this shape
with just one mask. It's gonna probably create
multiple mask, you know, one for this piece here, one for the middle,
one for the center. It's gonna be very,
very hard to manage. So one way to brutally make after effects detect this as
a single shape is to use, again, this effect
here, median here. So use median on the shape. Enable this check box and
then slowly increase. So it will kind of merge the
edges as much as possible. So find some kind of compromise. You know, for example, okay, this search will be
useful because right now it's merging these edges, so it's going to be
detected as a single shape, and kind of increase
this number when it such you detect most of
the entire shape. That's the sweet spot, alright? Don't worry about these
elements at the center. Right now, most of the images detected as one shape,
and that's what we want. Alright, so let's
keep it this way. And now we can just go, again, on layer auto trace
and that's it. I will already use my latest
settings and just confirm. Okay, and that's
our second mask. You know, I will probably do
two masks, but it's fine. Sometimes there's
false positives, so we can delete
mask number two, and that's our single
mask, alright? Okay, that's one. And now that we have
these two masks, let's copy the
mask of the clock. Let's call it clock. Okay? And let's call this
controller X box. Okay. So copy the clock mask and paste it into the
controller layer. Boom. Okay, so now both masks
are on the controller. Let's go back to
the controller now. Okay. And at this point,
we're almost ready. We go to one effect
called reshape, okay? And apply it on the
controller layer and change the source
mask to mask to Xbox, obviously, Destination
mask to clock. Okay? And now what
happens is that as we transition with this percentage 0-100, let me show
you what happens. If the shape morphs
in a weird way, when you increase this number, it means that simply
you need to re tweak the correspondence point. That's how it's called, which is this line here, you know? So simply go to one of the edges that you see here when you see
this little circle, and now it's ready to calibrate
the correspondence point. So one way to do it is
to go to 100%, okay? And then simply move the correspondence
point probably we need to move this
guy up exactly here. So until the shape
remains calibrated, you know, to the right
verse, like this. So now what it means that now it should be better because
now if we go exactly, you see what happens, it brutally morphs from the controller to the
outlines of the clock, alright, which is what we want. Now it will natively morph
shape one into shape two. That's the point.
Alright? So now as gas, we need to animate this guy. So we enable the
percent automation. Now it's ready to go here, and we go from 04
frames, and then 100%. Okay. And that's the first one. Alright, so now,
as you can guess, we need to simply
do the same thing in reverse to the clock. So how do we recycle all this? Well, let's delete this copy the two masks on the clock here, and then simply copy also
the reshape into the clock. And then just simply reverse the key frames
of the percentage. So just select these two. Reverse. Okay. And let's see what is happening
on the clock in three, it should be the same thing, but in reverse, if there
is extra deformation, it means that simply you
need to swap the two masks. Let's go here. Okay, let's see. Eat is going from the
controller to the clock. Eat is going to the revers. And that's it. And then now it should
be pretty much ready. And then we can also do the fade as well, between the two shapes. So we go here and
controller fades out. And the clock does the opposite. So instead of copying and pasting the same
keyframes here, we can right click on
opacity, edit expression, and then just type 100 minus and then make a cable
from here to the opacity. And now it will
automatically fade in as controller
fades out, you know, so that's a little scripting to avoid copy and pasting too
many keyframes, you know. So at this point, we
can also copy and paste these keyframes,
reverse, Okay. And let's see what
we have so far. At this point, we can search time all of this appropriately
with the jump animation. So we go back to the
base timeline here. So the jump starts
exactly here. Okay. Let's go here. Okay.
That's correct. And then the jump ends here. Okay. So we go back
here and go here. Alright, so let me show
you what is happening. So this fades in we
have the controller. The controller fades
into the clock. We don't need a median effect anymore on
the clock, right? So we go to the median on
the clock and we disable it. So it's just a normal alarm. So morph into the clock, and then then I forgot just
the percentage. Let's go. We also reverse the
percentage, obviously. So Let's use the same
trick for the percentage, so we can disable
these keyframes. Don't need those, we
can use scripting. So simply 100 minus cable
to the percentage here. Boom. So now we can control
everything from here. So copy and paste and
reverse. Boom, boom. Right? So now, morph into the clock and then back
into the controller. Who, you know? All right. And at this point, the result is that we start to
have the controller. Let's go heal bigger. We have the controller natively
morphing into the clock. We need to at this
point, disable the fades at this level here. We don't need these
fades anymore. It can remain 100% all the time, it already internally
fade inside, right? So let's see
what's happening. Wow. You know, that's the cleanest type of morphine that you can
achieve in Afterfax, you know, because,
as you can see, it's now natively morphing
between the two shapes, you know, there is no
flaw at this point. And, yeah, that's
my present for you. And this method doesn't require
additional motion blur, so we can say goodbye to this. Doesn't require the morpher. The dual disadvantage is
a little more complex, right, because you need to
prepare the two perfect masks. You need to morph
those. But if you want the flawless morphing,
that's the way.
12. Precomposing: Precomposing what it is. So very simply is a way to make a layer
become a placeholder. That's the main feature
of precomposing. So for example, we have this
object animated right now, and we want to be able to
change it with a custom image, video where we want layer
because we still don't know what we will use for
this animation, right? So to do that, we simply right click on the
element, precompose, and then just leave all then in case you can give a
custom name again like this and then just confirm. So that's how it looks like. And let me show
you, first of all, all the key frames, all the
effects are still outside. We need to do this.
So in this case, there were not effects, so there was no but let's
add one again on purpose, Latus so that's the
main bare minimum of precomposing simply making become a place holder, right? Which means that now I
can just simply go inside here and I can change this
with something else here, and I drag it here, and as soon as I go outside, it will instantly change the square with this
guy here, right? So let's let me just remove the background
super quickly like this. Alright. Okay, here's. Let's go. Let's go to Cosco. Alright, simply
like that, right? So the same animation
now is applied on whatever image I put
inside the pre comp. That's the way it works.
So you can also work on this dedicated timeline
to do whatever you want. You can design the object
better on this frame here. And then as soon
as you go outside, they will just apply the
animation. You know, that's it. So that's the first use of pre compositions, place holders. The second feature is
syncing multiple objects. So obviously, if I
duplicate this precomp and I shift it now we have
two of these guys, right? And I can still show
you separate changes. I, for example, mirror
the second one. So I unlink this and I flip the horizontal
axis with minus. And obviously, now, it's
going to look different. The second one is flipped or I can change the size
differently, right? So we have separate
control on these things. And right now, the content of these
precomps is synchronized. It means that as soon as they do any change to either
of these two, it will change for both. So if I go back to the
square and I go back here, as you can see, both
return to B at the square. So right now they
are synchronized. Sometimes you want this
behavior to be like that, you know, especially
in templates. So this allows you to change
like 20 layers all at once. Just by changing the
content in any of those. So that's one of
the uses of this. But sometimes you
may need to have a separate control on the content of each
composition, right? So to do that, you
need to first of all, duplicate the pre comp at
the project level from here, not from the timeline, because
if you duplicate it just from here, it's
still synchronized. So to really detach it, you need to right click
on the composition here, the first one, reveal. Reveal the source, and it will automatically highlight
it here on the project, and then you duplicate
it from here. Okay? Now, this here is
a true independent copy. Means that if I double
click on shape Q here, now I can change to the
beatboxer image here, and then I go back to
the composition to the main comp here, right? So at this point, all you
need to do is you select the copy here on the timeline and select
shape Q in the project view, and then keep press and
overwrite on this copy here. You will see the name changing. That's the confirmation
that I worked. And now, now we have actually independent
content on both containers. And now we can do what we want. So if you resize a layer
and you make it go outside of the boundaries
of the precomp. So for example, if you make
this guy outside like this, when we go back to
the main composition, as you can see, will
be cropped, right? So to preserve the information that goes outside
of the boundaries, simply click this check box on. So shape two, now
click this guy. Boom. Now it will preserve
the additional information. And then you can further
tweak it like this. We can do what we want, alright? You're probably wondering when
you precompose like this, total shift C you have also
a second option, move all. So move all is a more
drastic precomposition. Fundamentally, it
will make AffecT that the content has been
pre rendered, right? So it will kind of pack
everything inside. Also all the keyframes, all the effects, everything, all the properties will kind
of return to default values. So let me show you
if I now select these two precomps
and re precomp them with move all on here, and I call this all
and I just confirm. Now, let me show you
this new pre comp doesn't have any key frames. So I'm just pressing
you, as you can see, there is no keyframes,
nor ex. There is nothing. And all the values
return to 100% just standard default
values, right? So sometimes you may want
to have this option. To kind of normalize
something, you know, to make it become almost
as if it was pre rendered. So, especially
with the technique that I used this earlier, if you remember, with a reshape with a morph technique, right? Because I needed to normalize layers to kind of pretend
that both were the same, both work with the same
exact values, 100%, 100%, you know, even
though they had different scale values by
pre comping multiple layers, After facts will think that
they are just rendered, you know, just simple MP
fours, fundamentally, you know, and that's a very
important feature that you may need as you saw in the
previous section, right? So leave versus move all. The only disadvantage of pre comping usually is first of all, you don't need to pre
comp every single thing. So do this only if it's needed. Alright? Second thing, the only disadvantage
is that you lose this check box here, the frame blending,
which is important only if your project
needs time remapping. So we're going to talk
about that later. So just be aware of this. You lose this chat box here
when you precomp something. You can still time remap inside. So if you go to
the actual layer, if it's a video, right, in this case, it
doesn't make sense. So, in fact, this
was not available. But just be aware that
when you precop something, this chat box disappears. Finally, sometimes you
realize that you don't need to have something
precomposed, right? So you need to kind of put
it back outside from here. So one way is to simply double click to enter that pre
comp and cut, right? Cut everything, go outside
and paste it outside. That's one way,
you know, simply. If you wonder if
there is some kind of button to actually un
precompose something, well, you need to
install a free script. I'm going to send you the link. That. It's a free
script that you can install from this website. Let me show you. Okay, here
it's called Unprecmpose. You can find on this website. I'm going to send you the link. And all you need to
do you put $0 here, and you can purchase
this for free and just follow the guide to
install it very easily. And then when you reboot
Afterfax from that moment on, you will have a new
button available. Let me show you here. When you select the
layer, in this case, all, you go to layer and you have this new button here,
un precomposed. So when you press
this, simply confirm, boom, and delete
this weird thing. We don't need it.
And as you can see, it just unpacked the content. You know, it just exploded the precomposition
back to its content. You know, I still do it manually because
it's kind of safer. You know exactly
what is happening. So that's one way to reverse or simply go to a previous
version of the project
13. Kinetic Typography (Basics): Alright, so let's talk
about text for a second. So to make text, you need to right click here. New text layer. And first of all,
type where we want. So for example, hell, oh, how are you? Right. And then, first of all, let me show you a
basic customization for the tax staatics, alright? So where I recommend
that you use always a drop
shadow plus a glow. So let's apply both effects. First of all, here. Okay. And we place the glow
before the shadow here. So the order of the effects
in effet is important. So they go from from
the top to the bottom. So the glow now is applied
before the shadow. Alright? So in this case, we have a black
background, so we can reverse the color of
the shadow to white. Like this and we make it a
little more distant like this and we smooth
it. All right. You know, so we can integrate
this a little better. And then let me show you
the basics of the text. So first of all, you select the text layer,
right, like this. You can change the
content at anytime by double clicking it like this. So you can do, like, test
like this, you know. And to customize the font, the color of the text. You can go to character here. If you don't see this, go to
window and click character. And here you can change
all the basic things. So you can change the
color, as I told you, like this, right, or the font. So just select a
font that you want, like this, very
simple, you know, and the size like this, all the things that you
know about the text fundamentally are here
and the paragraph. So you can kind of
just find the text to the center from
left to right, right? And the anchor point, change this as soon as possible. So you can place
the anchor point to the actual center of the text or to the left according
to what you need. You know, do this
as soon as possible before doing any
animation. Remember this. And in case you want
more cool fonts, you can download them
from this website. I'm just you just see right now on the bottom left
of the screen, right? So I just give you two links
to gather fonts, you know. So you can download those
files install them, and then reboot upper packs. And you will see those new fonts available here in the
character section, alright? So that said, let me
show you how to animate, first of all, the content
of the text dynamically. So instead of using multiple text layers
like this, very boring. Each one will have
a different word, why not changing a
single text layer dynamically according
to what we need, right? So to do that, we need
to expand the properties and go to text source
text, enable keyframe. This keyframe contains
this word here. So if at any moment,
you want to change it, just go here and double click the text again
and just change it. So second. All right. And second T will automatically
make a new keyframe, and this contains that word. So now, it will switch between the two
words. Wow, just like that. And then just repeat this for as many words
as you want, like this. So you can make a
verbatim description like this with very high
costimization, right? And every single word can have a completely different settings. So you can change the size, the text, the color completely. You can change it even for specific letters, you
know, into the word. And one of the coolest
effects that you can apply to text is saber
glow, the hyper glow. So saber is a free, amazing plugin that
you can install it. I'm going to send you
just the link right now. You see it and install that plugin and then
reboot Afterfact. So you will see the plugin here. It's called saber
just like this. And to use it, you need
to simply create a solid. Okay. Let's call it text. Let's call it hyper glow. Okay? The color doesn't matter. Apply the, make sure that
the solid is full screen, at least, like this right. And then we apply the
saber effect here. Okay? And we're going to
talk about this plug in probably later a little more. But for now, let's
use it for the text. So we go to customize
core change it to text. And then link it
to the text layer. So text like this,
Poof. Alright. So now what it's doing is to exactly do the hyper glow
on the text that we want. So let's change the size of
the intensity like this Poof. You know, see what's happening? And then you can just disable the original text if you
want, if you don't need it. And you choose one of the
presets if you don't have time to do your own like
this, you know, like this. And you see what is happening? It's crazy. Change
to where we want. The size is little too big, so change the core, just tweak it. It
depends on the word. Alright, and you see how
many pieces you have. So this plugin is very powerful. It's
crazy that it's free. So just customize tweak the size, you know, it
depends on the text. So at any moment, you go
back to the text here and you change the
actual size like this, and it will just work. Fantastic, you know? And this is a very cool way to make the text look super cooled very
quickly, you know? If you just need basic
customization of the text, but still be efficient, these things are pretty much all you need
to know, you know? And then remember that
the text can be animated standard with everything
that we saw so far, so we can animate the text. Normally, obviously, all
these things can be done. You know, all the things
that we saw can be done on the text completely. So you can use adjustment
layers above it. You can move this with
the overshoot expression to make it bounce
back and forth. But if you need for most people, these are just the things that you need to
know about text.
14. Kinetic Typography (Advanced): One way to start
to use the text in a more advanced way is to animate each character
independently. So to do that, we need to
expand the full properties, and we can see one of
the unique features of text layers, the animators. So this button, I'll also
target one property, first of all, and as you can see there's full choice
in where we want. So I invite you to explore every single one
of these things. But for now, let's
let's go simple. Let's go with a very classic,
a typewriter effect. So we need to target
the opacity, for sure. So we go here Okay. And first of all, it
will tell you, okay, what type of change do you
want to do to that property? Okay, we need to do
0%, first of all. So the text for now will
disappear. All right. Okay. Then we have one thing here. We have the animator
that we just created, which is targeting the
opacity property, okay? And then we have
the range selector. So the range selector
allows you to apply this change progressively
to the text string. So it means that
now we can move, for example, start and poof. We see the text appearing
progressively, right? And as you can guess,
then we need to just animate this 0-100 over time. So one, two, three,
four, five, six, and 100 Woof, right? So now, we can just
type this in, right? And then you can simply just time this
with a change of words. For example, this
goes in like this. Then maybe as soon as
we change the word, we can simply reset. So copy and paste this, right? And now there is
one issue, right? Between this keyframe and this is going to kind of
go backwards, right? So to avoid this, you can right click on the
second keyframe here. So it's going to hold the value, so let's keep the
word completely open. So it kind of types
in the first word, then keeps it open. And then as soon as there's
any change to that value, finally, it will jump
to that new value. So in this case, 0%, which makes sense, and
then type the word out. So it means that
the best keyframes for type out are these. So now we copy and paste. And with this
system, you can just simply type in the new word. For example, double
click again here, third and then copy and paste. Again, this is true, boom. That way for every single
word is going to be perfect. One, and three. Alright. That's it.
And Yara, Yara. If you have a longer string,
for example, multi words, one, two, three, four, five. Okay. You can also target entire words instead of
every single letter. So you can change
these behaviors from inside the animators. Then you go to advanced and then just tweak
it to what we want. You can target, as you
can see entire words. So in this case, when
we tie that out, we'll do this, right? Just the words themselves. Or what we want, even
the entire line. You can completely
customize the behavior. Or you can also randomize the order. This
can be a cool feature. If you turn this on, when
you then animate the start, you see what's happening,
it kind of randomizes it. So let's go back to letters. Okay, so you can get a
cooler feature like this. Okay? So these are just one of the infinite
things you can do. So then obviously, remember that you can
target where we want. So you can go here and create a dedicated animator for every single property
that you want. So let me just do a second
example. We go to rotation. So now we have a second
animator for the rotation, and it's going to rotate
every single character like this, okay? And we change the
range selector, advanced randomize on. Okay? And then we put start to
maybe 50% like this. Okay? So that way, only certain words will be randomized like this, to me, slightly skewed, right? And one trick that I
like usually is to duplicate this second
animator here. Now we have animator
three down here, which has the same thing, right? So we reverse the rotation
to -13 degrees, okay? So that way, it will
actually target more words to the
opposite direction. And we can do this. We expand the range
selector again. We go to advanced and if you
scroll all the way down, there is one cool
feature called sed. The SD is allows you to shuffle the randomization
in a unique way to differentiate it with the
previous animator, right? Otherwise, they have
the same shuffle. So just change this
number randomly. And as you can see, every
time you change it, they will shuffle the
words in a different way. Alright? So remember
to do this to differentiate your
randomizers, right? And, yeah, as you can see, it can become very, very
complex very quickly. But that's the cool thing. Then another very
cool feature of text is that you can convert it
instantly to shape layers. So if you right
click on the text, and you go to create, you can convert
either to a mask, a solid with masks
or shapes layers. The most versatile
solution is shape layers. So if you do create
shapes from text, it will make a shape
layer, obviously. And you see what is happening? Every single word,
every single letter now actually is a
dedicated layer. So let me show you if we expand the contents of this
shape layer here, every single letter now has its dedicated layer,
which is crazy, right? Even subcomponents, because in this case, the
font is very complex. So it means that then
you can do what we want. So one of the things that
you can do if you have the time is to morph
letters into each other. But that can take a long time. But okay, if you have
the time, you can do it. So you go here, you animate the path fundamentally
of each word, and then you kind of cut and paste new
shapes into each other, and then it will kind of morph, like we did in the previous
section with masks. You know, it's kind of a slyly inefficient
way to do this. Uh, but, you know, if
you have the time, yeah, you can do it because
it's kind of fully vectorial. You don't lose quality
resolution, just to show you. But one of the cool things that you can do when you
convert to shape layer is that you have access to the exclusive effects
of shape layers, right? So when you go here
on a shape layer, you have Ed here, and you have these effects
that only shape layers have. One of the coolest
ones is wiggle paths. So you go this guy here, okay? And now at the bottom, we
have this effect here, and let me show you
what it's doing. It allows you to regulate these things if
you increase size, for example, now it's lagging because there
are so many layers. You know what, create
a lighter layer. So you go here to the word. Okay, now it will work. Wiggle pass Poo, here's. Let me show you if we increase size. You see what it's doing? It allows you to explode, text in a cool way, you know, and then you can all
these things can be animated. You see
what is happening? So you can achieve
very cool designs to kind of glitch the text. For example, if glitch
art is your thing, you can achieve very
cool effects like this because then you can animate
this guy over time, right? Like this. So you
start from zero, and then you just
regulate the speed, maybe 15 seconds like this and then go
back to zero, right? So you show, for
example, a word, and then kind of glitches and
then back to zero, right? And obviously, for this, we
can use motion blur, right? So let's go to an
adjustment layer, and we apply a nice
pixel motion blur. Like this. And now it's
detecting the movement, right? To increase the
amount of motion b, you can increase this value
here. So let's double it. Okay, and let's see. Maybe we can increase
also the amplitudes. So let's try detail higher. Okay, let's try the size
a little higher, exactly. And then just tweak
it, you know, tea infrared
combinations like this, see what we can do
like this, like this. Let's randomize everything a
little bit. Okay, let's see. Let's give you some
time to render. Right? Boom. And then return to normal. Alright? So you can do very
cool things with this. And finally, one of the
coolest things about text is that you can use
expressions on it, you know, to display content dynamically
or where you want. So one of the coolest
things you can do is to apply an expression on source text to show a number that's
progressive over time. So we link source text to a slider that we have on the text itself, like
this called source. And now, obviously,
it's going to show the number on the slider. And if you animate
the slider number, obviously is going to increase
the number progressively. So we go 2-20, and
it's going to do this. Obviously, now there's an issue. The number is not rounded. That's why it's showing
all the digits. So assumes there is
a way to round it. Let me show you the expression. It's very simple, actually. Simply, you add this thing
here, value to fixed. And then here with this number, you can choose how many
digits you want to show. So if you increase it,
you can allow one, for example, like this, right? Or more than that. Let's
change the font to a more simple one because
this can be a confusing. Let's go here.
Let's call calibre. Change the font.
Change. Okay, here is. Yeah. You know, and then
let's go back to zero. That way, it will only snap
to integers, you know? Even if we do this. You
see what's happening? Just like that. And you can add a symbol after the
number if you want. You know, if we are
talking about money, you can do it very
simply by just going to the expression
of the source text. And then after the
number source, you just type plus. And then quotes. And
inside the quotes, you type the symbol that you
want to add after the text. Like this. So in this case, we're talking about dollars. So we add the dollar symbol, and now it's going
to natively add the dollar symbol always at the end of the number,
just like this, right? And obviously, this, you
can add more words like this or to them also at the beginning of
the text if you want, so you can do this.
Completely free. So it's going to
print that content. Remember to add the plus, and then you can do, for example, account like this and it's going to print all that
stuff automatically. Wow. You see how cool
it is? Just out. Alright? So, full freedom. One of the things
that I created using these things has been
a clock in the past. So if I go here, let's create a fresh layer. Actually, we can probably do it on this itself.
Let's go here. And let's see should
still have that preset. It's called AM Clock. Let's see. Let's see Let's
see here AM PM clock. Dob click? Alright.
And let's see. We have a single slider here. Were you should go.
That controls the time. Okay. And in theory, when we reach 12, we will go
automatically 2:00 P.M. Right? Let's see. Boom. Okay. And as you can guess, to change these things, dynamically, there is some if conditions
behind this, right? So let me show you in case
you want to study this. Okay? That's the full expression to make something
like this, right? So we have some
at the beginning, some definitions to make the
code more compact, right? And you can define the
hours, the minutes. And then we have some
if conditions to simply assess if we are
on AM or PM here. And at the end, there is simply some printing of everything. So we have the hours, the symbol of the column, and then the minutes,
and then a space, and then simply PM or AM, which is the variable called C, right? Here. That's it.
15. 2D Tracking (Basics): Alright, so let's talk about
tracking for a second. Tracking is the ability to
detect movement in the video, and then you can do many things. For example, you can stick an object to that moving object or you can stabilize the video and many things in
between these two things. Alright, we're going to see all these things case by case. So let's first of all, drag a clip like this. Okay. And right. So as
you can see this video, I'm just moving my
remote control, right. And we're going to track the remote control so we
can stick an object to it. Alright? So that's first case. So there are many ways
to deal with this. One way is using the
internal tracker in Afrofax. So if you don't see
this tab in your case, you need to go to window and
call it from here, Tracker. Alright? Okay, so now we click the footage
and then we go here, tracker and as you can see,
there are four options. For this first case, we
need to use track motion. So click that. We see this thing here. So the tracker is
divided in three areas. We have the target with this little thing
here, this cross. So that goes exactly the
center of the target. We're going to target
this button here. Then we have this bigger square, which is kind of the main
area occupied by the target. So in this case,
just make sure that it's slightly bigger
than that, okay? And finally, the search area. So this analyzes the difference between
the different frames. So it means that
the wider this is, the safer the tracking will be, but also it will take longer. So you need to do a compromise, you need to do a trade off. So according to how much you move, for example, let's see. So you can see, I don't really move it down much even
in the worst case. So I think that this size
of frame should be enough. Let's go a little bigger, maybe slightly bigger, something like that,
should be enough. Okay. And then just
track forward here. And before we do that, we need to choose which properties
you want to track. In this case, we need to
just the bare position, X and Y position. That's it. Before we do that, we
are going to increase the precision of the tracking
by going to options. And then we go to enable
all these chat boxes. Always keep these on. These just increase the precision
of the tracking. And then we increase
this percentage, so that only 10% of
error is allowed, right? And then we just confirm, and we're going to
cross fingers and we track forward to using
this button here. Alright, boom. Okay.
Now it is tracking. So let's just skip
forward until it's done. And that's it. Alright. So
when the tracking is done, we need to apply it on a layer. We need to apply the
data now, right? So you can actually check how we went just by playing the video. Just a little bit. Do jumps. If it doesn't jump in weird
ways while it tracks, it means that it's going
to be pretty much stable. So it seems that it's
pretty much stable. So let's create a layer
that we want you to stick into exactly in the spot. For example, a text. So we
create a new tax layer. Okay, we're going to
call it here, okay? And then the way it
works is that we simply go back to
the tracker data, which is here on the
layer here, layer, and then just select
the tracker here, okay? And we're ready to edit the
target of the data, right? So we click here, Edit
start target and we say, Okay, it's going to go to the
layer called here, right. Okay? And then simply apply
X and Y, obviously, right? So we can also choose
selectively only X and Y, X and only Y. So okay, I'm kind of tempted by using only X in
this case, but you know what? Let's use X X and Y. Let's go. Boom. Alright. And
then let's play. Let's see what is
happening so far. Poof. Alright. Seems reasonably
good. Alright, poof. Then we can start to
enable the motion blur, first of all, you know? And so we can maximize the
aesthetics of this text. You know, let's use
our standard routine. Here, some nice glow. Like this and some shadow. Okay, Ant, we increase distance, and we smooth the
shadow a little bit. So it's more integrated
into the real environment. Let's see, right? So maybe we can reduce the opacity of the
shadow litt bit. We can hide all
these annoying lines by pressing this button. Okay, let's see. Alright. Wow. Alright,
you know, that's it. So that's already first way to integrate something into
a moving object, right? And if you wonder
if there's a way to tweak the actual
position of the text, yes, we can simply by
going to the position. Oh, okay, there's an issue. CNC now is busy, right? The X and Y is busy with
something else, right? So how do we tweak this?
If it's already busy? Well, at this point, you
should already know. Well, we can use a
transform effect to further tweak it, you know, double click on the text
with transform like this, and we have a second
transform now so we can further tweak it
like this, right? Or use simply a parent. So just no object. Okay. And we parent the text to the opposite,
like this, boom. Now we can then tweak
it from this, right? So like this, boom. Just
use the method you prefer. Now we can also calibrate it, so it goes actually at
the center of the object, so that way we can also
scale it more easily. Woof. You know, that's it. So this is the first way
to track moving objects.
16. 2d Tracking (Advanced): Alright, so now let's go
a little more complex. Let's suppose that
we need to stick a text actually onto my face and not into
the remote, right? Okay, so you may
probably already sense the issue that
it's going to happen. So let's as let's do
a step back, okay? And let's do a new tracking. So we go back here, tracker. Track motion. Okay? And we need
to track this time maybe maybe the hole
in my nos trill. Let's try that. Yeah. So you need the high contrast point,
that's the point, right? So or maybe the
tip of the nose or the straighter this because it's pretty high contrast
may be useful. Okay, let's try that way, okay? So you may already guess
what is going to happen now. So we're going to attempt to
track in that precise spot. So let's see what happens
if I try to do this. Okay. You see what happened? It messed up because
obviously the remote control, the remote passed in
front of it, right, and interfered with
the tracking process, which makes this useless when you have a occluding
object fundamentally. The base tracker in
Afrofax works only for the first case when you have
a non occluded object, okay? So in all the other cases, you need to escalate and use an effect called mocha.
So let's do that. So we don't need this. We
actually select the layer. Let's go back here for a second. Okay, we're going to use
one effect called mocha, which is already
pre installed in AfrofaxT guy here, Mocha AE. So let's apply it
on the footage. Make sure that you are
in full resolution. And then simply click the logo, and it will tell you, Okay,
do you want to log in? Just skip everything in
case it tells you that? Just skip still skip. Now we need to go
to the first frame. Let's go just exactly
before the moment. Okay, around here. Okay? So we need to do two masks. One on the face and one on
the remote control, right? So let's click on
this button here. Okay, let's make a
simple rectangular mask around the remote and
don't do it exactly on it. For this purpose, since
this is the occlusion, we need to make it slightly
bigger than the area. So around here. Okay, slightly wider than the area of
the occluder object. Okay? Just like this. To close the mask, click right click. So left click points
places the points, right click, closes the mask. Okay, so then we
call this remote. Okay. Then let's click
on this button again. Let's make a second mask around the face this
time. So let's go here. Again, it goes slightly
wider than needed. You know, around
here, around here. Click, click click, and then right click,
you close. Okay. So that's our second mask, let's name it face. Okay. And now let's tweak the
settings of the two masks. So in both, we need to disable everything
except translation, which means that it's going
to track only X and Y. So click on both, disable
everything except translation. All right. So then remote is
going to do a wide motion. So let's keep it
on large motion, whereas the phase is going
to do a small motion. So click small like this. Okay, then we're still on phase, increase on both masks the
precision to 90%, 90%, 90%. Alright. Okay, that's good.
And now they are ready to go. So you can just track forward. So let's click this button. Okay, before we do
that, be careful. Remember to place
the occluding masks above our target mask, right? So the face is our target. We need to place the
occluding object above it. So place it above like this. Boom. Okay. Down
fundamentally will tell the phase to ignore
the mask over the remote, and that way will not screw up. You know, that's the
point. And finally, before starting the tracking, that will check
this button here, press it S, and that will show where actually we'll
be able to stick stuff on. So, click the phase
mask and place this rectangle around
the actual center of the phase around here. Alright, that's fine. We
can also regulate this. But for the purpose of this, the only thing is just the need to check is that it's centered. That's it. Okay, don't
worry about that. Then click the other mask and double check that
also is centered. Pretty much it is. So we
can keep it that way. And finally, we can
track. So forward. Yes. The speed of this process
depends on your GPU power. When the tracking is complete,
you can double check. Everything is fine. Alright,
so we can search you save Control S, exit from here. Now, we're back in the software. So at this point, we can just create
a fresh text layer, so let's make a new one here. Let's call it mocha.
Okay. And let's stick the style english
at this. Alright. And now we're going to apply
the new tracking to mocha. Alright? Let's keep it actually centered ready on the face. Already here up. Just smooth it. Okay, here. And now
the point is this, we're going to go to
the Toca here, right? We expand the options,
tracking data, create the track data from
the face design, right? So we use the data
of the face, right? So confirm. And then as you can see, we see the points there, and now we just need to
send this data to mocha. So be careful for this
specific purpose, make sure that the position is not separated in the target. So in this case, it is. So when you right click, separate dimensions
get disabled, so it returns to normal. In that way, we go back to Mocha and we can just go
to scroll down here, export, layer export, ch mocha. Make sure that you
enable transform. Alright then just apply. Let's see what is
happening so far. Poof. So now we have the text, stick to my face, even though the remote is
interfering with that. It's crazy, right?
And that's the thing. And also, in case we can even reuse we can also
use the remote data. So let's also do that. Let's
double also this layer. Actually, let's make
another fresh layer. Like this. Again, text. Let's call this remote. Okay, let's move it somewhere
where it's visible, okay? Apply the style.
Okay. So now we go back to the footage here
and so in this case, just click on Create
data again, right? This time, we need the
data of the remode. So click that, confirm. Okay, at this
point, double check the X and Y on the target. Remote is not separated. So in this case, we need to re merge it again
a right click, disable this. Okay? Now it's ready. So go back to Mocha in the footage video, scroll all the way down,
change the target to remote. And finally, just confirm. Alright, so now we can
see what is happening. Poof. That's it. So that's this whole data is used by Mocha in this case, right? And, yeah.
17. Lock-on Effect: Another very cool thing
that you can do with the tracking data
is to stabilize, actually, the true
stabilization, but not in the way you expect. Let's just disable
everything, right? Let's keep only the footage, and we go here back on the
mocha on the footage, right? If you remember, we
have now two masks, one for the remote
and one for the face. So let's make sure that
remote is selected, and we create the
data of that again. And this time we
invert the data. Boom. And finally, we apply it on the footage
itself, so layer six. Okay, and make sure that X
and Y are not separated. So, right click, bring them back together and
finally apply here. Woof. So let me show
you what is happening. Alright? So it's keeping now the remote centered
and on moving. So it's doing the
opposite, right? I kind of reversed the
movement in the real video. So that's so it
kind of stabilized the entire video based
on the remote, right? That's a very powerful effect that you can do
on where we want. So we start to enable
motion blur here. Okay? And to compensate
these borders, you have multiple options. First of all, we can scale up, so we create a known object. Simply and we parent the footage you go
so we can scale up. Let's see. Okay? Let's see how much we need
to scale up more. Okay, more, okay. Alright. And then if you want
to scale up too much, you can also help using p tile on the footage
and then just increase 500 in all directions and change you unfold or slide. Okay. That way, we can kind of sacrifice a
little less of the frame. And then we do a simple
animation on the exposition of the non object and also maybe also the rotation in sync with the movement
that I do on the video, so it will kind of
stay fully centered. So because as you can see, by default is not sell
at the center, so we can move it back
in place. So let's go. Let's go like this. Let's move where we want, you know, here. Okay? And then we wait for
the beginning of the moment. Okay? That's where I
start the moment, right? So let's go. And then we go at the end of
the moment here. Okay? As you can
see, there's also a slide rotation, so
I'll bring it back. Normal like this
and just change. Just bring it back
in place, then play. Okay? Again, here
there is an issue, so we go here, the
next movement starts, so we place these keyframes here and then we start a
new correction here, right, so we bring
it back the center plus rotation back like this. Okay, and then we do a nice
busier to all these things. And that way, you
can keep the video completely centered on the
remote control, you know? So that's a very crazy
effect that you can do. And obviously, it's you can repurpose this
in infinite ways, but that's the core
trick for this. And, you know, in the end,
this is simply just how to stabilize a shaky video, right. But the point is that if you stabilize it on a small object, it will kind of think that
we need to we need to kind of lock the video
into that specific spot, you know, and it
will kind of create this particular effect, you know, that's the point. So you can use this for
technical reasons, right, to eliminate shaky
footage or for creative reasons to achieve this specific effect.
That's the point.
18. 3D Tracking (And More): Alright, so let's talk about
three D tracking, right? So first of all, let's
import nice video. And let's create a
empty timeline here. Okay, as you can see, this video has some issues in the
first place, right? There's some stuff
going on. So we are walking in this environment. We're gonna stick some objects
into this environment, you know, in a three D space. Alright, so Nisha solve
a couple of issues. Before doing the tracking, we optimize the footage. So first of all, let's trim only the part of
the video we need. So, for example, around here. Okay, so then let's start to
um rotate the video, right? So Miu it's a vertical video, so we rotate 90 degrees. Okay, and we change the
resolution of the project. We swap these numbers simply we already know
these things at this point. Alright? So first issue,
we have the fish eye. So the fish eye allows us to see more information but distorts the image around the borders. So let's remove that by using
compensation, it's called. Boom. And then we increase
this number to about 45. This depends on the
degrees of your fish eye. And then we go to
diagonal and revers. Boom. Alright, and then
increase this number until the lines of the video return to be
kind of straight. Let's see. Alright. That
can be a good compromise. This is called aberration,
this artifact here. But it's still pretty good. You know, now we have a
pretty much straight video with straight lines. And
we can start to use it. Now there is one issue. The video is kind
of shaky, right? Because we're just walking. We didn't use any gimbals, any cage to make the video the camera
heavier, more stable. It Spear handheld, the light
the camera was pretty light. So let's correct the shakiness a little bit automatically
using warp stabilizer. So we precomposed the video. Like this shot the fished. Okay. All right. And we apply wor stabilizer. So when you use this, I recommend using
detailed analysis for better resolution,
for better results. Okay? And we reduce
the smoothness to maybe 27% and crop
last 27 as well. Okay? And let's see
what is happening. As you can see, it's very fast. Make sure that you
crop the video to trim only the part of
the video and you analyze. Otherwise it will
take way longer. Alright, let's see what is
happening just for this. Alright, ready, way more stable. But as you can see, there are some weird
things happening, some weird artifacts here, some kind of micro jumps. Okay? So if the artifacts
are too high, you can try to reduce the
smoothness and this number, play with these two or change to a less drastic correction. So let's try to go back to
your perspective. Okay. All right. Let's
see. Now it should be more stable and we
have too many art nets. Okay, maybe 20%. And maybe a little
smoother, 40% here. Okay, let's see. Alright.
Okay. That seems better. So we can work with that, right? Compared to just starting
with this. Let me show you. Right? So it was completely
out of control now, compared to using this, right? Now it's way more stable. Alright? That's very
important thing. Having a stable footage grants you a very
stable tracking, and then, you know, crazy shots. Alright. So now we
recompose this. Let's make our fax thing that we already had
this stable video. So stabilized. Okay. Boom. Okay, so now we
have this footage. Fantastic. So now we're ready
for the actual tracking. So we go here tracking the effect is called
camera tracker. Okay. Let's apply it and enable
detailed analysis as well. Boom. Okay, let's wait
a couple seconds. When the tracking is complete, make sure that you select the
effect camera tracker here, and we'll show you the
tracking points, you know? So we can see those as we
move into the footage. So right now we are able
to stick wherever we want on any of these points in the video. You know,
that's the cool thing. So we can make these
points a little bigger. Sometimes if they are
small, you can tweak those. And now, for example, let's as you stick a text on the ground of this
footage, right? So we go around here and we select just with
the left click. These points, you know,
that we are interested on. And as you can see, we'll
create automatically a polygon in that area and then right click
at the center of that. Okay? And then create
solid and camera. Boom. Okay. So the solid
right now is gonna have a placeholder to just check how stable the object
will be in that spot. So let's see what is
happening already. Woof, let me show you. So we're moving. We're moving. And that shape is exactly
there on the ground. You see how crazy it is. So now we're able to stick where
we want exactly there. Alright, so let's
create a text layer. First of all, new text. Okay, let's call this happy. Okay, let's maximize the
aesthetics of the text. First of all, let's go
here, default glow. We already have it. Okay, and
maybe let's add the shadow. Okay. So first of all, we make the text, three D, so three D on on the text. And then to make it teleport already automatically
where the solid is, we use the truck I showed
you earlier with parenting. So shift parent to the solid. Boom, and it will
go exactly there. At this point, we can
delete the solid. And then we just need to
tweak the size of the text. So just bigger, like this. Woof let me show you
what is happening. Whoo. It's already there. You see what's
happening? Just that. So this is a very
powerful technique for branding or music videos or movies where we
want, you know? So, Okay, now at this
point, we go there, we act oriented the
text where we want, in case the slide you off, we can move with the
rings like this. You can do wherever
you want, like this, you know, you choose the style that you prefer,
you know, like this. You can choose the font at
anytime, completely free. And thanks to the shadow, we can integrate it, integrate it even better into
this environment. So, for example, the shadow
distance a little bit. Okay, maybe a little
bigger, you know, tweak completely free, you know, like those maybe
even bigger. Okay. The Kristi thing
that stays there is completely stable.
That's the point. No jumps. Boom.
Alright. The text can also be animated over time. So if you enable a automation on the Y
position, actually, let's enable on all
three dimensions, X Y Z, we can do this where
we go slightly earlier. Okay, let's go. For example, let's go around here, okay? Okay, we go slightly
earlier around here. And then with this arrow here, you move it up in the
video like this, right? So you can literally make
it fall down in the video, you know, like
this. Boom, right? And then you can enable
our overshoot expression. So we go here, overshoot
on all three dimensions. In this way, it will kind
of fall down in place. So let's see what is happening. We go there. Boom, right? So obviously, let's make
it become a little faster. M, right? And then you can
add some kind of impact to the video so that it kind
of it's fully integrated. So it kind of makes sense to make a slight vertical
shake on the video. So you can also add a tiny detail exactly
when the text lands, you can go to the
position of the video, the wide position in
this case, right? And then you enable overshoot
as well on that property. And then you go one
frame after, like this, and you just place change the vertical 100 pixels,
something like that. And then you go back
to the original value. And that way will kind
of bounce back in place. Let's expand the video to avoid the black bars. So we go here. We just reptile like this. Okay? So just use a
classic reptile like this. So now it's expanded.
Alright, in that way, we'll be
more integrated. Let me show. Ff, right? Boom. You see how much
integrated now is. Boom. Alright? So
that's the trick. Okay, so let's go back to the main quest the
three D tracking. So we go here. And
that's how you integrate elements into the environment already with the three
D tracking you can see, is very very intuitive
at that point. You know, you simply track
the footage after you optimized it for stable
tracking, and that's it. You know, the double frame rate is one of the most
important things. You know, when you record
your footage in 50 frame, 60 frame per second, it will
grant you more smoothness, you know, and all these things. Okay.
19. 3d Tracking (Mocha Method): A Let's go now one
level above this. Let's suppose that we
want you stick actually a text actually here
on this first column. So let's go at the
beginning of this footage. Let's suppose we want you to
stick a text actually here. By the way, I know that you put this probably even higher. Let's go this even
higher here. Okay. Oh, okay. Actually, the
issue of this text is that we need to simply
disable it before that point, it makes sense to trim
the text element, you know, from here. That way, it will not exist, and that's fine. Boom. Alright? Okay. So let's suppose now we
want to stick a text here, exactly this rectangle, right? So, now we're going to
use a second method for three D tracking. A we going to use
Mocha? So let's go at the beginning of
the footage exactly here. Be here, okay? Okay.
And let's use Mocha. And then we click
the logo, obviously. And we're going to
now, first of all, place a mask exactly on
this rectangle here. So one of the additional
features of Mocha is that we can fundamentally do a
localized three D tracking. So we just close a mask
here on the target. Okay? And then we press the S to calibrate exactly
where the text will be. So we tweak now these edges
exactly in the desired spot. So this time it's very important
to calibrate this fully, not anymore just the center, but all four edges of this. Okay? And then we enable a
perspective down here, 90% precision is good, and then just zoom out and
start to track forward. And usually this
works fantastic. So let's see seems that
it's pretty stable. Okay, obviously, when the mask starts to go too
much out of frame, it may stop because we placed
a too high percentage. You know, there are
not enough pixels. So let's go with 50% when
we arrive at this point, and you should be able
to at least push it out. Okay, still too high.
Okay, let's try with 30%. Okay, let's see. Cross fingers, and, okay, one more.
One more pixels. Alright, let's go 28%. We just need to make it
go out. That's the thing. Okay, even at this point it's literally the last
frame, so it's fine. So let's try 10%. Okay, perfect.
Okay, that's done. Okay, so at this point, the text is completely
out complete, so we can save and
then go out of here. Okay. So now let's create
a second text layer. So we go here, new text. Okay, and we apply the same
sling of the previous one. So copy and paste, wait on these things.
Okay, here is. So now we do one thing. We precompose the
text this time. So this one here
gets precomposed. So we call this love, okay. Alright, so now we go inside
Love and you make sure that the text is fully stretched to the full frame, so
it will look weird. Transform, Fit comp. Boom. So, it must
look like this. Fully stretched to the
full frame, alright? Then you go out of here, and you go to the video
here, the footage. You go to the mocha
settings, tracking data, create the just export
the data from the mocha, you know, that's the
single math that we did. Okay. And then simply port the corner pin with
mochonleu on the love word. So we go here,
layer por to love. And then make sure in this case, we have X and Y separated. So always make sure that they are not separated
for this to work. All right, so they should
be like this, normal. And then finally,
mocha export, love. Ooh. You see what happened? The text teleported there, and let me show you. It's pretty stable.
It's there. Proof. Just do that, you
know? And, meanwhile, the second word falls
down, it's crazy, right? So that's the
beauty of all this. And obviously, let's start
you maximize the aesthetics. So we start motion blur
everywhere on the words. Okay? And then you
integrate this even better. You can keep it like this
if that's your style, or you can do, for example, some
blending mode here. So you change the
blending mode to overlay, for example, so it will kind of interact with the colors
of the footage, like this. So this depends
entirely on your style, what you want, what
you need, like this. So that's completely
your choice. Anytime, remember that
you can just go inside love here and change the
word to where we want. If you change the
word, make sure that you re stretch
the full frame. So right click Transform,
fit to come again. So make sure that the word stays inside the outlines
of this frame. Otherwise, it will be cropped. So that way, then it
will automatically be fixed at this level. Alright? Remember just that. And this can be
actually where we want. Wherever you place here,
even an image also works. So, this case is for text, but even if you place a logo, nemage, a video, where we want. So we have a generalized
three D tracking with automatic effect here, camera tracker, and a localized three D
tracking, thank you Mocha. Okay? And the truth, the final trick is should then just combine these two things. You know, that's the point.
20. Speedramping: All right, so let's
talk about time remapping for a second, okay? So here I just shot two videos where I kind of rotate
around two objects. And so let me show you
just the source videos. First of all, let me go
here, here and here. So I simply orbit around
object one, you know, and then I go I repeated the same type of shot
with a second object. Try to keep the same movement, you know, same
position like this. Okay. And I'm going to skip some parts
because at this point, if you reach this
point in the course, you already know what it did. So I first of all, stabilized the video
with Warp stabilizer, you know, with these
settings here. And then I just used
Mocha to stabilize the video at the
center of the object. So I used the lock on effect
that I showed you earlier to try to keep the video stable around only
that object, you know. So these are the two
processes that I did already on these two videos, where you know these
things at this point, and we can skip to the
time remapping part. So these videos are
now just ready to go, are stabilized, but they're still not time remapped, okay? So the trick for achieving very clean time
remapping is also the music. So you need to have
some kind of audio that guide the flow of
the entire video, right? So, in this case, we have
a simple music beat. Let me just show you just
listened for a second. We have a drop. All right. So we're going to synchronize the two clips with the
first bit of this drop. Alright? So let's first of all, enable Ttermimpping. Let
me show you how it works. So you select the clips where you need to
use this and you press Control Alt Or
if it doesn't work, just right click Time,
enable Timmy mapping. Okay, so when you
enable time remapping, it will automatically
place two keyframes, one at the beginning and one at the very end of that clip. Be careful because if the video is actually longer
than it should be, it may place the keyframe somewhere else,
for example, here. So the true end of this video, for example, is not here.
It's actually earlier. So let's manually
place a keyframe on the very last visible
frame of the cube video. So let's go there.
Let's see what it is. So it should be actually here. Alright? So we place
one keyframe here, and we lead this here. So we should have two keyframes. One at the beginning, one
at the end of the video. Let's double check that the
JBL video is also right, so the beginning is
right. Let's see the end. Yeah. Okay. Okay. Make sure that actually it's not
a blank frame as well. So double check, go
one frame earlier and place one keyframe,
delete this one. So we should have always two
keyframes only one and two. Alright, so now the time
remapping is ready to go. So the way it works
is that by default, when you place doesn't matter where you place the
second keyframe. It kind of gives
you 100% certainty that the video will end
by that point in time. That's how it works. It's not a percentage, it's not anything. It just tells you that
the video will end there. So it means that if we place
this on 1 second here, the video will become
extremely faster, right? Let me show you. You see that just doing the movement extremely
faster now, right? Because it's been compressed
in that amount of time. So let's actually time this. So let's go with the video
with the cube first. So we place the
cube in solo, okay? And the attack is easier to do because
it it you simply move the beginning of the video and sing with the beat of the song. So now I'm going
to scrub through the music using
Control and click. In that way, as
you can see, I can listen a little bit
the audio as I move. See what's happening?
And if I stay still, while holding control, I can also if I keep the
mouse pressed, it will play repeatedly that
sample where I am amped. You see? So that way I can very precisely go where there's
the trends in the boom, the thump of the kick, right? Okay, here's. Alright, I just
found the attack. So that's where the cube
video needs to start. So we just place there.
Alright. That's the first part. Then the end of this
mini clip needs to be. So well, okay, let's first see how many movement we do.
We do two movements. We go here to the right, and then we return
back. Alright? So, it means that we can
well, actually, you know, to make it even
simpler, we can simply use only the first
half of the movement. So we go here in the cube. Okay? And we go exactly
where the movement ends. So exactly here. Okay. And we place one keyframe here. Okay? So we don't need
the rest of the video. We just return back. We'll
just reverse it from here. So we delete this keyframe here. We only have two keyframes now. Okay, so this is where
the first movement ends. So it means that here now we to choose in the music where to synchronize
it. So for example. Let's go with the snare.
So the snare is so we have so we're going
to alternate. So let's see. Okay, the snare
is there, so let's see. Here is. So I just
found this snare. Okay. So here we place
this keyframe here. Okay. And then we
need to repeat. We need to return back at the beginning of
this video, right? So we copy the first
keyframe simply, and we paste it somewhere
here and we're going to paste it exactly in sync with
a kick of the next beat. So let's see where
it is. Here, right? You hear it. And then we means that we need to
place this keyframe here. So now, fundamentally, what
is happening is that we go from point A to
point B in the video, and then we return
to point A, right? So we have one, two,
and then three. Okay? So it kind of
completes the cycle. And the clip also ends here, so we also chop it
here. Alright? Okay. That's the first part
of time remapping. Let's just do the
same thing that we did with cube with the JBL. So at this point of the video, we also switch to the JBL. So we start to
enable that, okay? And we do the same
thing. So let's place one keyframe at
the end of the movement. So it's exactly here,
okay, one here. And then we will
just return back. So we need all two
keyframes for now, exactly. And let's
do the same thing. So we find we have the
cook here already here. Let's find the snare.
Okay. And we go. We place that keyframe here. Copy the first
keyframe so we can go back to complete the cycle. Let's find a kick. Here is. Okay, paste. So as you can see, music is extremely
important for this. Okay, so now they are
both ready to go, but they are still not
interplated correctly. So let me show you
what we have so far. Let me loop this part here. Okay. Also, I think we can go double speed. I
think it's still slow. All right. Okay, so we can
actually go double speed. So let's go like this. I mean, the clipo sheet can
potentially enter even here. So to do that, we can simply select the key frames
that we did so far. And while holding Alt or
if you use Mac option, you can select the last
keyframe of the series and you can compress
everything proportionally. Alright? So we can just make the clip end here
at this point on the second one instead of
the third on the snare. So we can simply select
these and compress those and then hold shift so
they will snap also here. Okay, then we make the
JBL do the same thing. So it starts here, and then we go to the snare. Okay. Here is okay. Okay. And then we
compress, boom. Okay, now let's see
the timing. Okay. So at this point, we
can start to enable Bezier on the time like this. And then we go to
the graph editor. And the curve that
we need to use is entering fast
into the footage. So like this, we enter fast
into these videos like this. And then we do exponential curve to go out
of the video, like this. So this is the
curve that usually people use, something like this. Okay? And let's see
what we have so far. Ooh. You see what is happening? Now we start to talk,
right? So, and then let's start you add some
additional effects. So a mushroom blur is placed
on top of everything, plus some kind of
color correction. You know, we're gonna
deal with this later. I'm going to show
you what it is. And for now, let's just see
what is happening so far. Let's go here. So you see
how perfect loop stat to be. So we have the hu clips alternating perfectly on time because they have a
very similar moment. They are both stabilized
and they are both time remapped to have a special
curve that I showed you here. So we enter fast into the video. We kind of slow down at the
end of the B movement here, and then we return
at the beginning of the video with
exponential curve. That's the fundamental
principle for a smooth time
remapping, you know. And yeah, so And then just repeat it
for every single clip. You know, that's the work. You know, you just synchronize each to the beats of the music. That's the very important part. Make sure that
they're not too slow, no too fast. That
depends on the music. And then add some
post production. You know, that's the
thing that really made it very smooth, right? So the motion br really
does a lot in this. So then, let me just simulate
more clips like this, double double double, like this. Okay. In theory, it should
be still symmetrical, so you can just move this here. Exactly. Yeah. So let's just look these four
bits repeatedly, you'll see. Give
some time to render. Maybe reduce the quality
of the motiombr right. Use tend AP, you
know, four K is so efficient. Perfectly on time. One, two, three, four. One, two, three, four. Alright? So that's a
secret for time remapping. And in the end, it's
literally just this. You know, you simply control where each frame of the
video will land on. And the music is
extremely important. You know they get confused
with this, right? And the only thing that
needs to make you aware of is that when you
precompose a layer, you lose the ability to use
the so called frame blending, which normally is
available on the videos. So let me show you
if we go back to the source video, Okay? As you can see, we
have this switch available, this frame blending. So this allows create additional frames where
normally there wouldn't be any. This is very important
in case you do heavy stretching with
the time remapping. You know, if you do
it like this, if you have a couple of frames here and you stretch
the video a lot. Now obviously, we have
less temporal resolution between these two points
of the video, right? And that's why you may need
to turn on frame blending. So frame blending creates some kind of echo effect. You
see what is happening here? You kind of stretch smooth the the slow down
that you did, right? So, and this is
available only if you import MP a video,
fundamentally. So if you precompose it, you
lose the ability to do this. So, and that's why shooting a double frame rate
helps you because you already have so many frames that it kind of doesn't
matter to turn design. You may not need it
because you have already 60 or crazy cases, even 120 frames per second. You can do it. You just set up the composition to up all
those frames and it will work. You know, I will give
you those frames. Or even if you don't
change this to 120, and you just import
the video that has 120 frames internally, that information is
preserved when you stretch it violently
like this, you know? And that allows you should
not use too much CPU. That's the point because this
makes everything heavier. In the cases where you need
this, okay, that's fine. Remember that there are
two modes, echo mode, which is the one
that we just enabled here that creates a echo, as you can see, or morph mode. So if you really click this, it will enable a more
subtle interpolation. As you can see, even not
noticeable right now. You see that it's really
almost impossible to notice. So it's kind of some
morphing between the frames. Very clean, you know, I like
to do very high movement. And then to disable it just
repress this button. Okay? So one check that I did
sometimes was to, um, export the precomposition as an MP four before doing
the time be mapping. That way, I can use again, frame blending because it
becomes available again. If you import this
as an actual P four, you know, that's the point.
But, you know, it's a choice. So in this case, it was
not necessary because I already recorded at
a high frame rate. So that was that
was fine, you know?
21. Keying (Rotoscope + Ai): King is the art of cutting out a specific
part of a footage in order to put
something behind it or erase the object
from the scene. So let's suppose that I recorded a video of
myself moving like this, left and right, okay? And I realized
after the recording that I forgot to place an object here,
this is exact spot. But the issue is that I can
if I put something here, then eventually I pass
in front of it, right? So how do I recreate
the illusion that the object is
still behind me? Well, we need to use rotoscoping
or something similar. So let's first of all, trim
the video in just that part. So we don't need
to work too much. Let's go around here. Yeah, let's go from here to
here, and then back away. Alright. Okay, that's enough. Let's snap it here. Okay, that's the main video. Then I recorded without
moving the camera, I just recorded a second video where I just put the
object where I wanted, you know, there, right? So and we're going to call cube. So that is kind of
a small detail. You know, the cube needed
to be there, right? So first of all, we make a simple mask
on the cube video, so we make simple
mask like this. And also, this doesn't
need to be a video. We can just freeze
the frame, right? Make a simple frozen frame of this second video.
Let me show you, right? So right click on the cube
video, time and freeze. Alright, so now is
like a simple image, you know, the issue now is that, obviously, when I
pass in front of it, it will still be above me, whereas it should be
behind me, right? So let's fix that. First of all, let's just increase
the feathering of the mask a little bit
so a little smoother. A right. Okay, so let's
proceed with the Ro scoping. So the way it works is that
you click on this tool here. We can also press Alt W, and double click on the footage. So before you do that, make
a copy of the footage. So just double it and
place it above everything. And we can call this Roo. Okay. So double click on this
guy here, above everything. And now you will
see the brush here. So the way it works is
that you tweak the size. So let's make a medium size, in this case, like this. Okay. And you draw inside the
subject to isolate. So boom, and it will automatically detect the
points of a higher contrast. So just correct it, you know,
just tweak a little bit. You can do some clicks
or drag like this. If it gets points
that you don't need, hold Alt and then bring it
back in place like this. Boom, you know, the
more you correct it, the more precise it becomes
because it's kind of a micro AI that learns
what you need, you know? So we need this, we need this, alright, that seems decent. Don't worry about the very
fine tune details here, because actually the hair can be detected by the
second tool here. So when you have a main
mask that is decent, you go here and you
click again with Alt this is called roto edge. So you just go around the
perimeter around the hair, and it will also
feather fundamentally the edges for even better
dettion like this. Okay. And now let's double check what is
happening in the main comp. Alright, so far seems fine, but let me isolate
the rotoscoping. You see, now we start to
have myself isolated. Alright? Okay. So at this point, we just need to track
the rest of the footage. So good news, it will happen automatically
for the most part. So we go forward 1
second at a time, as you can see, tracking
the frames. Okay. And let's go forward
all the way to the end. Less renders the full
thing full quality. Remember to remain full quality
when you use rotoscoping. There is a way to
roll lower quality, but when you are still
creating the mask, you need to be the
highest quality possible. Okay? Let's go to
the first frame let's see what is
happening so far. Alright. Seems decent. You know, that's it. Okay? At this point, let's smooth the cube a little
more. Let's go. Okay. Alright. And, okay, even like this, already,
the issue is solved. As you can see now
I create a delusion that the cube was already there in the
first place, right? And okay, then at this point, we need to consolidate
the rotoscoping, because if I do any
change right now, it will recalculate this area here, which is very
annoying, right? So to consolidate this work, we can simply go full screen, double click here on layer. You see a hidden
button called freeze. So when you press
freeze, it will just save, does work, you know? And that way, we
can also finally go back to lower quality, you know, core, half,
and that's fine. Now it's completely frozen, and the video should
be fine. Let's see. Exactly, there are
no weird jumps. That's good. And we can
finally go lower quality. And what recommends you enable motion blur in the roto scoping, because in the end,
it's just an effect. As you can see, have
some features here. Sometimes you need to
increase feathering, contrast shift edge and
same for the refined edge. So the refined edge is the second thing that we
did for the hair, which has fundamentally
the same settings as the main pass. You know, increase the slightly sometimes when you have
additional issues, but in this case,
it's not needed. But the most important
one is the motion blur. If you enable this, it
will kind of smooth the edges exactly here. You know, you kind of
integrate it even better. Okay, now the point
is that we have fundamentally a perfect
rotoscoping already, and the cube is exactly
where it needs to be. It's behind me. Alright? Who. So that's the first solution. But as you can see, it's a little painful because
it uses all CPU. It's a little slow if you don't have enough power your CPU. So I'm going to show you
now a second method that, in theory is even
better than this. Alright? Very quick,
easy to learn. So let's just go
back to our footage. Okay, here, and we're going
to export it as an MP four. So simply press Control M, and you select just a
random preset here. Select 15 megabytes/second here and then just save it somewhere. Okay, so you just call it
pre AI. Okay, and safe. Alright, so now I'm
going to show you a tool called Matt Anyone. So this is a free open
tour tool that you can install locally if you
have a Invidia GPU. If you don't have it,
I'm going to leave you a link to use it online for free or
with a membership, to have a quicker
processing time, right? So, in my case, I have
installed locally, so I'm going to just
jump on the tool here. And this is how it looks like. Drag the video here, first of all, so we
go here, pre AI. Okay. And then I click
on Send to masking. Okay. And then I need
to do two clicks, one on my head and
one on my body. Alright, in this case,
I got very lucky, just detected the whole body already with the first click. Maybe I need you to detect
this part of the headphones. So let's do some additional
clicks here. Let's see. Alright, so in case
it's problematic, you can clear the clicks and redo a new attempt from scratch. Okay. That's it. You know, one click here.
That's it. You know. That way, when I
process the video, it will still be
completely clean. Let me show you right now, when I process it, boom, boom. Okay. So what do we do
with this weird video? Well, we download it
first Alpha output. Okay? We call this post AI, and we import that video
back into Afterefat. So we go here, import, post AI and we put it back above everything here
in the project, right? In sync with our video. So in this case, just the
beginning of the timeline. All right, so at this point, we just need to
select a new copy. Let's make a new copy
of the main video. Let's call it Roo two, okay? And we're going to keep the
cube enabled. All right. So we simply now
use the Track Matte from Roo tu to post
AI, like this. Boom. And then click
this button here. So normally, as you probably
saw already earlier, Track Matte allows you
to tell a layer to exist only where the
Alpha information of another layer is, right? Which is not what we want. In this case, this
black and white thing is the black in this video
is not true transparency, as you can see, is
actually opaque, it's actually rendered, right? So in this case, to detect this, we need to use the Luma
mode of the track Matte, which, you know, the words
is simply this button here. When you click this button, it will use the
Luma information. So that's fine. So in this case, we can finally disable this mat. We don't need to see it anymore. And right now, the roto che, in theory, it should
be already isolated. Listen. Poof. You see
what is happening? So it's using the
white and the black of that video chee due
the difference, and it's kind of isolating what is not white as a
transparent thing, you know? So, in other words, now we can just put Roto
che above everything here, and we reenable the cube there. The main is below everything, and in theory, now
it should work. Poof. That's it, you know? And remember that
once you have a rotoscoping setting ready to go, you can have fun with this. So I can now put whatever I want behind me so I can make
it a nice taxi boom, you know, and that's
it, you know, so I have my video, and I can
put information behind me. So that's a very classic thumbil
creation routine, right? So, so, yeah, you usually don't you will
not need more than this. You know, you can
already do whatever you want just with
these two tools, rotoscoping, and not
anyone, alright?
22. Keying (Green Screen): Alright, so before
we change topic, let me just show you
a couple more methods to king stuff, right? We cannot get out of
here without talking about a green screen,
for example, right? So, obviously, if you have a green screen or blue
screen behind you, we already know that there
is an effect that allows you click on the background
and remove it, right? So, but that's just not the
single way to use the effect. You can actually use it on whatever point of the frame has a very high contrast point, alright? Let me
show what I mean. So let's get this example here. Let's make a project. Alright, so for example, we
have this footage here, okay. So as you can see this
piece of paper behind here have a very high contrast point compared to everything
else, right? So you can simply use
this effect here, which normally is used for the green screen
here, key lite, okay? So we drag it inch
in the footage, and then we simply
click this button. You choose the color
that we need to remove. So normally, if you have
a green or blue screen, you just click here somewhere and it will sample that
color and remove it, right? So let me show
you what happens. If I click in the background
here, it will try to remove. Obviously, it's not perfect because it's not
the green screen, but this will work better because it's a
very high contrast point. So let's try that. Let's try one of
these points. Wow. Okay, okay, way better. As you can see, removed
all these sheets here, but also remove more
stuff that we don't need. So there is a way to make this effect only work in a specific area of
the frame, right? So as you can guess, we
simply make a mask here, so simply a rectangle mask
around this area. Okay. And then we double
click on the effect and scroll all the way down here until you see
compositing options. This has a plus
and minus button. Click the plus.
And this will now tell this effect to
work only in this area. Oof. You see how cold it is. So you can link effects
to specific masks. Thanks to this feature, right? So now, obviously, the
rest of the frame is safe, you know, no issues. And we can just work on removing
those sheets, you know? So let me show you
how kilte works. First of all, sampling the color this way
is not the only way. If you hold control,
as you can see, the tool becomes
slightly bigger, and fundamentally, it
will sample a wider area. So sometimes doing this trick will sample the color
in even better way. Remember that? Okay, that
said for now is fine. Normally, after you do this, you don't need to
regulate too many things, but if needed, you can do this. First of all, try to regulate the screen gain versus balance. So increase this in case, let me show you what happens. You know, we'll kind
of excavate more. And balance does pretty
much the same thing. You know, if you
increase it, we'll kind of excavate a little more, a little less, if needed. So this kind of a first
quick fine tuning of the green screen
removal, alright? Then if more
correction is needed, go here on screen mat instead
you tweak these settings, for example, let's search you make the borders
a little more clear. So for example, let's search
you play with Clip black. Okay, that's too much.
Let's try with clip white. Alright, you see
what's happening. So if I decrease Clip White, it will make the borders clear. Alright, let's kip it this way. And then just play
with clip rollback. So just increase these
values slowly to see to kind of fully
understand what these do screen shrink growth. So let me show you
if I increase this, obviously, it will
make the mask. Shrink or grow, right? So let's make it smaller like
this and more contained. And finally, softness is kind of a feather, you
know, for the borders. Alright. And so you don't need to touch
anything else, right? So these are the things
that you need to touch in most cases, just to go in a little
more advanced removal. And this is valid whether
you're removing a green screen, a localized green screen like this case or a green screen man. You know, you know, one
of those people that wear those weird suits. That's also valid. You know, that's also
one trick for this. So in most cases, it will solve all your needs already
with screen mat plus this and plus having a well lit
environment. Alright? Now, obviously, we
can actually put something behind that point. So we can put, for example, a nice texter from now, just random slid is fine. Yeah, noise. Let's generate a random texture here,
turbulent noise. And let's make it alive
automatically where you know, these things, time
multiply by 400. Okay, and then just make a random thing contrast
right, just like this. Okay, and then put it
obviously behind the layer. So just here. Boom. You see what's happening? Okay. And then just tweak
the contrast of everything. So for example, let's go
back here to the mask. We forgot to add some
feathering on the mask. So just a little
higher, like this. You see how it starts to blend start you blend it better
with the original, right? And then it's just
a merry tweaking, again, these numbers
like this, right? So you can choose how much
this happens, you know? It's really all here. Just balance these numbers until you find your sweet spot. So and that's one of
the ways to key out stuff using effects instead
of rotoscoping, you know. Okay, so let's go forward
into this footage. If the high contrast color
is a white or black, you can take advantage of
one of these other effects, extracts very very good. So let's go here. Let's apply one extract after the kelte. The keylte is working
only here, that's done. So for example, if you have a mostly a white color to
remove or a black color, you can use extract. So this is fundamentally the full spectrum of luminance
right now into this frame. So if we reduce, if we increase black point
to see what is happening, I removing everything
starting from the dark colors until it leaves only the
whites, obviously, right? Boom, right? And just find
a spot where your color, your desired area remains
visible, you know. So and then we can
smooth it a little bit. And that's a quick way
to have, for example, a instant rolescoping
of this card. So we can put now stuff behind this card
specifically, right? So, so and then, as you can see, is this completely working in the video like this,
you know, that's it. And then second case, very similar to extract, there is also a Luma key. That's also a very good one. Fundamentally does
the same thing. So if you increase
threshold key out darker, it does exactly the same thing. You know, some people use this same exact thing, you know, very similar and increase
feather to smooth the edges, just like that, you
know? That's it. And the opposite is true in case you have a
black color to remove. So if you go here, also, it
needs to do the deposit. So let's go back
to your extract. Let's bring it back
to square one, okay? So if we remove the colors from the other
side from the bright side, we will end up having only
blacks at a certain point. Let me show you slow, slow slow until we have
only the darks, right? So, in this case, it's more complicated
because you can see, there are already more
colors in the frame that already share this specific
tonality of black. So it's not truly absolute
23. Face Tracking: Alright, so let's talk for
a second about tracking, using the masks in Afrofat. So so far, we tracked the masks using the
mocha tool, right? But the truth is that
you can also track movement using the masks
in Afrofax already. The reason why we didn't
tackle those so far is that after the
introduction of mocha, they kind of fell kind
of in the back burner. You know, they kind
of got used a little less as tracking tools, and just for static
masks more, you know? But for completion, let me just show you also what you can
do normally with the masks. There's a couple of features
that are very cool. Alright, so let's go
with some footage here, okay? And so let's go here. As you can see in
this video, some face expressions because we're going to use exactly one feature that
works like that. So let's go exactly
here give me a second. Okay. Alright. For example, we go here here
here, and Alright. Okay, let's just trim
a little part here. So first of all, let me
just show you very quickly. Normally, in theory,
you can do this. You can use simple
masks like this. You know, let me go here. You can make a rectangle. Los like layer like this, right? Okay, so we have a layer here. And if you right click a mask, it will tell you track mask. And let me show you what happens when we pull
up the tracker. Whew, we have the tracking
options for that mask, you know, and very
similar to Mocha. You know, fundamentally, the higher the level of position, the more precise
the mask will be. So, for example, we want you
to track all the position. Okay, you just track forward, and it will literally create a mask that tracks the
face in this case, right? So, but the thing is now Mocha became a default
plugin available, so it's always better to do it there fundamentally,
that's the point. But there's still one
exclusive feature that we can use in the masks, which is the phase
tracking, right? So instead of making
a square like this, let's make some kind of circle. So we go here, old click. Yeah, let's make some kind
of circle around the face. Alright, like this. Okay. Then we right click
on the mask track. And here, we're going
to go all the way to the lower options. Here the last one phase
tracking detail features. Alright, so you select this guy, and then we're going
to set the rest post. So this is pretty much
the rest post because I don't have any relevant
facial expression so far. So we just set that. And then we just track forward. Make sure that you kind
of trim the layer, and also you constrain
the work area, so that way will
automatically stop the tracking when
it arrives there. So now it's fully trimmed. Let's go tracking. And Whoo.
You see what is happening? Now it's detecting all the facial expressions
as I go, you know? And what happens now is that, first of all, we have a
perft mask around the face. So, in some cases,
maybe for funny scenes, you can do this
where you can have only the face floating
around, right? And then you can move
it like this, right? So that's really one of
Maybe for tutorials, you know, funny tutorial can be like this, you know,
just with the face there. Right. But another cool thing is that every single
point that we see here on phase track points now is a coordinate that we can use for
something else, right? So let me just show you. For example, let's first of all, slice this pace a
little bit, right? So let's create an
adjustment layer here. Okay. Let's call this Distort. Let's make this face a
little more cartoony. Let's go to cartoon. Okay, I'm looking improvising
right now though, let's see. Just let me randomize
this a little bit. Try only edges. Okay. Let's try it with mosaic. So we kind of pixelify this first of all,
that will click. Okay, and then we link vertical
blocks with horizontal, so we can edit both
numbers at the same time. So let's try 50. Okay,
maybe a little higher. Alright, then we use posterize. To kind of decrease the
bit depth of the colors. You know, it's kind of a
jiff effect like this. Alright, then let's
use vector blur. Okay? And let's see how we can squeeze the shape a
little bit. Alright, of. Okay. Alright. And then finally, let's try cartoon at the
end of the chain here. Okay? Maybe that's too much. Let's try fill. Alright, and then we just
put a glow where you knows. Okay, so some kind of quick
processing, you know, to make this phase a little more like a character, you
know, of a story. And then finally, we duplicate
mosaic and pasteurize, duplicate, and put at the end of the chain,
more one more pass. Maybe we increase the resolution
of the second mosaic. Let's see. Alright. Okay.
Maybe you disable this. Okay, so we start to have
a more sliced phase, okay? So let's suppose that we
want to stick something exactly on one of the pieces
of the face. We can do it. So we right click, create a solid, for example. Let's suppose we want
to make some statistic. So we make a line that starts from one of the
points of the face. So for example, line one.
And now, you know what? We squeeze this layer
like this. Alright. And we change the anchor point. First of all, we change it to the right like this
because I want you, for example, stick it to this point here around this
point of my mouth, right? So we go, we click the
layer with the face, and as you can see this effect,
and now it's available. Face track points.
When you click it, it shows all the points. So the way I do this is to just touch one of
the point that we need. So in this case, this point. So when you touch it, it will
automatically highlight it. So let me show the keyframes. So every single these points
is busy with keyframes. When you touch one of these, it will automatically highlight
you see what's happening? It says, Okay, here
I am. All right. So now we need to just parent the position of the solid
to these coordinates. Mouth, right. Alright. So now that we know this, we simply go to the
position of the line, make sure that it's
not separated, and then link straight to mouth right at this
case. Let's go here. Go go go. Okay, mouth right. Let me show you
what is happening. Now we can just design finish designing the line so we
make it a little thinner, first of all, Okay, and shorter. So we go like this, just with a scale
brutally, like this. And that's it, you know,
now we have a line that says exactly on my mouth the
whole time there, you know? And then we can just take
a text with instructions, so we make a nice text. We know this. So for example, mouse. Okay, like this. And at this point,
as you can guess, we just parent the text
to the line like this. And then we can just tweak the initial position
of the text. So we calibrate it as well. We put the anchor point where
it makes more sense to be. So here, and we can just
snap it to the line. Boom, right? And then we just
scale it down, like this. We enable mochondu on
everything, like this. And let's see what
is happening so far. Oh, you know, that's it. And let's put these
things below everything. Let's see what happens. Okay, obviously,
it got screwed up. So let's just apply
some glow to make everything a little cooler.
Simple and efficient. You know, that's it. And then the possibilities are infinite. So you can also so
you can use this for a case like this where
you need to stem something from that
point of the face or to stick something on the
point of the past, you know? So for example,
when you use cases where you need to remove a spot, you know, that you have
on the face, you know, some kind of issue on the skin, you can use this
tracking technique to stick a little shape there, and then you can, for
example, remove that. So I'm going to show
you that later how to remove stuff from a footage. So you can use that as
a mat as a mask, right? That's there, like a patch. And this is a very reliable technique to do that
stuff like that.
24. Remove Objects: Finally, let me show you how to remove objects from a
footage or an image, which is something that is still related to tracking, right? So let's take this
footage as an example, first of all, okay, you
see this thing here. We need to remove that, right? And we're going to see also
how to remove static objects, you know, maybe one of these sheets on the
wall, you know? So let's go with
this second case because it's the simplest one. So let's first of all, change the work area
to only one frame. So just constrain
the area like this. You can also press B and N in case the circuit
doesn't work. Alright, so now we are
going to all on this frame. Okay, and then
let's make a mask. Using this tool here,
simply pen tool. You can use whatever method
you want for the masking. It depends how complex
the situation is. So in this case, I know that I can pull it off simply
with a standard mask. So I go here, here, here, we just surround the object
to remove simply, you know, with a couple of
clicks, click this, surround it. Alright else. Alright, by default,
the mask is going to isolate only that spot, but we need to do the opposite. In this case, we need to
cut a hole in that spot. So we pull up the
mask and we invert. Boom. Okay? If you
have multiple masks, you need to not use invert, but just keep change the actual mode of the
mask to subtract, right? So in this case,
it's only one mask, so we can simply invert it. Alright, so now we simply have we're a hole fundamentally
on the footage there. There is just
transparency right now, which is what we want, because now at this point, we can pull up this tool here. Content aware fill,
which is this guy here. And so you can see
it's already detecting that single spot
there in the video. You see this hole here. So now we just copy my settings here. So we have three types of cases. We have object when there's
something that moves, surface, which is what we
need is just a static object. And then we have some additional processing
that we enable. Usually, keeping
this on is good. So keep this on and
also on trunk, right? And then simply make sure that
the work area is enabled, like this work area,
and then generate. Boom. Alright. Let's
see what happened. Poof. Okay, we
have a first case. But as you can see the
mask was not big enough. So we see this line here, so it's not fully blended
with the rest around it. So let's go back one step, and let's try to expand
the mask a little bit. Maybe 15 like this. Alright. Okay. Let's try again. Now it should be better, Poof. Alright. Okay, we're
better, right? And what this tool does
is that uses I to kind of guess what's around it and
kind of feel it like a patch. You know, In fact,
it creates a PNG, which is a patch image. You know, let me
show you. You know, it's really a patch, you know, I post there to kind of fill it. And for people that
don't know your video, if you do a clean job, they will never tell that
there was something there. You know, if you want
to wait to make a seamless, most perfect filling, you go to Adobe Firefly, and then you use the inpainting model
to import your image. And then when you can actually have the best
contextual refilling or simply go to Pod shop and use the
content that we're fill. Of Photoshop, which is better because you
can type a prompt. Here is no prompt, automatic. But from small
objects, this is fine. And in case you need to place
something above this thing, it's going to be fine
because you won't let her see these imperfections, you know, that's the point. So it's always a game of
trade off, as you can see. Okay, so let's go back for now. And so that's just one frame, which is an issue, right, right? We need to make it extendable for the entire video, right? So we simply freeze this thing. Time, freeze? And now we can extend it simply
like this, go to the end. Boom. Alright, just like that. Now, here's an issue,
as you can see, even though it was
a steel frame, there's a change
in colors because the camera was in
automatic mode, right? So let's try to actually simply extend the video to a
longer area, right? So let's let's go back here. Okay, maybe let's go
back a little more. Okay. So even though it's a still place there's some
issues because actually, there's some
fluctuations in light. So let's re enable our mask. Okay? And actually
reprocess. Okay. What it does is that it analyzes
the footage and then it renders a dynamic sequence
of PNGs right now. So now should be
way more subtle. As you can see, po
way better, right? Because now it's
more integrated. It's not just a still image. So in the end, if there
are fluctuations, yes, you still need to use a video, even though it seems to be
just a steel frame, you know? So so now is really more subtle. At this point, let's stretch
you remove the second thing, which is this thicker
here in my shirt. So since it's a simple
shape, in theory, we could attempt to use the
standard mask simply here, here, here, we track it, right? We do a negative, a subtraction. But by default, I recommend
developing the habit of using mocha for these things
because it grants you the most precision
and reliability. You know, we have just more control for these
things at this point. So let's just select the footage and we go to FX and
we apply Mocha. First of all, just drag it
into the footage. Open it. First of all, let's
go here and here. Okay, so first of all, we create a mask around the
subject to remove. So in this case,
let's go here and go slightly wider than
the actual object. So it has space for error in
case there is high movement. And then let's disable
everything except translation, and don't change anything else. Change this automatic. And then let's track forward.
Should be pretty fine. Poof. You know, pretty
fast, pretty reliable. And, and then we just save Okay. At this point, we need to
duplicate the footage. Okay, let's call this Mt. And in the mat, we simply enable one
thing on the mocha. We go to Matt and show mat here. Alright. You see
this thing here. And you can start to recognize
where this is, right? So then we simply connect the footage to mat using track
mat in Luma mode, right? Because this thing here
because the mat by default, let me check exactly. The black of the mat is
actually rendered, right? So we need to use the Luma mode instead of the
Alpha mode, right? So you see that it's
actually rendered. So it means that here, simply, we check this on. Okay? So now, in theory
on the footage, exactly, we see all that
part of the video, but we need to do the opposite. We need to cut a hole there. So simply just invert, right, and that's it. So at this point, let's
see what we have. So we have a hole exactly there. And that's it. Now we can
also felter it a little bit. Let's go back to mat here. We can fer it just a little bit, a little, you know, okay? And then go to our
best friend here, the the content aware. So let's go here, work
area, strong surface. Okay, in this case, we
need to change object, right, because
it's moving. Okay. And if we don't need expansions, it should be pretty good
already that way and generate. Wait a couple seconds. Alright, let's see
play the video, and let's see what
all looks like. Poof. Alright. Okay, so
reason will be good, right? Because if you show
these two people, they may never notice that something is going
on down there, right? So, in case we can change
the look of that slightly by just shrinking the
mask a little bit, or just change one
of these modes. You may be subtle.
Let's see. Let's see. But, you know, the point
is that for most people, they will never know that we removed something, that
there was something there. You know, it's fundamentally perfect, you know,
and that's it, because people will
just look in the eyes, you know, the talking person,
you know, that's the point. And yeah, so that's
how to remove objects. So it depends on the
complexity of the object. But if you use Moka,
if you have patience, you can remove very complex
objects, even, you know. Yeah. And then sometimes in
the most dangerous cases, is to rebuild something. You know, for, we've
removed this cube, it may not fully understand
TV versus walls. So ins you kind of create
a little patch here, and that's where using Photoshop helps you rebuild
what's behind, right? That's the point. But yeah, for the most part,
it's crazy, right?
25. Displacement: Displacement maps are an extremely important thing
because fundamentally, you can use one channel of one layer to modulate
another layer, right? So every single layer has
at least four channels, RGB and the Alpha
channel, right? So fundamentally, red, green, blue, and the transparency. Let me show you in
practice how it works. So, for example, we have
here a texter file. Okay, let's make a new comp. Alright, so let's
suppose that we want to integrate a text, int this texture in
a realistic way. Alright? To do that? We
need to First of all, call this texture. Okay? In we create a text. Alright? Where do you
know this so far? Uh, let's say, happy, okay? And we center it, right, let's apply our classic globe. So normally, we can use the blending mode.
That already helps. Let me show you like
this. Alright, that already makes an
amazing job, right? But the text right now
it's still just flat. It's just perfect
symmetrical, right? But what if there was a way to kind of make it
actually interact with the texture to kind of make it actually spread into the
texture gradually, you know? So to do that, we need
to use displacement map. So we go here and we
search displacement. Okay. And we apply
displacement map on the text. First of all, here. Boom. Okay. And now
we need to tell it. What is the source of
the modulation, right? So the source is
the texter here. So we go here and
choose Texer Alright. And then we're going to choose the channel that is going to be used for the
modulation, right? So as you can see, these are all the choices
that we have available. As I told you, we
have all the channels plus the brightness, luminance. So in this case, since it's kind of a black
and white texture in a way, we can use pretty
much just luminance. So let's use that for the horizontal
displacement and the same for the vertical. And now what happens is that we can regulate the amount of this. So let's ho with 00 for now. And let me go slowly. Increasing the
horizontal displacement. Watch carefully what is
going to happen here. Wow. You see what is happening? Now I can slightly spread
the text into the texture. You see that now it's way more integrated.
So now it's zero. Let me go slyly. To the left. You see how now it's kind of spreading the text into
the textar, you know? And then let's try
also vertically. Let's see how much
we can push it, you know, so now it's fully interacting
with it, you know? That's the beauty of the texture and the displacement maps. And this has infinite use cases. You know, if you
have, for example, a glass distortion that
involves displacement maps. If you want to use this on
video, that also works. So just have fun because
it's incredible.
26. Depth Maps: Now that we familiarize a
little bit with displacement, let's escalate a little bit, and let's go with the
concept of depth maps. Alright? So we have
this image, right, and we want to be able to create some psudo three D
movement into the image. Alright. So there
is a way to do it thanks to the so
called depth map, which is something
that looks like this. Let me show you. Is an
image that looks like this. Alright. So
fundamentally, this is an image that detects what is closer and what is
more distant from the camera in an image
or a video, right? And then thanks to a
displacement map, you know, we can create a three D movement into the image or more things that I'm going
to show you in a second. All right? So first of all, let me show you how to create
this weird thing, right? So there is a free tool
that I'm going to send you. So all you need
to do is you drag the image into the tool and just press submit and we'll just give you this image and download it and put it back here, right? And in case the size is different when you
import it like this, for example, make sure
that you stretch it back to full screen
and precompose it. So for this rock to work, the two layers need
to be normalized. So the scale on both layers need to be
fundamentally 101 hundred, right? Okay. That said. Let's actually go
here and let's call this depth and this is
going to be the picture. Okay. So the depth map
means should be invisible. We don't need to see it in us. Let's keep it where we
want. Let's keep it here. Okay. And now we apply a
displacement map on the picture. Okay? And we change the source of displacement to the
depth map, obviously. Now we change the source of displacement to
luminance again. Okay. And let me show you
what is happening already. So if I move this. We see what's happening, we
start to be able to kind of rotate into the
entire environment as if we were in a three D
place. You know, it's crazy. But this was just
a three D image. How the hell is that possible? Well, that's the beauty of the three D displacement
with that maps, you know? So obviously, you
cannot push it damn much because then it's gonna
show the trick, right? But if you move gently just
a little bit, you know, there's an area where you
can use it, you know, to kind of make a slight
three D movement, you know, very quickly. This is also called parallax,
in a way, you know, so this is already some form of so called
parallax, you know? Okay, let's engage
these two parameters since these are now interesting. So and we can make
it automation. You know, we can
make, for example, 0-3 seconds here we
go to zero, zero. Okay, and that's
really. One thing. You know, you see that now we actually have some movement. But depth maps can also
be used in other ways. So we can also use one effect
called camera lens blur. Okay? Which by
default, this effect, all it does is you just blur in the most
realistic way, right? Simulating an actual
camera, right? But it has a feature
called blur map. So if we connect this to
a depth map like this, now we can control the
focal distance, right? So let me show you what happens. If we move this value
slowly, let me show you. You see what happens?
Now we can really change the point of focus in the image as if we were controlling the camera into
this environment, right? So we can choose near
focus versus far. Right. It's crazy. Right. Okay, so that's a parameter
that we can automate, so we can engage
it with keyframes. Alright. And now let's use, for example, we can also synchronize all this with a scale in the image. So let's engage, also
the scale. Alright. And let's scale in
slightly. So we go here. We go to 115
probably. Like this. Alright. And while we zoom in, we go from this focus
level, a new one. Let's go with a cube and focus. Let's see if we
can find the spot. All right. Maybe we
can halt this number. Okay? And finally, we can expand the borders of the image by
using a classic reptile. So we go here, like
we already know this. We apply the reptile
before everything. Remember this. And the blur needs to be applied after
the displacement map. Otherwise, you will see
some imperfections. Okay, let's see what
we have so far. Let's smooth everything
with some busier. Okay, let's make half so
it's quicker to render. All right, right. We're
getting there, you know. And let's also automate. Let's also smooth
the D map, right? So we can simply apply a
Gaussian blur on that map. We do, and we increase it to 50. So the death map right now, technically looks
like this, right? We are kind of smoothing
the edges of it, you know? So to make the displacement map aware of this Gaussian blur
that we added on the layer, simply go back to the
displacement map settings and change here from source
to effects and masks. So it will also take
account of that, you know? All right, so now we start
have some more movement, okay? Let's also maybe add some
more moment. Let's go here. Let's see how much we can
push this value choose, so maybe we can go from this situation to
this. Let's see. Okay, that's way better. Alright. And then also let's
automate the amount of blur here from 15 to
maybe way lower number. Maybe it's five or three. Yeah. Okay. And finally, let's
add maybe a camera shake, you know, some kind
of handheld shake. So we can simply
use some wiggles that we already know
about. So we go here. I just saved a preset that uses the same
principles I already explained to you in the
past. So we go here. We made the wiggle a little
slower and a little wider. So 0.5 degrees 1
hertz frequency here, we can say, again, one and ten pixels. Alright. Alright, poof. You see how now the scene starts to really come
together, you know? And, yeah, and then
we can just have fun. We can have some Color
correction. You know, let's see. Let's I'm gonna
explain this later, but for now, let's just do a
quick adjustment like this. Alright. Let's see how the full scene starts
to come together, you know? That's it. That's it. So as you can see, that maps plus
displacement maps, plus this trick with
the blur allow you to adapt to an image and make it look more three
dimensional, you know? So, so remember
these things because the uses of this are
fundamentally infinite, alright?
27. 3D Camera Principles: Et's talk about
three D. Alright? So let's make a new
composition here else let's make ten
ADP should be fine. All right. Okay, so let's
work with text for now. Let's go here let me show you one thing.
We type something. Every single layer normally
can be converted into three D by clicking this icon
here, this cube here. When you turn it
on, fundamentally enables the third dimension
on every property. Rotation, scale, and position
now have the Z axis. Alright? So, the first thing that I do when
I make a layer, three D is to calibrate it. So we just change the anchor point to the
center of it, right? So you can simply like
with three D layers, go at the center of the layer
first when you see XYZ, and then you can
change like grab it from there when you see XYZ and snap it at the center of it. Alright, now it's calibrated. So let's put at the
center of the frame. First of all, so go here
and move it. Alright. Okay. And that's now fully
calibrated and centered. Okay, so where do you know that we can move the layer now
like this with a mouse. Were you know these things.
So we can skip that, right? So instead, let's
talk about cameras. So cameras are
fundamentally a way to have the point of view
control on the entire scene, you know, so as if we were
exactly in the scene, right? So you can imagine
the camera like a giant null control that controls everything
at once, you know? That's what it does
fundamentally. And it's very convenient. So let me show you.
Let's right c here, and let's create
the first camera. Okay, there are two
types of cameras, the one node and two nodes. So let's start with one node,
which is simpler to use. Okay? Let's call this Cam one. Okay? Don't change
anything else. Keep everything default. If you have specific requirements, you can simulate specific
types of cameras. So in this case, let's go with 35 millimeters and don't use the field for now
and just confirm. Okay, so as you can see, when you enable the
camera, normally, the view doesn't change. It remains exactly the same. But let me show you one thing. As soon as we change the
position of the camera, the Z position, we get closer to the element. You
see what's happening? Because we're Lily now moving into the scene,
closer to the object. So we're not making
the objects bigger. It's just us getting
closer to it, right? Okay. So before we do anything, let's actually arrange multiple layers in the three
D environment. So let's duplicate
this layer, okay? And we shift slightly like this, and then we change the
content a little bit, just make a different word. Okay. Two. And so we're going
to push it somewhere else. So instead of making it smaller, we're going to
actually move it away. So we push it with a Z axis somewhere else, somewhere there. And here's the first issue, as you can see, doing this just with this view can be
a little confusing, right? So we can enable
multiple views by double clicking on
the composition tab, and then we enable
two views here. And then we can
see the scene from the top as well, so
we can zoom out. You see what is happening now, the camera is here. We're here. This is our point
of view right now. The word one is here, as you can see here, and the word he is
exactly down here. You see where it
is? That's where we pushed that so far, you know? And we can go as
distant as we want. We can push it all the way. You know, so just push
it to a useful distance. So for now, this
should be enough. Let's put it here. Okay? So here in the video
looks smaller, but actually, it's just distant
from the camera, alright? Okay, and then let's
make a third word. So let's duplicate this one
more time. Let's go back to. This view, okay, let's
call this three Okay, and we're gonna put it on
the other side of the frame. So let's go here somewhere
here should be fine. Yeah, exactly. And maybe
even more distant. Like, those Okay. And also, you know what, let's rotate
slightly these three layers. So that is going to
have that rotation. That's gonna be slightly
different like this. Okay, that should be fine. Okay, so let's first of all, increase the ecstatics
of the layers. So adjustment layers
work as fine, even if the layers are three
D. So let's call this post. We use our best friend,
the default go. That should be fine. All right. Exactly. Okay. So now, let me show
you one thing. When we move inside the
environment using the camera, fundamentally, we can also
rotate our point of view. So remember that fundamentally the camera right now represents our head into the environment. So it means that if
I go into the scene, if I go like this, I go closer, let me enable the chew view, so we see exactly
where I'm going. Alright, now we're here, right? So if I go closer
to the first word, obviously, we see
a bigger, okay? But suppose that I surpass it. Obviously now's behind me, okay? So now I'm surrounded by the
words have word two here, word three is there, word
one is behind me, okay? So if I want you orbit, I can do it by
rotating the Y axis. So I just move it with a mouse like this, and see
what is happening? Right now, it's as if I kind
of looking around the scene. So I kind rotate until
I will see, obviously, the first word, but reverse because we are behind it, right? That's the point. So so that's I think it's one of the best ways to
fully understand the first type of camera, you know, so you simply move
around the environment, and you can rotate your point of view at any point, right? And all these things can
be automated, right? Okay. So before we start
with the animation, let's first do the same thing, but with a two node camera. Let's disable the camera. Okay? And actually, let me
show you one thing. If I go to a specific position and I then disable the
camera, you see what happens, the view shifts to a default view because you
can bypass cameras instantly, just like this, you know? So the eye actually turns the camera off and makes the scene return
fundamentally to D, you know? Or just without cameras. That's the point
that keeps the point of view static, you know? Okay. And one more thing, only one camera can
coexist. Per time. So if you have multiple cameras, only one of them would
be active at a time, which means my recommendation is to if you need more
cameras, you can do it. But make sure that they
are kind of sequential, you kind of trim
one camera when, you know, until the
point where you need it, and then you switch
the second camera, like this that way, it will switch to
the second point of view, with that
camera, you know, so they cannot coexist, even if you get them
simultaneously, unless you precompose them. So if you precompose the
entire scene with a camera, yes, that's one trick to make
multiple cameras coexist. But normally, they are
not meant to do that. So you keep those
sequential, right? Okay, also, just to
tell you that using multiple cameras
like this can be useful if you want to experiment with different types of cameras. So maybe they're gonna have
the same exact automation, but they're gonna be
different cameras. Maybe one has a wider angle, one is narrower, right? That's one case where
you may want to have multiple cameras at
the same time like this. But you can only turn
on one at a time, right? Okay. That's said. Let's just step back. And let me create a second
camera for now. Let's go here, new camera. And let's make it to
node camera instead. Okay. So the difference
with a node is simply that we have an additional value called point of interest.
So let's go here. You see what we have here.
So the point of interest, in other words, is
the anchor point. So the one node camera
doesn't have a anchor point. The two node has. So by the anchor
point for a camera, if fundamentally
this thing here, and the feature of this point
is the fact that doesn't matter where we go with the position of the
camera, let me show you. So so right now we're here. The position is here, right? This is the anchor point. So now even if I move
away horizontally, you see what is
happening, the camera keeps looking at that spot. So that's the
fundamental difference with a one node
camera, you know? So, in case you want to have a specific target to
look at all the time, you need a point of interest to make it easier
to manage. You know? So if I go horizontally, vertically, it doesn't matter. I will always look at
that target I chose. And obviously, the target
can change dynamically by automating a point of
interest position, you know? And if you don't
need this, simply go with one node camera. That's the simple question
that you need to ask yourself in case you need to use one node versus two nodes. So in this case, as you can see, if I rotate, you'll
not care about it. It will just change
the point of view like this because there's no target. There is no point
of interest. So
28. Camera Animation (Basic Method): Let me show you how to approach animation when you use cameras, right? There are two methods. Let's first choose
our initial position. So we're going to start
with this exam position. Let's actually enable
keyframes on X Y Z on both rotation and position and also orientation. All right. So we stat to have
all the keyframes ready for ready to go. So let's go to 1 second
into the timeline. And the cool thing about
this method is that we can select the target where
we need to go to. So in this case,
the word che here. And then we simply click on View and look at
selected layers. Boom, of. You see what happened? It will automatically go to phase the target automatically. And since we had the
keyframes already enabled, now we have the coordinate
to go there, exactly. And then we just need to regulate the rotation
a little bit. Let's go here. All right. And as you can guess, now
it goes from position one to position two.
Wow, just like that. Right? So then, for example, we can go back to the
first position again. So we go here, for example, around second two, right? So we copy and paste
the keyframes, and then we go to word three. So we go again on second three. And then we just click it, look at selected layers. And then again, just regulate
the angle like this. Alright. So now we
start from center, then back to center and
word three. Alright. Okay. So at this point, I'll start you make these key
frames a little smoother. So we simply enable
sie everywhere, and then I'll start to make
everything a little cooler by enabling motimbler Okay, and that's already a
first way to manage animation for elements that are arranged and in the environment,
you know, very simple. So we can case also time
the curves like this, you know, remember that you can also do this at
any time, like this. So there's some timing
that you can still do, and that can feel
also very good. You know, you can do
stuff like this, right? Kind of slow down. So there's a certain level of expression that you can achieve
already with this method.
29. Camera Animation (Advanced Method): So far, with first method, we are pretty much
able to go from first target to a second target, and then go back to
the source position and then jump to wherever
or target in sequence. So this is really good solution. But if you really want to have smooth motion, we need
to escalate this. Well, let's go to the very start of the composition, right? Let's delete all the keyframes because now we're going to
do something completely new. First of all, let's make a
noll object. So let's go here. No object and and that's where most
people make a mistake. So you make the null
object three D. Okay. Let's name it Control one. Let's go here. Okay. So let's see where three
D where the null is. Okay, as you can see,
the null is here, whereas the camera
is here, alright? So, most people start to parent the camera like this
straight the null object, and then they start to make
the technique I showed you earlier with
multiple null objects, one per movement.
Okay, that's fine. But they forget
always do one thing, which is to calibrate
the null object. So before parenting the
camera, we actually do this. So before parenting, we
select the null object, and we use the trick with
shift parent to the camera. That way, look carefully what is happening here
to the null object. Shift parent to the camera. Boom. Woof. You
see what happened? Now it goes exactly where the camera is completely
synchronized. That way, when we finally
parent the camera to the null, fundamentally, the
numbers that we see here represent fully the
coordinator of the camera. That's the point of this, right? So it's going to be
way easier to use, and you're going
to see the reason in a couple of seconds, right? So the camera, obviously, we don't touch it, let's
keep it like this. Alright. And let's start
make the automations. So let's go there are
two ways to do this. We'll enable XYZ, as always, may be also here. Let's enable all these things. So the method is pretty
much the same now, so we can simply go
slightly the future here, and we can either
grab now the null. And we just move like this until we snap to the target.
So the target is there. So we can either approach
it slowly like this, one exit at a time, like this, you know, and that's
already one way. Or you can just grab
it from the center. So if you see X Y Z, you can also grab it
and snap to the target and then go slightly
far from it like this. And that's already one
solution, you know. And then there's also a
slightly more complicated way. If you really like the
feature I showed you earlier, where you select a target, and you just select the
target and you go to view, look at selected layers. Thanks to the fact that
the null is calibrated, we can still use that feature, but it's slightly different
the way we use it. So we need to do this. We open the coordinates of
the camera here. So we select the
target here and we go to view, selected layers. Alright, so you see
what happened now. The feature is still working, but we need to do one thing. These numbers here on the
camera, we need to cut. We need to cut these. And add those to the actual key frames of the position of the control
that we're using, you know? So it means that we need
to add those to the X now, so plus that number. Then again, cut from
Y and paste it here, plus, again, so
plus that number. The camera in the end
needs to stay 000. So those numbers are
just the numbers that I need to extrapolate,
you know, from there. So now from this number, again, we go here at the end of
it, plus that number. Boom. You know, so Thanks to the fact that we
calibrate the null objects, we can still use that feature. All right? Okay, let's continue. Let's go to 2 seconds. Okay, let's go back. Now we
need you a second moment, right, so we need to
duplicate controller. Okay, and we return to the next position using the new controller.
So we go here. And obviously,
remember to parent. Obviously, parent Control one to control two before
doing any change, okay? And now it should be
fine. Let's go here. Exactly. Alright, so for now, we're going here, then we
go to two, go back here. Okay, let's go to number three. And then we start to do the
smooth the smoothness step. So we need to a new movement. So Control three now, right? So we delete all the
previous key frames. Parent Control 22 Control three. Alright, now we're
ready to move it. So let's see where we are. At any moment, you can
see where you are by just clicking the latest controller. So we are exactly here,
which makes sense. We need to reach this point. Okay, so let's do it
manually, for example. So we go here, X Y
Z, move move move. Go to the target, snap it, and then go
lie away from it. Alright, that's it.
You know, and Okay. Let's also maybe also rotate
it a little bit. Let's see. As you can see, the Zac is still working completely
synchronized. So remember to
calibrate ole object. It's so important
because otherwise, these rotations will start
to not make sense anymore. So that's a very
important trick. Alright, let's rotate it
a little bit, like this. And finally, maybe let's go
back to the initial position. So Control four, parent, And actually, we
just copy and paste. We just move these in theory. Should be fine,
let's see. Exactly. Wow, that's it.
Alright. Okay, so now we start to have all
the key frames, right? So to smooth these guys, we need to, first of
all, select all of them. And such you enable
busier. Boom. Okay. So now we simply use
the trick I showed you in the motion
section where I explain the technique
where we start each next movement before
the latest one ends, right? So, fundamentally, we start to overlap these slightly
like this, you know? And how much you need
to overlap these depends on how smooth the
animation want to be, you know? But the
principle is this. We cannot intersect each movement with each
other, you know, that weight is going
to be a area where each movement starts to
average with each other, you know, and that's what
really produces the smoothness. So let's go here, here. Okay, let's make the first
move a little slower. Alright. And then if you stretch
the key frames a lot, you may need to reset the Bzier. So reselect everything
and repress Fn one more and more. Okay. And what else? We have
the motion blur enabled. Let me check. Exactly.
Okay, let's see. At this point, we don't
need the key views anymore, so we can go back to one view, full screen, Alright, let's see what we
have so far, Render. Let's give me some
time. Alright, so we are here at the
beginning. Let's see. Okay, so for example, right now, the interpolation is so
high that it doesn't approach the word
two even remotely. It's just it already starts
the next moment too early. So we need to tweak it, right? How do we tweak it? Well, simply go to the first moment, okay? And we need to simply tweak
the first few moments. So for example, we can make
this start slightly later, not by moving the key frames
by changing the curve. So the second moment,
we go to its curve. You make it start a little
slower, you know, like this. And make sure that each
moment fundamentally ends with a slow inertia. So in the end, it doesn't
matter if you do this or this. Here it needs to end in a slow
fashion. That's the trick. So here again, let's delay the Z axis of the
second movement. That way we have more time for the first moment to complete. So let's go here.
Maybe we can make it a little faster like this. Alright, so now we're
regulating the first movement. So you can see this
technique obviously is a little more complex because you need to do these regulations, but, I mean, you need to choose if you have
time, you do this way. Otherwise, if you
don't have time, use the first method, you know? So let's see. Wow. You see
what is happening now? Klas also continue with the smoothness for all
the other movements. So we go here. Movement three, same thing, so we go here and slow these curves at
the beginning of it. Same here, and same here. Boom. Slow it at the beginning. Oh, right. And then what else? Fourth moment also. Let's see. So, this can also start
slightly earlier, like this. Alright. Let's go
to the key frames. Okay, so this can be
delayed slightly like this. We're almost done. And
this goes like this. Oh, right. Okay, so let's see. Let's render all this.
Let's give it some time. Okay, Woof. You see
what is happening? Now, instead of zigzagging, which in the coordinates of the different steps
of the animation, there is a averaging
between each step. You know, kind of slows down, then smoothly changes
to the next target. So it kind of navigates into the scene in a smoother
and more integrated way. So that's the secret for
this second method, alright? So again, the trade
off is that it takes longer to realize
something like this, compared to this, this kind of quick, you know,
quick and dirty. But actually, sometimes
this is pretty efficient. You just do this way and just pull it off, you
know, that's the point. So when you do have time,
use the first method. You know, which is a little
more staggery, right? But it does the job, right? When you have the time,
use the second method, you know, kind of
whoo whoo, right? So, so, and this is incredibly powerful.
You can do what you want. People use it for anime edits or incredible transitions
in real footage. The implementations of this
technique are infinite. You, you know, so have fun.
30. Depth Of Field: Before we go away,
let me just show you one couple of things
about the cameras. So the camera has also a
feature called depth of field, which fundamentally is
like in real life property where according to the
distance elements, we can keep only one element
in pocus at a time, right? So let me show you what I mean. We expand the camera
properties here. Okay. And we enable depth of
field, first of all, boom. Okay. Then let's go for a
second back to the views. Okay? Let me show you what is happening here on the
view on the right. Alright, so this is our camera. We already know it, okay? So, now that we enable
depth of field, we should see one
additional parameter, which is the focus distance. So when I move this line, you see this line here, that's the line where everything will be fully in focus, you know. So it means that if something is outside of that line either
before it or after it, it will progressively
get defocused, you know? So let's, for example, see where the number
one is. Okay, is there. So let's make sure that it's exactly coinciding
with the focus line. So we go here. It's gonna
be exactly there, right? Whereas, we already know that the other
elements are there. So now let me show
you one thing. If I increase the aperture,
you see what's happening? Poof. Now the other
elements that are outside of the focus line, obviously they are getting
out of focus, right? So that's the very
important thing to add realism to
your scene, right? And obviously, this reacts dynamically to the position
of the camera, right? So you can already guess
what's gonna happen if I play. Poof. Obviously, when we approach the words,
they get in focus. And again, back out of focus when we get back to
word one, right? So this is additional thing that you can add
to your projects. Be careful because this makes
thing heavier to render. So maybe you can
toggle this on and off dynamically as you
need to check, right? But this is very
easy way to use it. So regulate the focus distance
first with this prometer. You know, and then just
increase the aperture. You have more or
less blurring for stuff that is out of focus. And then, these are kind
of more nerdy things. Don't touch these
if you don't care. And finally, the zoom level here is fundamentally the type
of lenses that we use. So, fundamentally,
if you want to have a shy effect, change
the zoom level. Let me show you let me show you what happens
to the camera. If we squeeze it like this,
you see what happens, it kind of allows us to see more of the frame
at a time, right? As if we were
installing a fish eye, but it comes with a
cont of a fish eye, which means that we distort the image a little
bit more, right? So whereas if you want a
macro view or a tunnel view, you can do it by doing
the opposite, you know, kind of a telescopic zoom lens. You can do it that way, you
know, like those, right? But as you can see, obviously,
it's very narrow now, so we're going to have
a very narrow view.
31. Basic Extrusion: Let me tell you one
thing a little extra. In Arafax you can change
the renderer here by enabling fundamentally changing some features
for three D, right? There is one mode
to make the text truly three D by changing
to Cinema four D here. And this has an advantage, which is the ability
to make text truly three D. So now that I have Cinema four D enabled, let
me show you what I can do. I can go here and I can enable geometry options
and extrusion depth. So when you enable this,
to see what is happening, now the text is actually
three D has truly depth, you know, and also reacts with light we're
going to cover later. But you probably notice that we lost the ability to
have depth of field. So that's kind of a trade off. You either have depth of
field but only two d layers or truly three d layers
and not depth of field. But actually, even
that that's not true. There is a way to
work around it. So actually, well, let's go
back to the normal renderer. Okay, now we have Dulfeld. Okay. What if we precompose one? So we precompose this word, right into its own
container, Okay? And we make a three D. Alright, so now the field is
still working, right? But inside inside it, we can now enable a local cinema four D
just for this word. And that way, we have the ability to make a three D again. Without breaking the depth
field on the outside, right? So we can kind of
rotate this word, keeping, you know,
the perspective of the word a little
more interesting, right? And here at this level, we are now still with
depth fulfilled available. So that's a workaround
for these things, right? So always don't get miscouraged. You
know, think about it. Think about all the
things that we can do. There is almost always
a work around issues.
32. Character Rigging (Puppet Pin Method): Alright, so when it comes
to animating characters, you can always find yourself
in two major situations. The first is the
case where you have only a pre rendered frame
of the character, right? So the second case
is where instead, you have full control on every single piece of
the character, right? So we're going to analyze
the first case first, which is the easiest
to kind of understand, and then we go more complex
with the second case, right? So I suppose that we
have just single image. So right now, let
me simulate that by precomposing all the
pieces into one, so we say character one. Alright. Okay, so we have this
single image, right? We can already notice
the first issue, the fact that it's
just pre rendered, so we can now really
have full control on making a pose that
is significantly different compared to the
initial one, alright? But there is a way to still have some control over the pieces
of the character, right? We need to use one tool
called puppet pin. This one here. So you click
this tool and we start to zoom in and create one joint on every single joint of the body of the character. So in this case, we
place one on the neck. One on the shoulders, you know, when they attach to the body and then on the forearm hand, let's repeat the same thing
on the other side. Okay? And then one between the
body and the waist, right? So exactly here. Okay? And then one between
the waist and the legs. And finally, between the ankles, fundamentally on the
knees, you know? And then down here between
the feet and the ankles. Alright. Okay. So now
we place one joint, one point on every single major
joint of the body, right? And you will actually
notice one thing. After you Do that, see
what is happening. Now we should be able to move
every one of these pieces. In an elastic fashion, you know, so we start to give elastic property to the
body of the character, which gives us some control. You know, that's already
a micro victory, right? So we start to have
some level of control. So we can not do one thing. For example, we
don't need to really control certain joints directly. So this, for example, is
useless, this is useless. So let's just make this become automatic.
So how do we do it? Well, first of all, as
you create each point, it will automatically place one keyframe in the timeline
inside the puppet effect. Let me show you here.
So if we press you, we see all the
keyframes already. That I have the coordinates
of every one of these points. So when you move these, it will obviously change the
position of that joint. So if I move this, let
me see which one it is, obviously, it's
down here, right? If I move it, you see
what is happening? Okay. So let's stop that
issue I was talking about. This, for example, we don't
need this elbow or in, so to make it become automatic, we just identify it first.
Okay, so it's number three. So this is the way it works. We don't need keyframes on this, so we disable that, and then we compress it and re expand it. That way we see
this option here. This is the type of pin. There are three types position, starch, and a band. You can ignore band because
it's kind of useless. So, starch is what we need, because when you enable starch, it will not give us any
key frame because now it's kind of a rigid
point that reacts automatically to the
points around it, right? Okay, and in fact,
I can now move instead the hand which
makes more sense and who. You see what happens? The elbow automatically reacts
according to that. Alright? So we can do the same thing
here on the other side. Let's see which one it is, okay? Here, here, and
just enable starch. Alright, so now also this reacts to that
automatically, you know? So, in yada, yada. So, same thing for the
knees. So let's try here. Let's go's see what happens. We just enable a starch. And that way, have
a behavior that makes a little more
sense, you know, right? As long as you then animate
the feat in this case, in a way that makes
more sense, you know, in a coordinate that makes more sense, it will fully work. So obviously, here we have keyframes, so we
have keyframe one, and then we can go
to keyframe two around here and we go in
this position, right? So now, obviously, the character will do that movement, right? And also we can use
a little trick to auto make it loop
infinitely, right? Because maybe you want to
make idle animation, right? So to do that, we need to first close the loop with
the keyframes. So we have keyframe one, keyframe B, and then
return to keyframe A. So we return here. Alright. So obviously, cup and paste in these keyframes infinitely is
going to be very annoying. So we're going to use
an expression as, yes. So we simply just click on this property
with to enable Expression, right? Where do you
know how to do this? Old click. And then we
simply go to one of the presets here,
go to property. Property and then
loop out duration. This is what we
need. This is one. So when you click
this fundamentally, it will read all the keyframes in that property
from the beginning. And then as soon as it
ends the last keyframe, it will immediately restart the loop of the three
keyframes in this case, right? And in other words, now, it will automatically loop
that moment forever, right? So that's one way to
make a loop very easily. So you just create one cycle
fundamentally and then it will repeat it the whole
time, right? That's it. Okay, so let's now, for example, go to the neck. Let's say that we want the neck to move
slightly like this. We can simply apply our best friend, the
wiggle expression. So we go down here, we apply a simple wiggle,
and we say, Okay, it's gonna move one
times a second and five, maybe ten pixels max omnidirectionally.
Alright? Let's see. Now what happens. Alright.
So the characters should be a lot
more alive, right? And, yeah, and all
the other things are just connected
work the same way. So you either move it manually
with key frames, right? So you create the different
poses over time, right? Or you use expressions to kind of create loops or
randomizations. But the only limitation is there's gonna be some artifacts if you move too much in
the position, right? Because it's gonna squeeze
the character a little bit. So, this method has some
limitation. That's the point. But for small motion, this may still work as
a method, you know?
33. Character Rigging (Duik Angela Method): If the puppet pin is not enough to animate
your character, I should escalate the situation. So first of all,
you need to have separate control on
every single piece of the character, right? So you need to break it down. So if you have a starting image, you need to break it down using just duplicating the
layer multiple times. And for each layer, you make a mask to isolate exactly each piece
of the limbs, right? So or if you have native
control over those parts, you already have those, right? So in my case, I generate
these using shape layers. That actually can
be a good idea. You put the image blow and you rebuild natively
using shape layers. That's valid. That's
a strategy, right? So after you have broken down
all the pieces for safety, I recommend precomposing every single piece in
the final ring, alright? So let's because sometimes maybe if you duplicate
some pieces, they may still result as copies of each other and
not independent pieces. So, to make sure that they
have a unique identifier, I'm going to precompose
each piece like this. Like this. And I'm going to
do it for every single layer. But there's an issue. As
you can see, already here, already positioned the anchor
point in a better place, you know, where the
joint should be, right? When I precompose it, I lose
that information, right? So I need to redo the
same thing again unless I check this checkbox on Okay. And at that point, when I
reposition the anchor point, it will actually detect the anchor point that already I already positioned inside. So then I can just easily
just reposition it like this. And now I'm just skip forward because I'm going to just do
this for every single piece. Alright, so first of all,
let me show you what it did. So let's go in order
through all the layers. So we have arm left, which is the first
piece of the arm here. And forearm and hand. So you can see watch carefully where I place
the anchor points. One, w, three, one,
three. Alright? Cool. Then I repeat the same thing for the other arm, same thing. Okay. Then we have a leg left, the knee, and the conjunction between the ankle
and the feet, you know? So one, w, three. Let me show you so watch carefully the anchor point
again, one, two, three. Okay, so then same
thing your leg, right? And then the most confusing
the head, chest, and waist. Alright? So the waist, you should place it around
this spot here, which is connected then to
the spine to the chest. You know, I call the chest,
but you can call the body, where and the chest is
the one that's different, place at the center
like this, okay? And then finally, in
case you have a neck, you can have additional piece, or in my case, there's no neck. So simply straight to the head, and the head has the
anchor point there, right? Okay. So that's how use this as a rule of thumb to make all your raking pre
raking fundamentally, because we still didn't do
the actual raking, right? Okay, this is the
preparation of the layers. So once you have this situation, you're ready to pull up
a plugin called Duik. So **** is an
amazing free plug in open source that you need to download from this website
that you see right now. Just follow the guide,
download for Mac or Windows, wherever you use, and fold the installation guide it's
very easy and reboot Afrofax. And then you will
see it pop up here, Window scroll down here and you see ****
Angela. This one here. That's the plug in.
So you click it. This is how the
plug in looks like. Alright? So first of all, we're going to create the bones. That's how they are called. We click here, bones and we select the type of
animal that we're going to use. So in this case, just human. And as you can see, we had all the different
types of limbs, and we need to select the pieces of each limp
in a specific way. Okay, so we need to
fundamentally select each limb from its
base to the tip of it. Alright? So in this case, we click on the arm piece first, then the forearm,
and then the hand. Okay, so now they are
all three selected. In this specific order,
we click on human, and then we shift click on arm and we choose a
couple of things. So, for example, the
orientation must make sense. So now, that's right. Front back in this case, does make sense because, okay, let's say that this should be in front because it's the part that is visible above the body. So let's say that it's
in front, in this case. And then we choose
how many pieces. We are three pieces,
so that makes sense. One, two, three, right? And then while holding Control, remember this is very important. Keep control pressed
the whole time and then confirm arm here, confirmation. Alright, wit the bind completes, and you will see here a
couple of things happening. Then if you see your pieces
engaged with parenting, me the you did it
right. Alright? If this remains empty,
maybe you forgot to press control,
something happened, so retry again until you see exactly as I have it here
on my side, alright? Okay. Then we start
to see the bone. So this is one bone, how
it looks like, you know? And the tip needs to
be tweaked manually. So we go here and we
just move the tip of the bone exactly in sync with the tip of the actual
limp, you know, like this. Alright. Okay. And
that's one bone. Now we just repeat
for the older arm. And good news. Now we just
select it. So, same thing. Arm forearm hand and we go, again, human here, arm. So now we don't need
to shift click. This time, we can straight
just control click like this because Dweck is aware of the past limbs that we
did already of that type, you know, so it will
automatically say, Okay, probably you need to
do the opposite side now. And in fact, that's
what we need. Alright? So let's
tweak the tip of this. Alright. Here, I just forgot
to place the anchor point a spot so just take your time when you place
the anchor points, right? So then let's go to the leg. Let's go here. Let's
see what it is. Leg, ankle, leg,
ankle, and foot. Obviously. Okay, one,
two, three, right? And we click here. Human. Shift click Shift click before, because it's the first
leg still, okay? And this is the right side. This is the front. And
it's made of three pieces, thigh, cuff, and foot. Okay, then just control, click again. Remember,
control, click. Very important. Alright,
then again, tweak the tip. Okay. And finally,
the other leg. So let's go here, one, two, three, Human, and then
control, click straight. Alright. So as you
can see, the process becomes easier and
easier as you progress, we're almost down the way. And then we go to the
hardest one, the waist. So the body follows
this in principle. So from the origin point, from the waist going up. So we have then it's
like the waist first, then the chest,
and then the head. Okay? And finally, we go here, human spine shift click. Okay? So we have the head, yes, we have the hips, yes. We have one piece for the
spine or next zero pieces. Alright, so and then just
control click on spine. Okay, Let's tweak this here. And now it's fully rigged. Okay, we have all the
bones that we need. Okay? So, now it's
actually the magic. So we simply select
all these layers, the points, both these
layers that appeared. These are called guide layers. Fundamentally guide layer
has this icon here, and its property is it will be automatically excluded
from the render, right? So let's select everything
at this point. Control A. And then let's go scroll down here on the plug and
you see Auto Rig. So simply click it and wait a couple seconds.
We'll take some time. And now fundamentally at
the top of the timeline, we're going to see a bunch of
layers called controllers. Those are the only
layers that technically, we will just need to animate. All the rest fundamentally
can be hidden or almost. So I'm going to show
you a few seconds. The length of this
frosts depends on the power of your CPU, but usually it's
over within one, two, 3 minutes, you know,
depends on your computer. Alright, so let me show
you what happened. So at the top of
the timeline now we have these new layers
with a C. These are the ones that
we need to move to change the position off to animate the character,
you know, that's it. So all the rest can be hidden. So let me show you because, as you can see, we actually
have so many layers, right? So these are our actual
visual elements. So, in case you issue, keep
those visible because maybe you initial apply some
effect, keep these visible. But for simplicity for now, let's hide everything, you know, so we select these we don't need to see
the bones anymore. So we select all
the bones layers, which let's unlock these. So we can select
also the other ones. So unlock everything. So you fundamentally select the first layer with
the star that you see. And so you can select
everything until here. So don't select the ones with the C because those
are the controllers. So all these can become
off actus, right. And for now, we can also
select everything again, actus, except the
ones with the C, and then enable this icon here. The shy icon allows you to hide something from
the hierarchy like this, boom, and then press the
same icon again here. Right? So now we have
only these layers that are the only
thing that we need to add at this point. All right. So let's go here, so let's go here. Frond, we don't see
the plug in anymore. All right. So let me show
you what is happening. Now, we can start to verify if the rigging
happened correctly. So if we change the
position of this icon here, so let me just move it
slightly in a better place. Let me rotate it. Okay. So each of these layers
controls this icon here, which can be moved with a mouse, and let me show you what
happens if I move it. Woof. You see what it's doing? Is the so right now, we're kind of holding the
hand of the character, and obviously, the
forearm and arm are reacting accordingly,
automatically, right? So let's check the other
arm as well just oriented. As you can see, each layer has its option to orient the D, the icon make it more sense, you know, make it more sense. Okay, let's move it. Alright. Yeah. This also
works fine. All right. If the angle of the elbow
is not what you need, if you need to reverse the angle of the elbow or the knees, simply reverse this value here. You see this value here, side. So if you reverse this number, so in this case, from -100 to 100, it will flip the verse
of the elbow, right? So now, it will do
it this way, right? So, so but in this case, let's just return to normal
because it was an experiment. Alright, okay. And
then let's go. Let's check all the pieces. So check the rigging happened correctly
before doing anything. So let's search you
move this around. Let's see. Yes, sir, this works. Alright. And let's
see also this Poof. You see what's happening.
Everything is working fine. And finally, let's we probably need to disable
one of these icons. We don't need it. So the head is probably doing the
head. Yeah, poof. You see how everything
is connected. And then this icon
probably we don't need. The body. Let me see. Yeah, this kind of
changes everything at once. You know, how cool is it? You know, everything
is connected. You have an idea how complicated this would be to do
it manually, right? So there are kind of
complex relationships between every single piece now. Okay, for example, we don't
need this controller, really, so we can disable
it because we can control from the chest,
probably. Let's see. Yes. Okay. But even better, probably from the waist. Let's see here. Okay, yeah, it's fundamentally
the same thing. Yeah, so you can
control from the waist, from the hips, you
know, like those. Okay, so and then yeah, okay, now it's fully confirmed. So the rigging
happened successfully, and at this point, as you can guess, all you
need to do is to animate the position of each
of these, right? And you can change the rotation of every single
one of these guys. So if I change the
hips, obviously, it will rotate the
entire body now. Oof. Right? So you can create
infinite combinations, every pose you desire with
full consistency, you know? That's why creating characters this way will always be
better than just AI. You know, AI will always
have some limitations. So if you learn this, you will definitely find a job
eventually in the industry, you know, and this will
be a very precious skill. If this is specifically what you need to do in After
effats, well, I invite you to study projects that have been
done there, you know, for example, this amazing
channel that I invite you to do to study because they create their videos with Adobe doctors, you know, including After
facts, as well, you know? And in case, let me show
you a couple extra things. **** is not just rigging. You can also do
simple animations. So let me go back
into the plugin. There is one entire area that has a walking and
running system, right? So let me show you very quickly. We go here, we select all the pieces, all
the controllers, simply, and then we go to this button here,
the play button. Okay? And then you see
what we have here. These are all things
I invite you to try. But one of those is the
walk one cycle, alright? So we just click
this button and it will automatically
create a new button at the top of everything. It's wait a couple
of seconds because it's probably
processing some stuff. So now we have a new layer here, which is the controller
for the walk cycle. So for now, you know what? Let me just play to show you
what is happening. Woof. The character became alive, so now it's trying
to walk, right? It's a little stubborn
because we didn't calibrate the initial position
of the layers, right? So the position needs to make sense for this
cycle to work. So simply calibrate the position of the arms of the body, right? Usually, a standing position
is the perfect case. So you need to place
your character initially in a
standing position, you know, still like this, not a T pose, still like this. And the cool thing, though, is that when you click here, the controller, let's
make a different color. When you click the controller
of the walk cycle, you have these amazing values that can animate
dynamically over time. One of those is simply
the walk versus run. So if you do 0%, obviously, it's not
going to do anything. It's going to now it's just
to our still position. So you can return
to this for now, so we can create a
better initial pose. Okay? And then we
increase this again. We can also make
the character run. You know, there are so many
things that you can do. Okay, the cool thing
is they can morph, hoof between running and
walking dynamically, right? So let me do a little loop so we don't take forever to render. You can see this. Alright, so you see what is
happening, right? And you can choose every
single aspect of the movement. So you can go here,
you can choose the width of the step, right? You can kind of tweak also
the angle of the feet. So if you click on the
feet, specifically, so in this case,
legs, as you can see, if we have a foot roll, so you can tweak the actual angle of the feet when they walk,
you know, like this. So every single aspect can
be tweaked very easily, like this, you know? And, yeah, that's one of the most amazing
things about this. You just need to explore this.
That's my recommendation. You can tweak the
speed, the inertia, the swing, angle, you know, every single thing.
You know, now I see. You see what is happening.
Now it's crazy. Yeah. And it's just a matter
of tweaking this, right? You calibrate it fundamentally
for your specific case. And then you can find
it with some effects. You can apply our best friend on the glow and also some
motion blur like this, okay? Let's give some time to render. Yeah, so that's it. You know, so remember, this is a
very powerful system. So I think this is ready
enough to get you going. I want you to explore it by yourself because
that's the way you are gonna go you're gonna have the most
fun, you know? So but, yeah, I hope
that you enjoy this, you know, because, in case
you need to go even deeper, just look for tutorials, go even deeper into
this plug in, you know, but I just wanted to
make you aware of this plugin and its
basic routines, alright?
34. Expressions (Intro): Alright, let's talk about
expressions for a second. Fundamentally is a internal
scripting language in signed aftereffects
that allows you to use less keyframes or none, but still create movement. So fundamentally, create movement without
using keyframes. The second purpose
of expressions is to create special relationships
between parameters.
35. Mirroring vs Parenting (expressions): One of the first uses of
expressions is mirror values. So let's create a
second shape here and let's shift it somewhere
else a little bit like this. Alright, so we have shape two, shape one here. Okay. So suppose I want you to link
the horizontal movement of shape so that it mirrors
shape one, alright? So we pull up the
positions of both here. And as you can see, we have the cable available through
every single property, right? So what we're going
to do is should go to shape position and make a cable exposition
of shape one, and watch carefully what happens when it release
the mouse. Boom. You see what happens?
First of all, the color of the number of position and
shape two changed, which means that now there
is an expression going on. Let's double click to
check what is happening. There is a full text here, which just indicates
the reference to the exposition of shape one. So right now we are brutally applying the same number of
shape one X to shape two. That's the reason why it
instantly jumped there. In this link, it's
working in real time. It means that if we
animate shape one, like this, you see what happens, Shape two will also
follow along like this. Because every single
as you can see, we accelerate the number
being updated in real time. So it's always mirroring,
referencing to this, and it's fundamentally
overriding the internal value of that
originally this head. So we can still see that
value by clicking here. So you see originally
has still 326, but it's been overridden
by the expression, right? And we can bypass
this at any time by pressing this button here. Boom. So as you can see, as soon
as I disable the expression, it will easily return to the
fixed value that normally has or eventually
some key frames. So if we have keyframes here that even go to
the opposite way, for example, it will just
read those instead, you know. But if I just enable this expression that
just tells it, Okay, actually, the value
that you need to have is the one from shape one. In that case, it will
completely ignore these keyframes
and also the value that we give here manually. It will just refer to shape one. And that's why it just works
as the word parenthood. But be careful because this is different compared
to the parenting from here, from the cable here. So let me show you
the difference. Fundamentally, when you parent, let me just go back like this. Okay? So, right now, we have these two layers
like this, alright? No expressions, nothing, okay? So the difference with the
full parenting is that if we parent shape two to shape one with the
standard parenting, so from here, First of all, as you can see, doesn't jump. Immediately, doesn't go
to the same X value. That's the first difference. And what we just told Shape two to do is to
follow shape one, starting from this
relative position. So it means that now we can still control shape
two independently. So for example, this can
still go somewhere else. We can control this
value as we want, so it's not busy. And it will only move if we move the parent,
which is shape one. So if we move shape one, yes, it will follow along, but it will never jump
as with expression. So that's the
fundamental difference. Parenting keeps the
relative position intact. Whereas if you brutally mirror the value
from the expression, it will actually synchronize
instantly to that value. So even though these two things may sound or look similar, they still have
fundamental differences. So be careful, so that
way you recognize when to use the parenting versus
when to use the mirroring.
36. Value & Slider (expressions): Oh. Okay, so now,
let's just step back. We just deleted the shape two. We are just with one, and another expression that you
need to know is called value. So let's enable an
expression here, again, Alt click or option click
and just type value here. Okay? So value means the internal value
that we input here, you know, or the
keyframes that we use. So just read itself
fundamentally. So it means that right now,
it's as if the expression is not enabled because if we change the value manually,
it will still work. If we have keyframes,
it will still work. All everything is good. Let me show you. Look, this,
you know, still working. Alright. So, but now we can apply a relative
offset to this value. For example, we can apply
a operational change, so we can do minus, divide it,
multiply wherever we want. So, for example, let's
go simple, plus, okay? And if I just click
outside, right now, we will just break because the expression is incomplete, right? So we just specify where to get the offset
value from, right? So this can be
everything you want. You can just go here. After the plus and then continue the expression
just with a cable. You say, Okay, get
the number from here from something
else can be from here, from an effect value.
Where you want. But to make it make
sense the most, let's actually create a slider. That's how it's called slider, which just represents a number. We apply it on the
layer itself here. And now, as you can see, this by default just doesn't
do anything. It's just a number. And let's
call this offset, X offset. Okay? And then we go here
back to the expression, continue it and cable. Shoot the slider here. Now, just close the
expression, click here. Boom. Alright, so right now, it's still just working fine, is reading exactly these
values on the keyframes. But if we change
something on the upset, if we increase it, poof.
You see what is happening? It's applying a extra
increase on the value, right? And obviously, this is dynamic, so you can animate the
offset dynamically. So now we have a
new keyframe here. So that you can do extra
change of movement. So you have your
normal exposition, plus the offset in
case, you know? And that's obviously
a simplified version of this to make you just
understand how it works, right? So there are infinite
uses for this. So you can do very
complex relationships just for this simple principle, value and operation that are references to
something else.
37. Auto-opposite (expressions): Okay, then another case where you may benefit
from expressions is that you need a layer to do the opposite of
another layer, right? So we already saw
this case earlier. The most frequent cases where you have a layer
that does the opacity, you know, goes 0-100, right? And you need another
layer to do the opposite. So in that case, you have a second layer,
obviously, here. And instead of copying, pasting the keyframes here, you reverse those,
why simply just applying a simple expression on the opacity of the other layer? I'm just going to
place this number to a random number just to show
you that it doesn't matter. So you just enable
an expression. And if you know the peak
value of the other layer, so we know that the peak is 100. We need to go 100-0
in this case, on the shaped, right? So we simply type the maximum, which is 100 minus
this guy, right? Boom. In that way, when
this is on zero, this is on 100 and
the other way around. So they will automatically do
the opposite of each other. You will do a cross
fade automatically. And that way, you can just control it from a
single layer, right? So then you just do this deposit here,
just reverse again. And that way, you
have a cross fade of two layers
automatically, right? So this obviously is valid
for anything you want. As long as you know the peak
value, this trick will work.
38. Wiggle (expressions): Two let's go back to shape one. So another expression
that you will need a lot sometimes
is the wiggle. So wiggle is an expression
that allows you to randomize a value at a certain speed
and at a certain amount. So for example, we can randomize the position of this square to
make it more live. So in this case, let's
compress X and Y to one single parameter by right clicking and disable
separate in this case. Okay, so now even
though we can apply one independent wiggle
if they were separated. So right now we just compress those for simplicity, right? So now we just apply a single
expression on position, and we tie wiggle. And inside these brackets, we need to type two numbers. One is the frequency, which goes from zero
to the frame rate. Frame rate is the max
number you can get to this. So in this case,
just go to five, which means five times the
second. And then coma. And here, the maximum
number of pixels that the square can
diverge its position two. So let's try with 100 pixels. Okay, and close, and now
it's already working. Let's enable the motion blur.
You see what is happening? Now it's randomly going to 100 pixels max from the
starting position, you know. And obviously, if I
increase this number, the first frequency, it will become faster, right, or slower. So but as you can see right
now these numbers are fixed, and secas, we can make
this become dynamic. So we go to the slider here
and we can call it frequency. Make a second slider
called amplitude, and then we can just connect the two numbers like
this to the sliders. So this is going to
be the frequency, and this is going to
be the amplitude here. And now we just give obviously we give numbers that
make more sense. For example, ten and
25. And that's it. Now we can control frequency and amplitude dynamically
because these have keyframes here, right? So we can make, for example, a shake system with this.
That's the very classic case. So you go here and you start from zero amplitude,
obviously, right? Then you instantly go to a higher value,
for example, 150, and then just decrease back to maybe more relaxed
setting maybe ten, and also the frequency,
we can go to one and make
everything a bisier. Let's see what
happens. So we have the just no shake and the boom and then it returns to be a more relaxed setting. So let's give it two pixels
will be fine. Yeah, you know? And so that's one use of this, you know,
and as you can guess, you can apply a wiggle to every single property
independently. So be careful with rotation because rotation
works with degrees. So we need we don't need to
have those high numbers, so maybe we can do
two and let's say, one degree max like this, yeah, or maybe like this 51. Okay. You see that now, we have the slight shake also
on the rotation. Let's make it a little higher, maybe two degrees Max. You see that now
is more apparent. So all these things can be
controlled dynamically, you know, so it's your choice. The cool thing is that
wiggle doesn't bypass the key frames or the
internal the initial value. So you can still decide the
initial value from here and also use keyframes while it
still wiggles the position, you know, the current
position, so it will just keep working.
39. Index (expressions): There are some cases
where you need to create some kind of grid, some kind of custom design where we have multiple shapes
of the same type, and every single shape should be slightly shifted like
this progressively, right? So instead of doing
this manually, which will never be as
perfect as you want, why not using expression
to make it perfect? So we keep only the
first one for now, then we go to position and
we enable an expression, and we type value, which means right now
this number that we see here plus plus 400. So that's our shift in
pixel to do for every copy, multiply it by index. So index is fundamentally
this number here. Every layer has a
unique identifier, which is called index. It's this number
here. So right now, all of this is going to be
multiplied by one, right? And as you can guess, all you need to do now is to
shift this guy to the left. And then duplicate this layer. So let me shape one. Okay, so if I
duplicate this layer, Wow. You see what is happening? As I duplicate, obviously, the index, it increases
automatically, two, three, four, five, and consequently, every
layer is going to be perfectly 400 pixels to the
right of the previous one. That's it. Oh, trivia. You can notice that
even though you can notice that layers with
the same name can coexist. The reason is that
actually this is not the true name of the layer. It's just a temporary
local name. To see the true
name of the layer, you can just click
on layer name, and we'll show you the
source name, right? So remember this. I
will never rename the actual file in case you
import files from the drive. So when you rename a file
inside the Fafaxs safe. It will just be a temporary name that
it's only in our forfex.
40. Rescaling (expressions): Another very useful
expression is called linear. So fundamentally, being able to rescale a range of values
into another range of values. So a practical use
of this is when you need to make something
audio reactive. Let's make a clean, new shape. Okay. And we start to apply some kind of
fake mouth very quickly. So let's use split. Okay,
split very quickly. Okay, so this effect when
you increase this value, it will make the
mouth talk, right? Okay, so we know that
this value here needs to go from zero to
maybe around 30. Alright? Okay, that's our
first range that we need. Okay, and also we
need that this is the property that we need to apply the
expression to, right? So let's actually enable
an expression here. Okay. Then, we need
to drag in a sound. So I just dragged
in this audio here. Let's check out the
waveform. Let's hear it. Hey, Annia D gum. Okay. Hey, Annia gum. All right. So now
infaxs a way it should convert a audio
waveform into keyframes. Alright? So that way, we can then send those
keyframes to a property so we can make it react when you
hear the audio, you know? Alright, so do that
when you should right, click on the sound,
first of all. And then assistant and then
convert audio to keyframes. Boom. Alright, so now we
have a new layer up here. We can call it envelope. And if we check out the keyframes, we can see that we have
three streams of keyframes. We can delete left
and right channels. We don't need those. We only need both
channels keyframes. All right? Let's see what is happening if we see
the graph editor. As you can see, we have a S of keyframes are perfectly
replicates the waveform, right? And these values
have defined values. So let's see the range. So they go around from zero, obviously, to around ten,
maybe nine. All right. Okay. After we assess what
is the range of the input, this envelope, we start
to go back to our split. Property that we
need to animate, and we start to type
our expressions. So we go here and
we type linear. Then we define where the
values are coming from. So while the cursor is blinking, we connect it to the source. So both channels here. Okay. And then we need to
type in the source range. So the source ranges 0-9, if you remember, Alright, so 09. Then we define the
destination range. So we can now rescale the numbers between 09 to
whatever range we desire. All right, so we just
continue to type coma. We need to do this
0-30, remember, right? So okay, and then just
close the expression cts. Now what is happening is that
only when the audio talks, we should see this
guy talking, listen. Hey, Arno. An. Just like that. Hey, A. That's it. And a
little curiosity, if you want to
change the linear to a more curved interpolation, you can do it just by
typing ease. Boom. Okay? That's it. All right. So that's a type of case where you need to
re scale ranges of values. And obviously, you can do this. You can do this forever
where else you want. This is just only
one of the uses.
41. Time (expressions): Sometimes you just want to make a value automatically
alive over time. Another way to do
it instead of using wiggle is use time.
So let me show you. We have a simple shape again
and we want for example, this object to rotate over time like this, we can use time, so we enable an
expression and we type time multiplied by
a certain number. So fundamentally the
higher this number is, the more faster it
will go over time. So let's try with 20 per now. Okay. So now every second
is going to rotate by 20 degrees infinitely until the end of the timeline,
you know, forever, right? We can also invert
the rs by multiplying by -20. Like this. Okay? And this expression is
extremely useful, for example, on noise based properties that have evolution
as a properties. One of those is
called fractal noise. Okay, which generates
a taxor right? And as I told you,
evolution here, which is a degree value, right, which moves the taxor
always in a different way. So to create perpetual change on a taxor that has a evolution, and I can grant you that
many plugins have this. You can just enable
time here, multiply by, for example, 100, now it
will be always different. Let's disable the
rotation. Okay, boom. So now, it will always
keep changing, right? Like this automatically, right? So you need to use key frames just enable one of these,
and it will be fine. Alright? Too quickly display
the expressions on a layer, like similarly to how you do with you to show the
keyframes on a layer, you can double tap
E very quickly. So if you press E, it
would just show like this. But if you press E twice, quick, it will automatically show the actual
existing expressions. Alright. And yeah,
so remember this.
42. IF Conditions (expressions): Another behavior that
you may need is to make something happen only when certain conditions
are met, right? So we can set that
up in After effect. So we create again in our
simple shape again, like this. And we say we want the rotation. The rotation should be, we need the time to be applied, not all the time, but only when a certain condition
is met, you know? So let's make it go a little
faster exactly, alright? Okay? So let's
suppose to do this, we can use whatever we want
as a condition, right? But for simplicity, let's
go with a check box. So we install a check box on
this layer, okay? Alright. Check box is fundamentally simply a number that
goes from zero or one, you know, so it's kind of
a true or false statement. Let me show you if I enable. We can also animate
this dynamically, as you can see, right? So let's actually
do it. Let's go. Initially, it's going to be off, and then it's going to be on. Alright? Okay. So
let's go Lectus. Yeah. This is called hold keyframe. It's just a static
value, Lectus. I kind of holds that
value instantly. Alright. So now let's set up the F
condition on the rotation. So we go here and we type
make a little more space. Sorry. Alright, so we type brackets. If the check box is
zero is one, actually. And then square bracket is
going to be the result. So in San square bracket, we're going to type
the time, right? So if the check box
is equal to one, just apply the time
automation, right? Otherwise, so we type. Otherwise, just use
the internal value, fixed value, value like this. So that's a basic
simple if condition, you know, so and then we
just close it, right? And let me show you
what is happening. Now, as you can see, it's not moving, which makes
sense because this is zero. Still, it's closed.
But as soon as we encounter the true
statement, Wow. You see what's happening?
Let me show you. Boom. Right? That's it. So with this system, you can make certain
behaviors and expressions trigger only when certain
conditions are met, right? This is a very powerful
thing that you can use to build scripts and in general
in certain projects, you know, to avoid using
too many connections, integrate everything inside
the big if statement.
43. Presets (expressions): When you make projects, you
may realize that you use the same exact thing
multiple times every time. So to avoid losing time to
do the same thing manually, why not creating presets? So presets are a very
powerful feature of Effex because you can selectively save specific
things of a layer, so you can reuse those
later on another layer. Alright, so let me show you
how to create a preset. So let's make our
best friend again, our standard shape here. Okay. So suppose that I reuse exactly the same time the
glow, you know, every time. So let me make the presets. Always this glow with the
same setting like this. So I want you to make it become the starting
point, you know? So, okay, select the thing that you want to
save and the preset. So in this case, just the glow. Like this, boom. Okay. And then just go to animation, save and just give a name. So in this case,
default glow Okay. And from now on, he when you make a new layer, so let's suppose that we have
a fresh layer like this, maybe with a different shape, like this, completely
different, you know, Okay? If we maybe even we want to apply this setting
a different type of layer. So maybe an adjustment layer. Like this. Okay? So we go to effects
and presets and we search our preset,
so default glow. Alright, and then just press
Enter. Woof, that's it. So it will automatically call our settings to that layer
that we selected. That's it. Right? And the presets can also save where
we want can also include key frames and expressions or only one of
these things in the preset. So you're completely free. So you can also save
multiple effects like this where you want. The preset will include it. The most extreme case
is where you want to save the entire
settings of that layer. So it will really call that layer and make
it look like that. So for example, we apply
the default glow, okay? Alright. So suppose
won't you be able to call this exact
shape as well, you know, with these
exact settings. To do that, you need to
expand the full settings, and it will always show
you transform and effects. So for this need, you need to select both
effects and transform. And let's also, you know what? Let's also do some key frames
here, here, right, okay? And also, let's
enable motion blur. Let's just enable a
couple of things. And also maybe a way goal here. Way goal. Okay, maybe
ten and 20. Okay. Alright, so we have pretty much all the stuff
happening at this point. So we select both transform
and effects like this. And then we just save
it as a preset, save. Okay, let's call this
random bar, like this. So now we just this
layer completely. Okay, so at this point, we just go create a new layer. Complete fresh layer, so
let's call this fresh. Okay. And if we drag this into the layer or just double
click on random bar, Wow. You see what happens?
It will literally apply the full settings
or our preset. You know, I will literally
just apply the key frames, the expression is there,
the glow is here, you know, all the
effects are there. And that's how powerful
the presets are. So please use presets to reduce redundancy in
your projects. Alright?
44. Good To Knows (expressions): What I want to show you
is slightly unusual, but sometimes you need this. I call it long distance
cables, right? So let's make it duplicate
of this composition. So we have this main one and
then duplicate it main two. Alright. So they have
exactly the same content. Okay, so now, let's suppose that we want the square in main one to mirror the value of the rotation in
main two here, okay? So how do we do it? Because they are they are completely in
different compositions. You know, they are
somewhere else, right? But there's a way to do it. If you undock the composition
Mint here side by side. Now we can see two time
lens at once, right? And you probably guess
what we're about to do. We can simply connect just
the square in Main one, all the way to the one
in Mint like this. Boom. You see what
is happening now? Let's just go to not only
just give a value to, the one in Mint, okay? So we just rotated this. So now if we go back to main one, in theory, we should see the
same thing, the same square. Woof. You see what's
happening? That's it. So this is I call it long
distance connections, you know, because it just makes a longer
reference fundamentally. But just to make
you aware that you can also do this in Aftereffex. The only limitation is that sometimes the
long distance cables may not respond in real
time changes, right? So they may only snapshot
the first value, right? So but just be aware of this. But I think it's very
important to be aware of this possibility
because sometimes it may be resolving a big issue. I mean, so far, I just
brutally use always the cable to connect something
to something else, you know, like this, and
it will brutally just compile the reference to
the entire value, right? But the truth is that to keep clean to keep some
expressions cleaner, especially when you
use more complex, longer expressions, you
need to use variables. So variables are very simple. You just define
those at the very beginning of the expression. So here, we can just type wherever number or
void, where we want. So it can be anything you want. So maybe in this
case, for simplicity, let's just say X. X equals. And then we just compiled. So we say, Okay, X is
this the check box. Okay? And then type
a semi column. And then for now one, you can reference to this just using X in the rest
of the expressions. So here we can just delete all this thing here
and just type X. So if X equals one, etc et cetera. So let's see. Just works. Alright?
So that's the way you keep everything
cleaner, more organized. Alright, so we will
reprise talking about expressions again
later in this course. But just to make
you aware, here, if you press this button, it will make you access all the presets to do what you want on any
property, you know. But the truth is that
the ones I showed you are the most likely to use
in a project, you know. But remember that there is an entire universe of scripting that you
can access to, right? It depends on how deeply you
know the language, you know. And in case you don't
know it deeply, you can just use Chat
TPT as a coder partner, you know, to build custom behavior expressions
to what you need. You know, remember that's
a very powerful tool that can help you
at anytime, right?
45. RGB Split: Alright, so let's have
some fun with glitch art. One of the fundamental
techniques of this type of art is
RGB split, right? So we need to have control over the three separate
channels of an element, reds, greens and blues, right? So to do that, we
need to first of all, precompose one custom
image of our choice, in this case, our
favorite controller ever. Okay? So now it's already
precompose, okay, here, and we are going
to disable it, actually. Boom. Okay. So now we simply
create one solid Okay, let's call this one. And here, we're going
to apply one effect called set channels here. So set channels is pretty interesting because it
allows the solid to pull to fetch any of the four channels RGB
and Alpha from another. Layer, you know? So we have the four channels source here, ready to go. So we're
going to use that. So in the solid here,
we simply say, Okay, get the red from the source, the greens from there, and
also the blues from there, and also the Alpha from there. Alright, so now, even
though source is disabled, now we still have
the controller. You know, that's it. So this is already one
interesting use of this effect. Lost mirror completely the
look of another layer. Even when that
layer, for example, has so many effects, right, this can be an
efficient way to duplicate, fundamentally, an
entire chain of layers, you know, duplicate
the look of that, not just the layers, right? It's a way more efficient
solution. Okay, that's said. Let's now return to our purpose. We need to have control over
the three separate channels. So simply before doing that, let's set up also
the shake system. So let's make also
a null object here, and we can control this
call this control, and we add a nice slider
to this called amplitude. Okay? So we're going to use fixed frequency and just
control the amplitude. And let's make a
simple animation. For now, yeah, I'll just
enable one key from there. That's fine. Okay, so now on the position of of the solid, we're going to merge. We're going to merge
the two dimensions. So merge, and here we
enable simple wiggle, you know, but it's going to have ten frequency and
dynamic amplitude. So the amplitude is going to be gathered from the ampitude slider on the control
layer, right? Here. Okay. So now,
obviously it's zero. It's just still, right? Okay. So now let's apply the same
thing to the rotation. So copy expression only, and based on the rotation. And we just need to
maybe reduce the number received because probably
we're going to go to a higher number,
you know, like 25. So for a rotation,
we need to probably divide that number by at least, yeah, kind of 75%. So multiply the number 0.20 0.1 to reduce it
automatically, right? Now it's ready to go. So let's start you
making automation shake. So we go one, three,
four, five frames, and we go to 35, and then 25 frames
later back to zero. Okay, so far so good. And let's enable motion
blur everywhere. Okay. And let's polish
these keyframes, so they are a little more
interesting. So let's go here. Let's make it come impulse
animation. Boom, okay? Maybe also more drastic.
Let's say, boom. Alright. Okay, okay. So,
so far so good, right? Nothing weird, but now
it's time for the magic. So let's go to our solid, and let's just disengage one of the three
channels simply. So we actually we can already actually duplicate
duplicate this three times. Okay? So we now
have three copies. So on the one above earthing, let's just disengage the
first the first box here. So there's the first channel, right right now is gathering the red channel from the source. Just disengage that. None.
Okay, now, select the next. Let's just double check what is happening. You see
what is happening. Now here, layer two is fetching
all the reds from source. That's what is going on, right? So let's call this
layer reds now, right? Red. Okay? Let's go
to the next one, and let's disengage
the next checkbox. So this one here,
the second one. Alright, now we have the greens. And then let's enable this, the third one and disengage the third channel, and
we have the blues. Woof. So let me show you
what is happening so far. Okay, it's still not done
because we need you now reassemble the three
channels together, right? So to do that, we use blending modes. So we go here and select
the three channels. Like this, and enable
dark darkened mode here. Ooh. You see what's happening?
Now, it looks normal. But actually, it's just an
illusion because now we have separate controls over
the three channels, right? And as you can guess, as
soon as we shake this image, we're gonna see the RGB split, which is Poof one of the fundamentals of
glitch art, right? And, yeah, so that's already a RGB split system, ready to go. So at any time, then you can duplicate these keyframes and shake it even more
drastically if you want. For example, the second shake can go a little crazier,
for example, like this. Let's see. One, two, right? What we want. Okay? So because the three
channels split temporally, but then as soon as they return to the exact same position, they obviously resemble the
original untouched image.
46. Glitch Displace: Alright, so now let's level up, and let's make another
very famous glitch effect, which is the horizontal
displacement of an image, right? So to do that, we need to, first of all, chop
this sequence. So let's make it
a little shorter only 3 seconds, and
let's precompose. We just create a solid here. And we need to call this map. Okay? And then we recompose
the displacement map already. So move all inside it like this. Okay? And let's go inside here. Okay? So here, we need to use a turbulent noise or you
can use a fractal noise. Fractal noise or turbulent noise are pretty equivalent to facts. You know, it's turbulent
noise in this case. Okay. And all we need
to do is to make the evolution alive
automatically using some time. We
already know this. So time you go here, just type time,
multiply it by 4,000. Okay? And then we simply change noise type
to block. Okay? And then increase the
contrast a little bit, and then maybe
change the algorithm to something else, maybe smeary. This doesn't matter, for now. So let me show you
we need to wait to see what is happening
on the outside, right? So let's put this let's
return to the surface for a second here and
let's lock the window. And then let's go back
inside the map composition. And then we can put these two window Sab
aside like this, right? So now here, the one on the
left is the final result. The one on the right
is just map source. So on the source level, we don't need to see the map. So we can disable it like this. So we need to see the
actual controller, right? Okay. So let's actually
go back to the map here. So you see what's
happening, we start to see the source map on the right
and the result on the left. Okay, so here we can search
to design our noise. But before we do that, let's
go back to the controller, and we search to use a
displacement Map effect, right? So we apply it on the
controller layer, and we link it to
the map, obviously. Okay? And we're going to use luminance as a source
for the modulation. Let's increase the
horizontal displacement to at least 200. Yeah. You see what is happening. And so this technique is not just for the classic
horizontal displacement, but also for granular
displacement or for disintegration
of a layer, right? So the uses are
pretty much infinite. Okay, maybe also slide. Very cool displacement
never hurts. Okay? Alright, so let's return
to the map design here. And so you see here, we have all the
settings for the map. So, let's first of all,
make it a little bigger. So we should transform
here and scale up. Like this. You see
what's happening? So that way, we make the
chunks bigger, right? So there's still, even
though we scale up, it's still, as you can see, too detailed, you know,
it's too detailed. So to reduce the
level of detail, you can either reduce the complexity value which
is here, right, reduce it. It will exactly do that. It will kind of it's called
posterization, fundamentally. You know, see what is happening, becomes simpler, alright? So for this specific effect, we need this to be simpler. So let's keep it around
too, probably, yeah. Okay? And what else? This is also changing
already dynamically, right? You see how it's breaking apart. The image. Woof, right? You see where we're
going with this, okay? And what else? You know what? Let's keep the settings of the noise now visible
at all times, okay? So we can design it even
without this window open. So after the major
part major settings are decided here at this level, we can cart close this window. We don't need to see it
anymore, so let's close it. So let's return to
one window mode, and we see just the
final result here. But we still have
the noise settings ready to go, thanks
to the lock here. And then let's start you making
automation on the amount. Of the displacement, right? So when you go
here displacement, displacement amount,
and let's automate it. Let's make it kind of jump instantly, choose
specific values. So let's make the star
from zero, right? 00. So does our shake initially? Let's lock these keyframes, as well, toggle hold. Okay? So does our initial
shake, right? And then it explodes
to fall displacement. Let's also keep these
keyframes still. Okay, just for a
few frames will be about 20 frames and
then jump back to 00. So boom, boom. Okay. And let's see
what we have so far. Okay? One che. Okay? Maybe a little shorter. Let's keep this pretty
compact. Let's see. One. Okay. All right, so we're going somewhere now. And at this point, let's
start some motion blur, so we put an ice
adjustment layer after everything, we go here. Let's go where the
displacement is happening. So we go here. Let's use
a force motion blur. Okay, and let's see. Alright. So now
to finalize this, we need to maximize
the flexibility of the glitch map, you know? So I'm going to now add
a couple more things to then just have fun regulating the exact type
of regulation that we want. So let's first of
all, fix one issue. Right now, the change
of the texture is just almost uncontrollable. So we can quantize we can
quantize the change of this value here by adding one line here called
posterize time. So if you put that number here, it will update the value only that number of times
a second, right? Right now, for example, it's just one fixed
value every second. Okay. So let's make that more interesting by adding a
slider, first of all. Here, we just call it
pose postise frequency. Okay, and we link it
obviously to this. Okay. Second thing,
as you can see, we have this line here, this cross the whole time. So in order to move
this thing away, we simply need to add a wiggle to the position of the noise. Let me show you. You
see what is happening? So it means that we can
add simply a wiggle. Should this guy
here, wiggle, ten, and maybe 500 and then
pasteurize this as well. So just comp and paste the same line here,
shoot this guy. Boom. Okay. And now we're controlling both things
from this slider right. And then why not
being able to stretch the text search to horizontal
or vertical, right? So to do that, let's see if
we can do it internally. Here, we can. We can simply uncheck uniform scaling, okay? And that way, we can brutally stretch the text search to be only horizontal
or vertical. Even if this slider
gives us only 600, we actually go over that
number. Alright, that's good. Okay. We have an internal
control over that. And finally, let's add
an actual final effect called posterite again, but this is for the visual part. So this kind of decreases the resolution of the
shades of colors. Okay, so now with all
of this in place, we can simply lock
this thing here. Okay? And let's return
to our main composition. So the surface simply. And foo, you see
what is happening? Now we actually have just
full control on our glitch. So kay, you see
what's happening? Right now, it's just
one single glitch, you know, just fixed.
It's not changing. So because the posterization is, the frequency is literally zero, so let's make it a little higher maybe four. Okay, let's see. Alright. You see
what's happening? So it becomes kind of
a little more live. Let's double it. Alright. And at this point, we have
full control on the glitch. So you simply go where
the glitch is happening here and you can decrease
the posturization, the visual posterization of the map, for example,
that's one thing. Then if you pull this number up, it will make the texture
more horizontal, right? You see what's happening? Right? Or vertical by
doing the opposite. You know, this back to 100, this goes to 10,000. And now we will do mostly
vertical line glatch, right? You see what's happening? So, in this case, let's go over the classical
horizontal, okay? That's already start to
be interesting, okay? And then we can regulate
the width of the lines, obviously, by increasing
this, as well. So let's increase this, right? See what is happening.
Let me show you Okay. You know, so you can really
have now at this point, infinite combinations, right? And what else? And then remember that
you can simply change the type of algorithm
here to any texture, you know, and you have really crazy infinite
combinations. You know, it's
just never ending. This is very fun, you know, if you increase this, people will kind of
return more detailed. And but the point is
thanks to the system, we start to have a full
custom glitch system. And, you know, the funny thing is that there are paid plugins
that do the same thing, but this is really
a legit replica of a very expensive plugin. Oh, yeah, and finally, let me just show you a final trick. When you find the
settings that you want, for the texture, for the speed, for everything, you
can finally go at the very bottom of the
turbulent noise settings. And you have a very secret
feature here called seed. If you change the sd, it will just shuffle the whole noise. It will create straight away, a new combination of that noise. You know, that's it. So you can simply kind of play
it by lock, you know, just navigate the different
seat until you find one that does exactly
the thing that you want. You know, that's
another approach, you know, is completely
legit, you know, like this. So you see how crazy it is, and that's completely
free for you, alright?
47. Trail Methods: Creating a trail
effect where you have control over a line
form like this and morphing the align into
custom shapes and still tracing the align is a
very typical technique because you may use it in
so many different cases via effects on footage or motion
graphics, so many things. So there are
different approaches, and each one has advantages
and disadvantages. Okay, so let me show you
we create a simple solid. Here, stroke one. Okay? And we're going to
use paint for this method. So we go here on the
paint tool, brush tool, double click on the layer, and then make the
first line like this. Okay? And then enable paint
on transparent. Alright. And also, you will see the options for the
brush that you're using. You can smooth it, here, you can tweak what
we want the color. And we need to make sure
that we enable single frame. Even though if you need just to be just a fixed
frame, you can do it. But in this case, for
this, we need animation. So we need to enable single
frame mode like this. Okay. So now what happens is that starting
from this spot, I can simply go one
frame forward like this, and then you know, I can just continue
tracing the line, so I can just go from
here and I say, Okay, the next frame is going
to be like, you know, then again, from here,
next frame, continue. Next frame, Lectus. So this is a very classic technique for
music videos, for example. So if you have a footage
behind this layer, you will see the subject moving, and you can trace
every single moment of the character, right? And that's it. So now, obviously, when we play the timeline, so let's go back to composition, we see the animation. Poof. You see what is happening? So, you probably saw
this on social media, and that's the way they do this. Then we can customize the aesthetics of the line so we can kind of make
a little cooler. Well, that makes it look
one to times cooler, right? And okay, the second
method uses a shape layer. So we go to make a shape layer, first of all, like this, okay? And we create the custom shape. So produce, make sure
that fill is disabled. So fill has to be off. Okay? And you can
use either one of geometrical shapes or you
can do it manually again. So let's do this
manually like this. Okay. And you can either keep
the shape open or close it. You can do both.
So in this case, let's close it, okay? And the shape layer is always
divided in two categories. You have the contents here
that contain the shape, which is this one here, and the shape is made by two components, the
fill and the stroke. In this case, the fill is invisible, so it doesn't matter. But the stroke is this
line that we see, so we can expand it here. And we can customize the aesthetics of the
shape of the line. We can make it dashed like this. If it's needed, we can customize where we aspect the desire. And one of the cooler thing of this method is that we can
taper the line, right? Let me show you. You know, star should be cooler
like this, right? And then we can also
make it move over time. So to do that, we
need to enable one of the effects called
the trim paths. So we go here, click
Add Trim Paths here. Boom. And now we can control
that from these two values. So if we control start and
end to see what is happening. We can open and close the line from point A to point B, right? And we can also make
it shorter and make it cycle automatically with
offset value like this. You see what is
happening? So this is very useful for animations, loops, right, in
case you need to act to put something for
a video game or wherever. So let's automate this automatically with some
nice time like this. So now becomes live
automatically, right? And let's the timeline, okay? And the cool thing about
this method is that we can also morph the line dynamically to different
shape over time. So for example, we start
from this shape, okay. But then if we go to
the path property, we can simply enable
one key frame. Go, for example,
slightly forward here, and then you select the path
property only, like this. And when you see
the line is ready, so you hold shift
and click one of the points, and now it's ready. So we can morph it
into this new shape, for example, like this. Okay? And obviously, now it's gonna
morph between two shapes. While it keeps looping. You see what's
happening? That's it. And now you can morph this
to whatever shape you want. There's no limit, you know? We can also make toggle hold, so it will instantly
jump through these shapes with
one single layer. That's equal part,
right? You know, let's put a nice
adjustment layer as well. Like here, boom, boom, boom. So it looks already cooler. And let's suppose that you
have a custom image here. So we want you morph into
this image outline, right? So we can do it by simply precomposing the
image like this. We make the composition one
frame long for that image. That way, we can use a hyper optimized
bro scoping, right? So we can simply detect very easily the edges of the image. That's done, and then
we can simply freeze, and it's going to be
only that frame, right? So in that way we have already the image outlines
ready to go, right? And then we can simply select
the layer, go to layer, autotrace, current frame, and
let's put this back to two. Exactly. Let's make sure that
it detects only one mask. So let's increase the
blur a little bit. In case we're going
to remove that one. Okay, just confirm, we simply remove the
second mask, right? And then we just copy the mask path key frame
from the solid, so copy. At this point, we
can delete this and then just go back
to the shape here, go forward here and just
select path and paste. So you see what happened?
It converts the mask from the solid into
the one for the shape. And that's the way morph into more complex
shapes automatically. Who, you see what's happening? So that's it. And final bonus feature
is the fact that you can use multiple paths at the same time on the
same shape layer. So you can really use one shape layer in the end but use multiple
masks simultaneously. So you can return
here on the pen tool, make sure that you're not
on shape layer anymore. Make sure that you
select stroke here. And then with the pen
tool, you can simply draw a second shape like this. You know, and that will work in conjunction with wherever you
already do it, you know? And that will have its own
separate path property, and that can also morph. They all work simultaneously. Or you can make them work
sequentially, actually. So if that's your
thing, you need to simply scroll all the
way down to our effect, the trim paths here and simply change this feature from
simultaneously to sequentially. And obviously, it will
do a sequence instead. And finally, there's a third approach
which is kind of outdated, which is simply
creating a solid. You make a path, you know, simply you can close
it or keep it open, and then use this effect stroke, which kind of does
the same thing, but very more limited. You know, you have
very similar things. You don't have the ability
to taper the line. But in some specific cases, you may need to use this
instead of the shape layer. You know, that's the
point. You know, so let me show you, you know, so it's kind of a basic way
to do the a stroke line, you know, fundamentally,
but it's way more limited. So yeah, it's important that you're aware of
this anyway. Alright?
48. Looping Methods: Looping a piece of footage
can be very challenging, especially if there
is high movement, and there are so many
different cases. So the ideal case is
the case where you have synthesized completely
something in Afrofax and that way you have native control
over what you see. It means that we can,
for example, here, natively decrease the
movement of each of these eggs back to zero, right? Or we can make them go back to exactly the first decision that they have on the first frame
of the animation, right? So right now, obviously, they end up in a different pose, but since we have native control on each
egg, we can, in theory, control each one to kind of match the first frame of
the animation, you know? So that's the ideal case, but not always this happens. We deal with footage, right? The first method that you can attempt is boomerang effect. So you simply enable timely
mapping on the footage here, and you simply
just go backwards. So you copy the first
keyframe like this. Double check the keyframe
here is not empty, so we actually we need to
place one keyframe here, okay, and simply paste
later like this. And obviously, the video
now is going to go in reverse, starting
from that spot, but at least we are
100% sure that it will eventually go back
exactly to the first frame, and we'll close the
loop right now, right? Because the last
frame is matching. Exactly the first one. It means that now
we can just export this 15 second thing and then just let the file
play back repeatedly, right? So that's one thing. But again, the disadvantage of
this is that if there is high movement at
the end of the cycle, it may give away the fact
that it's going backwards. Even if you use
bezier like this, you can try to smooth this
with some bezier like this. Alright. You know, this is
slightly better, especially if you
use, for example, a wide time effect here. You can try to use this, you can interpolate the
neighbor keyframes, so we can link this to this. Okay? For example, around the spot where time
gets reversed, we can increase this
value slightly. So now it's going to
interpolate three frames, four then three backwards. We go back a little bit here, zero, three, and back to zero. Okay, let's see what happens.
Let's give it some time. So kind of add some
additional interpolation to not give away that easily the fact that it's
going backwards, you know? This is the first strategy
that I invite you to use in certain cases where you recognize that you can
solve it already this way, especially when the
movement is very slow. In all the other cases, we need to use a
different strategy, which is to actually go forward
with time, so let's make. This called Math one. Hey, let's do a step
back, and here, let's disable time remapping. Okay, we're back to square one. Let's disable also
the interpolator. So if you cannot use
the reverse strategy, you need to go
forward with time. So in that case, the second
technique is simply cross fading between the first part of the
video with the last part. Let me show you what I mean.
Okay? So we do a chop. And the first 2 seconds
of the video like this. Okay? And then we snap this smaller chop at the end of the video
and we put above it. And then we do a fade in. So we just make it go
from 0% to 100, right? And that's the fundamental
principle of this. Fundamentally, it will make the beginning of the
video fade in, right? And that way, poof, you know, the time goes just forward and it just continues to go, right? So, the only issue with this is that if there's
too much movement, you need to do a couple
more compensations. For example, let's
start to improve this by making the second
video, you know, this guy. Fade out at the same time. So we enable a nice
expression that we already saw at the
beginning of this course, 100 minus this opacity. So we also fade out. Okay?
That's really one thing. To make the fade smoother, we can use also a
camera lens blur here. Let me show you. Fundamentally, it is a cara
lens blur that has the radius linked to the opacity
of this layer, the below one, so
that it automatically increases the blur amount
when this fades out. Fundamentally, it's
a trick to fade out more professional
and smoother, right? Then over this, we can also tweak the position
of the layer above, so we animate rotation, scale, and position
all at the same time. We know that for
sure. At the end, needs to reach those keyframes because we need to fully
match the first frame, right? So we cannot we cannot move these
keyframes at this point, but we can move those
earlier than that point. So around here, you see that
there is a misalignment. We can compensate it
by simply rotating. Wow, you know, the
video a little bit. Like this, like this, you know, just a little bit, will be way more aligned. And then use Bzier as always. Let's use Bzier everywhere. Like this, boom. Let's see. Okay, we're better. So let's
give sometimes render. Okay. Let's make it even slower. And then we can also add a adjustment
layer on top of this. Okay, we use the technique I
show you at the beginning, the very beginning course, you know, in the morph
section of this course, we use this technique
already, if you remember. So we're going to
reuse that now. So Gaussian plus median. We're going to animate radius
and blurriness here only when only when the morph only
when the video is ending. So from here to
here, we increase blur, plus median, right? Like this. And it will do the opposite at the
beginning of the video. So copying paste, swap. Okay, and then nice
busier. Let's see. Okay, so it starts that way, and then it ends also. That way, okay?
Let's just play it in real time, full speed. Don't watch the timeline.
Let's just go back here. Okay, play. Alright, so this is already a more advanced fix and allows you should have the
video going just forward the whole time and have, in theory, a perfect
artificial loop. You know, that's the
point because it's not natively looping, obviously. But this is a way
to still have it in a smooth way, reasonably
smooth, you know? It's very important to watch the result in full resolution to see that actually there
are some flickers that are not actually there. And finally, the third method that you should
look something is used simply some kind of
distortion to hide the cut, you know, create another adjustment layer
above everything. We call this glitch and we use an effect called
glitch again. We can use the one
we built together in the previous exercise. But for now, for
simplicity, let's just use another plug in that does the same
thing, you know. Also, let's use
some more shambler and we're going to animate
simply the amount. Of glitch from this value
here, small glitches. All right, so we start from a higher amount,
obviously, like this. Okay. Let's change this to
simply black and white. Okay. We go to simply ten
frames after and it goes zero, and then obviously, you copy and paste these key
frames at the end of the loop. Here? Oh, even earlier
at this point, the danger of this metal is you can do whatever you
want at this point, so you can cut and restart the loop a little more freely because there
is no problem. You see what is
happening? At this point, it doesn't matter
because thanks to the distortion, hide the trick. You know, we can literally
restart the loop at any point, as long as it's long
enough, you know, at least four, 5
seconds, you know? So, so answer, so these are the three main concepts
for perfect looping. And each case shines on a different type
of footage, right? And finally, in as you can use even AI nowadays to
morph perfectly, between a problematic
first frame and last frame of
the loop, you know? So for example, if you
have this footage, start with this frame like this, you know, so you can see, I don't return in the same pose, I either reverse
the video, right, which is going to be maybe it's going to be
apparent, right? So the new way to do these things to loop
a video perfectly is to export the first frame and the last frame
of the loop, right? So in this case,
these two frames. And then simply go to this
website here. Let me show you. You go to Video
Generation image video, and you drag in the
first frame here. Make sure you select
clink 1.6 for this, and then you go to end frame. And then here in the prom,
just type something simple. So that describes
the subject and action that will be performed to morph between the
chew frames, you know? So in this case, simply
a guy slowly moving, you know, between
che poses, does it. You know, something simple. And then make the video as
short as possible. One output. Let's make a little
more relevance, convert your
creativity, maybe 0.65. So you can also scroll down here and just add something
that you don't want. So, for example, blurry,
shaky, distorted, deformed, Okay, and finally,
you simply generate. Okay, 3 minutes left, right? Just wait that finishes.
Alright, here's list. Woo. Alright, that's it. Right? So so that's the way you loop the video without
going backwards in time. We actually use the yin to
go forward in time, right? So at that point, we simply download the video
here, download. Once you have the video imported back in the timeline, obviously, it just needs you connected with the last frame of
your video, right? And at that point, it will
just allow you to naturally. Go back to the first
frame of the video. Like this, Woom goes
back, at that point, you just pack
everything together, and that's the perfect loop. That's it. You know,
boom. So one che one che without going backwards in
time actually goes forward.
49. Particles: Oh. So after effects as an entire
category of effects inside the generate and simulation
categories. Let me show you. You have an entire folder here, simulation versus
generate as well. So I invite you explore these two folders and just play with the prometers
of all these things. You have very much
fun with that, right? And you will find out so many cool things that
you can generate. But one of the most useful
ones is the particle systems. So we're going to
explore that today. So we need to generate
a new solid here. So let's call the particles. And we're going to use one
effect called Parole world. So we apply it on
the effect here. Wow. Here is. By default,
it looks like a fountain. So not much
interesting, alright? So let me show you
how to tweak this for any single thing
you desire, alright? So with this, you
can do dust, snow, impact, effects or fireworks. There are so many uses of this. And also, you can use a custom texture and use
fulm design the particles. It looks like your custom
symbol, where we want, right? So let's go slowly
through everything. First of all, the
bith rate obviously determines how many particles
there are gonna be. So the lower this is, the
last particles we have. Longevity is self expiratory,
so fundamentally, the bigger, the longer each particle is going to
persist on screen, you know? So for now, let's go to
2 seconds per particle. And then let's go to producer. So here we can choose the actual origin
point of the particle. So for example, if we need
to design some kind of rain, we put it above the screen, right, and we're going to see
the particles falling down. Or if you want the opposite, we put it below the
screen like this. And then on the physics, we simply reverse the
gravity, right? So they're going to be
pulled up like this, right? So it depends on what
we need. And then we have also the radius. The radius is very interesting
because it allows you to give some depth
to the whole thing. So if you increase the radius
Z, you see what happens? We start to have some depth. And this is also reactive
to virtual cameras. So if you increase the radius Z, then you can zoom in inside
this environment, right? So it's very cool and also
react to depth of field. Okay? That's it. Then we have obviously
these all the radiuses. So this is very important
in case you want to have a general turbulent
system or the dust, you know, so in this case, we're going to design
some kind of firework. So we need the adi to be pretty pretty close
like this, okay? And then we have velocity
here in the physics, which obviously
determines how fast the particles already get
generated and resistance, which is fundamentally
the opposite. Fundamentally, you can set up a system where the
particles start very fast, but then they kind of
slow down very quickly because of the viscosity
of the air around it. So you can achieve kind of a firework effect
this way, you know, and then you can change
the algorithm here, that determines how the
particles expand fundamentally. So everyone is
completely different. Let me show you. Jet sideways
is one of my favorites. Vartex know, there
are so many things, so many cases that makes this tool very
versatile, you know? So for now, let's stay
on fire like this, okay? And then the extra here adds extra randomization to every single
particle fundamentally. You see that. The
more I increase this, the more they will
kind of disperse. Just skip to the
particle design here. So here you can first of all, change the color of the
birth of each particle and also the death color so we
can do two colors, right? So that particle
has one color when appears and then gradually fade to the second
color, you know? Or you can change that
behavior from here. You know, we can have custom
color or whatever you want. And we can change the design
of the particle, right? So let's first of all,
zoom in a little bit. So we go here to the camera here and swig a little closer. Let me show you so
fundamentally here, we can change the
type of particle, and you see how many
options we have. So we have, for example,
motion polygon. That can be very
interesting. If we tweak it, we can make the
birth size bigger. That size is very small, like this, let's increase
the amount of particles. Okay. Change the type here. So there are so many
things that you can do. And then we show you we have also 43d elements
that we can choose. So if we do dra hedrin, every single one actually is going to be every
single particle. Now, is a three
element, you know? So we can make very big
particles like this that go actually also
very small like this. And then we need to just tweak the rotation values like this. So every single particle is going to be a little more alive. We can expand it. There are so many things
that we can do, you know? And for now, let's
go a little simple. Let's go back by back
through our simple lines. That's said. Now, let me
just skip to the fun part. We can automate the position
of this system, you know? So let me show you if I
do some key frames here, and I move it and I
move slowly forward LTs let me show you if I go
lower here and here. Okay? Automate the birth rate, I can open the particles
only when I need them, so I can go from zero birth
rate to a higher value, maybe even ten like this, and then back to zero. And let me show you
what is happening. Wow. You see what is happening? Now I start to have a directionality only
when I need it, right? So let me decrease this
value to maybe four. You know, so we can make very very custom systems
just with the stuff. And also, you can assign a custom texture
to the particles. So if we change the
particle type to textured, if we import a custom image and we precompose it like this. Okay, and we just prepare it, remove the background, where
you know these things, go to the particle system, and you assign the
texture layer. Obviously, we hide
the texture source. And what happens now
is that is gonna use that image for every
single particle, right? So that can be a cool
way to use that image, you know, if you have a
custom symbol, right? Okay, so for now, let's just
go back to the normal line. Okay. All right. Oh, yeah, now there's too much glow so
let's disable one of these. Now, we're going to go
a little more advanced. I'm going to show
you ways to make this system controllable a little more
automatically, right? Because as you can
see right now, we have to manually use the opening of the particles
versus the position, right? So what if we want you
to synchronize this to a custom trail that we
designed with a shape layer? Well, there's a way
to do it perfectly. So we just just step
back for a second, let's design the
trail very quickly. So at this point, where
do you know this? So I'm gonna skip some
parts. We go here. We call this trail.
Simple shape layer. We make the trail with a with a simple pentle
obviously, make those. While the path is selected, if you by accident
deselect it, don't worry. You can simply go
inside the contents. Here, you already
know it, shape one, stroke, select the path here. Alright? Boom, while
this is selected, just shift, click one of the points, and
you can adjust it. Alright? So for now, we have this nice path, okay? While the path is selected,
you can do stuff like this. You go go to window,
scroll all the way down, and there is one script that Afrofax gives
you by default, create nodes from paths. So we need to select this.
And opens this window here. Don't worry, press trace here. Boom. Okay. Obviously, we
need to select the path. Sorry. Alright. Alright, poof. So just slack the
path and click Trace. Then it closes weird window, and now let me show
you what happened. Now we have a new known object. Let's call this coordinates. Because also we have
two keyframes here that have a simple
percentage that goes 0-100. And as you can guess now, we have coordinates to go from that point to
that point, right? So first of all, let's disable the expression of this
percentage because by default, kind of loops infinitely, so we don't need that behavior in this case, so
let's disable this. Okay. And also, let's shift
these keyframes here. So for now, we go 0-100. Let's do it a little
quicker than that. Let's go from here to here. Alright. So let me
show you what is happening in the
coordinates of this layer. You see that they
are updating, right? So we're gonna use
that in a few seconds. First of all, and that's where I'm going to
show you one thing. There are two twin effects
for particle world. So when you don't
need this feature to sync it with a custom
trail, use Particle world. If you need that
specific feature, you need to use a slightly
different version of this effect called
particle systems two. So you just type here,
particle system here, and we drag here,
and now we disable. Actually, we delete the
first particles, right? Else because this one here has a slight kind of has
the same parameters, but some of those are
slightly different. So this version of the
effect has some limitations, which is why by default. I recommend using
particle world, but this has a specific feature for this purpose, which is here. Producer has a single parameter with the raw coordinates
for the coordinates here. So we need to have
this in this case. So in other words, we simply enable expression
on position here. Of the particle system, and then we link it to the raw coordinate of the
known object up here, right? So let's mix little space. So just a cable from here
all the way up here. Alright, so now in theory, we should see the particles. Foo. You see what's happening?
It's following exactly our path. Alright? Okay, we're going to
go back later there. So for now, let's ultimate
the design of the trail. We're going to smooth the line, so we go to the design settings, scroll all the way down, taper, let's give some directionality. So we're going
from here to here. So let's taper the tail. So we go here, you know? All right. Then
since the path here, if you remember, it's
just a percentage that goes 0-100, right? Go back to the shape
layer, add trim paths. And now we have the
two percentages to open and close
the line, right? Okay? And assume gas, we simply need to link to the key frames up there
here of the non object. And so far, now let me show you. Now the line is
opening automatically. Alright? Make start, also, do the same path
fundamentally to close it, you know, 0-100 as well, but slightly later, right? So to do that, we can
simply use a script, I'm going to send you
right now, alright? And this script allows you to delay delay some automation that is received from a
specific value, alright? So let me help you set this up. After you have
installed the script, after you have installed
this custom expression, the way I showed you already
in the expression section, you simply create
a new solid here. We're going to
delete this layer, so this is just to make
the expression work. The expression is called
delay delay automation. So select the dummy layer
and double click. All right. And now we just expose
the expression here, and we copy the raw text. And we're going to
paste that code into the start of the
trim paths here. So we go into expression
and paste, right, and then just go to the dummy, copy these sliders and paste
those on the shape layer. Alright, initial control
is simply delayed. So this value here is
the value in frames. Okay? So let's put it to up 24, which means a full second at this frame rate,
probably, yeah. Okay? So now we can simply go to this expression here and
we define the source. So the source is simply the end, so we simply link it. Let me mix space, right? So we simply highlight source. Okay, and we say, okay, get the value from end. And now delay it by the number of frames that we put here on delayed amount, right? So, in other words, now we should see the line opening and then
1 second later closing. Poof. Alright. So at this
point, it becomes very easy. We can simply
reduce this number, probably when you reduce
it to six. Alright, poof. That's it. You know? Fundamentally, in all this, we're using only two keyframes originally just coordinates. There are no keyframes on trail. It's completely automated. Alright. Alright, so let's
go back to the main quest. Now there is an
issue. First of all, we can let me show
you a very cool, very cool parameter that
I didn't show you yet. If we go back to the
physics settings here, let me reduce the birth
rate just for those. Oh, right. Okay. So right now, it's a set to be coal, right? But there's one more
thing that we can add to add realism
to this particle. So we can make it take account of the actual diversion
of direction, right? So to do that, we simply go
to the physics settings here, and we enable this value
here, inherit velocity. This takes account of the
change of the position, right? So let me show you if I
increase the hundred, Wow. So the particles kind
of echo that change. You know, let me
tweak it, alright? You see what is happening? So it kind of feels better, right? Now there's still one issue. After the trail has
completed its path, the birth rate is still open. Kind of remains there
hanging, right? So we need to kind of
manually close it. It's very boring. But actually, we're now going to do that. There is a expression
that automatically opens the birth rate only when the position
is moving, right? So this is called
automatic gate. So let me show you
because I'm going to send you now a second
custom expression, the unit to install
in order to use this. So let's simply create
a new dammi layer. Okay, and we apply
the new expression, which is called default gate. This is the name.
Alright? So, select the dummy layer and
double click. Alright. And we have a new slider here. So let's you paste
this sliner to the particles layer here
at the bottom, right. Okay? And then we go
back to the dummy. Let's copy this expression. Okay, paste, boom, boom. Alright. So let me show
you what is happening. Now, we don't see anything, just because we simply
initial click on now on this expression that we paste it inside birth rate, and we need to define obviously
the source of movement, which means that the reason
why we need to do this is that this expression is just not for the
particle systems. You can repurpose this
for anything you want. So in other words, you can make a parameter
go between two values. Only when another parameter
is changing, right? So when the modulator changes, you can go from value one
to value two instantly, and then back to value one when the modulator value is
not changing anymore. That's how this
expression works. It's not just for this
case. That's the point. So in this case, we need to read. We need to detect the change in the position of
the particle system. So we simply highlight
this specific line here. I left a comment for you. So
comments to make it clear. So let's highlight
this the source, and we get it from the particle
system position. Here. Fundamentally, when the
position is not changing, it will go to zero, so
it closes the particles. When it's changing,
it jumps to 100, but it multiplies that
number by this slider here. So we can tweak it a
little more easily, right? Without going through the text. Alright, so at this
point, we search working. Whoa. You see what's happening? So now we see the P holes opening only when we
are moving the trail, right? Isn't that great. And at that point, we can tweak how many power holes we want just by changing the
multiplier here, you know? So we tweak it to our liking. Maybe there are too
many power holes still. Let's reduce it, reduce it. Less. A little more.
Let's see. Alright. And I didn't tell
you the final trick. The final cool thing
about this is that we can still change the
path at any moment. So if we go back
to the path here, we can still select it. Let's go here. Let me
show you shape, path. We can still change
the path, and poof. You see what is
happening? The particles are completely linked
to all this, still. So it will still follow
whatever is happening. So it will kind of update in real time to whatever
new path, we give it. So you can really be
creative with this. Hey, so we can use Bzier for example, here
on the trajectory. Okay, we make it a
little slower, right? And then we make something like this,
kind of exponential. Let's see. Let's
say poof, right? You can design where
you want at this point, whatever trajectory you need
for your custom case, right? That's the point. That's
the beauty of this system. So it's completely
dynamic, and in the end, after all this is rigged, you can really control it simply from these two keyframes. So everything is controlled from the path and these two
keyframes. That's it. There are no other keyframes
involved in this system. It's all expression,
all scripting, right? And then we can just maximize
the aesthetics of this. At the end by applying a nice
standard motion bler top of everything like
this motion proller. Poof. So you can really
have fun with it.
50. Parallax: Right, so the parallax technique consists of creating
the illusion that some elements that are far from the camera are actually
also way, way, really big. So fundamentally,
when you move and you see the foreground
elements being very fast, at the same time, the
background moves very slow, you know, very similar to when
you are on a train, right? You see the cars next to the train and
moving extremely fast. But the mountains, you know, very far from the train
are still moving, but way slower, even though the train is
moving fast, right? So to recreate the same
principle in Affax, we need to really follow the
same principle in real life. So let's simulate a
couple of things. So let's first make a
background. Let's go here. Solid. Let's call this back. Okay, we're going to make a
simple grid for this here. Okay, we make it
three D immediately. Okay, let's and also, let's create a
foreground element. So we make a second solid Okay, and make it a little
smaller three D. Alright, so now foreground is the foreground element
is gonna be this guy, and the background is
gonna be that one there. Alright, so at this point, let's actually make
a three D camera. This trick works with both types of camera doesn't matter, okay? So right now, the
trick is not working because the two elements are
exactly in the same place, which is not what we want. So the trick is to simply move the background
away from the camera, so we push it all the way far, you know, all the
way, like 10,000 pixels a lot like this, right? So let me show you. Now
it's there behind that. And then we just
compensate the size of the background so that it fills the
screen again, right? So now it looks as if nothing changed, but
actually it did. So if I now move
the camera around, let me show you. You
see what's happening? We are now moving way faster for the foreground element than for the background
element, right? And that's the fundamental
principle of any parallax. At this point, you
just need to make the depth of field
more apparent, so we can simply use it.
The camera like this. Alright, increase, right?
So let me show you. Then we tweak the focused point. Let me show you like this. Alright, you see? So now we have our foreground elements
in focus, right? And the background is moving extremely slow,
you know, that's it. And then you can
expand the texture, can make it even
bigger like this, or you can tie it like
this with reptile. We already know this
at this point, right? And that will be fully
working, you know? And the more distant the
background element is, the slower will be, right? So that's the principle. So, in other words, simply
push the element far from the camera and
then compensate it by remaking it bigger. That's it. You know,
and now at this point, we can simply make this
position alive automatically, like some wiggle, and also
same thing on the rotation. For the foreground
element, we're moving way faster compared to
the background. And that's the fundamental principle of the
parallax, right? Then if you have Duik installed, you can enable this
behavior automatically for any layer just by calling Duik. Here. Let's bring it
back to square one. So transform to comp. Okay? Now it's
back normal, okay? But if we go to Duik and we
go to the camera section, there is an amazing
button called Z Link. So fundamentally, when you
select one or multiple layers, so in this case, let's do
it for both, like this. And we press this
button here, Zink, now all I need to do is
just push the back away. Okay? And you see
what's happening, the size will
automatically increase the more I push the layer
away, you know? So let me just
increase this number. You see that now it's
getting out of focus because obviously it's getting away from the camera.
Let me show you. So if we go to the
two view mode. Okay, now we are watching
the scene from the top. Okay, we are here, the
camera is here, right? So the background right now is let me see what it
is. Now it's here, okay? So, in theory, the more
I push the layer away, the bigger
automatically becomes. Let me show you. You
see what's happening? If I push it away also
automatically, it becomes bigger. So thanks to the ****
button, fundamentally, you can then simply
choose how slow that element will move in relation to everything
else flow control.
51. Ascii Technique: The fundamental principle
of ASCI art is to remap specific values of a footage or wherever you have in
cho specific symbols. As you know, every single
footage has dark colors, medium colors, and
bright colors. So if you remap, those specific three
ranges hoot specific opacity of symbols is going to create the ASCI
art in a dynamic way. So we're going to do
that. Alright, so let me show you how to do it. So we create, first
of all, a ten by 801080 by 1080 composition. This works with any
dimensar you want. But in this case,
for simplicity, I'm going to keep
it square, right? So let's do it. Let's
call this footage. So here's where we're going
to put our footage. Okay. So let's drag it into the composition and make sure that it fills the
screen completely. So we have this amazing
Jeff in this case. Let's go fit height. Don't worry about the
resolution because this is going to be
remapped completely, so it doesn't really
matter. Let's go here. Alright, we have our nice character doing a
simple movement. Okay. Oh, right. Okay, so this is the footage
I'm going to work with. Okay, so then we're going to
create a second composition. Okay. And here, we're going to choose the resolution
of the grid, right? In this case, let's make
it 36 by 36, right? Make sure that it has the same aspect ratio as
your footage composition. That's the only
condition of this. Alright? So in this case, let's make this
number because it's kind of a sweet spot for
good definition and speed. Okay? So let's call this
composition shape one, okay? And inside this comp, now we have we're going to choose the first symbol
fundamentally so. Let's create a simple square. So let's call this
simple square, make it slightly smaller
like this, and that's it. Alright, so now let's
make a third composition, 1080 by 1080 again. So the same resolution
as the footage, okay? And we're going to
call this main. So this is where all
the stuff is going to eventually happen
like this main. Okay, so inside main, now we're going to drag in footage and shape one together. Boom. Okay. All right. So now let's first of all, put footage in solo for now. Actually, okay, so now for now, let's put shape and solo, and we're going to
align it to the left, top left of the screen, right? Just like this.
Okay. So then we go to one effect called
rap tile here, and we're going to
tile it all the way to fill the screen to the
right and to the bottom, L. Very simple. As you can see, is perfectly synchronized with our footage with our composition, you know? And the reason is because
the math makes sense. Okay? So at this point, we just select the footage, like and we apply
one effect called Mosaic. And here's
the tricky part. We need to do a formula
to choose to create a perfect grid that matches
our shape one grid, alright? So to do that, we simply
divide the dimension. We divide the dimensions
of our original footage. So if you remember,
it's 1080, 1080, buy the dimensions of the grid of the
single shape, right? So we need to divide, in
this case, 1080 by 36. So we go here and
a divided by 36. And obviously, since
it's one, one, we simply do the same
through this, right? And now is doing a
perfect match grid, right? That's the trick. Okay, at this point, we start to hide finally the
footage forever, like this and simply track mat, shape one, chew the footage. Okay? And in the end, let's return to the footage
one last time here. We apply extract like this. And then here we're
going to choose the three ranges I
talked to you about. So for this symbol, for example, let's choose maybe
the dark colors. So we're gonna squeeze down this slider until we set you see F, you
see what's happening. We start to see the
character being resavled, you know? Let me
show you, right? Okay, let's do the aposte. Let's Archie actually invert it. Let me do a little loop here. Let's do the aparte.
Let's actually invert it. Okay, yeah, you see
what's happening? That's the straight
fundamentally. Now you see what is
happening, right? So we can now choose exactly to which range of the footage the
symbol will trigger. Or not simply. That's the fundamental
principle of all this. So okay, let's finalize the area where this
will be visible. Okay, that's going to be the
first range, and that's it. Okay. So at this point, all we need to do is to
duplicate these two layers. So footage and shape
one get duplicated. Let's give them a color. So duplicate, give them a different color so
we don't get confused. Now we're going to simply go to the project view and duplicate
shape natively here. So now we have shape
two and override it on shape one with all
to where you know this truck so that way, we can use a different
symbol for the second range. So let's first of all,
for example, rotate. We make it smaller. Let's
make it a kind of a d dot, yeah, something like this.
So it's a little tiny dot. Okay. And at this point,
we go back to Maine. Here and then simply map, shape, this footage now, right? Just link it to the new
copy of footage, right? So that way, we can use a different range for
that symbol, right? So if I move this Wo You
see what's happening? Now, obviously, the dots, the second symbol is
going to be triggered on a different range, right? So let's let's choose. Obviously, we need to
a different range. So let's for example,
go all the way here. Okay? So it's going
to do the opposite of the squares
fundamental, you know. Right? Okay, and what else? And then, let's do a
third final range. So we again, duplicate these
once more, put at the top. Link shape three with
footage three, okay? And also, let's
create a native shape three here and override
it on this guy, right? So we can choose a
different shape once more. So we got it up. Let's make kind of an
empty shape for this. We go, let's double click here. We use a simple mask, so we can quickly simply
expand, shrink the mask, right? Just like this. Y, then, let's go outside of here. Alright, and then obviously, let's change the different range again for footage
three. Let's see. Let's go kind of
intermediate place, yeah. So kind of Alright, something like this kind
of partially triggered. Okay, we can also smooth this a little bit
if it's required, if it looks good, right?
You see what's happening? Okay, let's smooth all
the ranges a little bit. Same for fog one. Okay. Let's keep
this very strict, like this. Okay? And
finally, you know what? Let's maximize the
aesthetics by using a nice classic glow at the top of everything.
So we go here. Boom. And let's see. And finally, let's add a
nice motion blur, as well. Like those Poof. Alright,
let me show you. Let's play through full quality. Poof. You see what is happening? So that's the fundamental
principle of this. So with this technique,
you can do what you want. You can choose as many ranges, as many symbols as you want. And the cool thing is, it generates a full quality
vectorial ASC matrix, right? So this is not our
footage anymore. It's just pure vectorial shapes. So it's full quality,
full resolution. And it can look very amazing, you know, especially
if you put this, you see what's happening. You can really achieve kind of retro retro style things, right, if you add nice grain like this add grain here to earthing. That can be one idea. There are so many uses for this that can, you know, that's completely
without plugins, right? That's a crazy thing.
So just have fun.
52. Clouds: All right, so to make
steam or clouds and affats only to make
a simple solid. This and then we're
going to use a mask. So we go here and let's make
a first ellipse like this, and then you feather it. So remember to increase
this guy, right? And then simply apply an
adjustment layer of bomb here with a tripling display. So you apply this
guy here, okay? And then we enable a evolution
time multiply by 100, so it starts to be
moving like this, right? And then we go to simply
tweak the complexity. To add detail versus the
amount versus the size. So the first size
needs to be smaller. And, you see what's happening? So if it's small enough, it will su to become more realistic. If you want it to
be more cartoony, use a bigger amount
like this, right? So this is where
you need to choose. So in this case, let's
go for realistic. So we keep a very granular
and detailed texture, right? And then we duplicate our post adjustment layer with a second one, a bomb earthing. The second one instead,
is going to have a bigger size distortion
and way lower amount. So let's keep around ten. Okay, maybe the lower five. Okay. And at this point, all you need to do is to choose
how many clouds you want. So you can go back here to the clouds and you can
change it safe, right? Or you can simply
duplicate this guy, so you have more if
you want a duplicate, if you want a longer
cloud, you can do it. Just simply duplicate it like those, right? Just like that. So that's the fundamental
principle to do custom clouds and so you can have fun,
right? I like those. That's it. One of the coolest effects
that you can do in fx is to scatter an image into pieces like if it
was in front of a glass, fundamentally, you know,
a glass shattering. Alright, to do that, we
need to first of all, create a dummy layer. So let's go here, solid. Let's call this dummy. Okay? And we need to apply one
effect called shatter, right? So we go here and first of all, change the mode to rendered. And then change the
type of shape to glass. Even though you can
use more weird shapes, let me show you what
is happening for you at the beginning, fourth. See what is happening. So one of the coolest way
to use this is to obviously use it as
a glass sharing. So we go here,
pattern, change glass. Okay. And right, no, now starts to look
cooler, right? So the first thing that
we need to be able to control is the moment where
the sharing happens, right? Otherwise, this becomes useless. So to do that, we
need to go to force one and animate one
keyframe on radius. Here. Alright, so we go here, and we need to now make
that value jump from zero so we can keep zero at the beginning of
the timeline, we hold it. And then it will just jump to that value that
already have by default. So when wherever we
place this keyframe, the sharing will happen. So let me show you.
So it's waiting, and then eventually it
will scatter there. Boom. Alright. So now
that we know this, let's go here and we'll regulate the exact detail
of the scattering. So, for example, if you want
more fragments or less, change repetitions
here, you know? So I can do less
scatter or more. Right? Let me show you. Obviously, this becomes super
detailed or less detailed. Then I can choose the depth of the fragment.
Let me show you. It kind of makes those
more three D or less, you know, there are
inferent combinations. And then if you like
the way it looks, but just want you reshuffle
the scattering pattern, simply move slightly one of these parameters
simply, you know, we can change their origin point or the direction
or the petition, slightly, you know, completely reshuffle earthing when
you move all these. Then let's call down here
to the physics here, and here we can choose
exactly what it says. So we can choose, for
example, no gravity. So the points will just
float frontally, right? Or we can change the
variance or the sizes. You know, you can
see how quickly I can make this become
unique for my needs. You know, randomness,
is constity makes those slow down after the
explosion like this, right? So you can become very cold very fast with this, you know? So let me just not too
much rotation, right? Okay. And so, yeah, so very simple to use. You can apply a
text so many things back layer, front layer. But let's suppose that
you should be able to apply this on a
custom image, right? So to do that, simply
precompose this guy. So as always, we precompose it. We, you know, leave
mode, share this. Okay? So now, if we go inside here, double click, we simply
disable the dummy, and finally we import one image. Here and let's stretch
it full screen. Alright, let's go outside. And who, it's really
working, obviously. Okay? And finally, let's
suppose that I want to be able to scatter a specific
area of the image, right? So there's a way
very precise way, which is to create
a solid here and we call this scatter
shatter mat, okay? And in this solid,
we're gonna make a simple circle mask like this, right, and then we're going
to feather it slightly. And we're going to
hide now the matting, and we go back to the
shatter effect here. And so when you scroll down, there is a gradient setting. At that point, all
you need to do is just simply place link the gradient layer to the shatter mat that
we just created, okay? Then you change these
to effects and masks, and finally place threshold
to 1%. Fundamentally, poof. I will read the white as the only point where the
shatter will you happen from, right? So let me show you. So if you go here
now, as you can see, only scatter from
that specific area. And you can change
that, you know, spot perfectly matching Wow. We can now scatter
all that area. You know, we can break
a specific point like this and obviously
reacts to the mass. So if you make it bigger,
it will be bigger, like this, you know, very
handy, you know, like this, so you can create very
interesting effects where you kind of destroy all the specific part of an image, you
know, like this. And at that point, you box the shatter settings here and make the depth a little bigger. In this case, for example, m3d, and just have fun with
everything else. Alright.
53. Glass Shatter: One of the coolest effects
that you can do in fact is to scatter
an image into pieces like if it was in
front of a glass, fundamentally, you know,
a glass shattering. Alright, to do that, we need to first of all, create
a dummy layer. So let's go here. Sold.
Let's call this dummy. Okay? And we need to apply one
effect called shatter, right? So we go here and first of all, change the mode to rendered. And then change the
type of shape to glass. Even though you can
use more weird shapes, let me show you what
is happening for you at the beginning, fourth. See what is happening. So one of the coolest way
to use this is to obviously use it as
a glass sharing. So we go here,
pattern, change glass. Okay. And right, no, now starts to look
cooler, right? So the first thing that
we need to be able to control is the moment where
the sharing happens, right? Otherwise, this becomes useless. So to do that, we
need to go to force one and animate one
keyframe on radius here. Alright, so we go here, and we need to now make
that value jump from zero so we can keep zero at the beginning of
the timeline, we hold it. And then it will just jump to down value that I
already have by default. So when wherever we
place this key frame, the sharing will happen.
So let me show you. So it's waiting, and then eventually it will
scatter there. Boom. Alright. So now
that we know this, let's go here and we'll regulate the exact detail
of the scattering. So for example, if you want
more fragments or less, change repetitions
here, you know? So I can do less
scatter or more. Right? Let me show you. Obviously, this becomes super
detailed or less detailed. Then I can choose the depth of the fragment.
Let me show you. It kind of makes those
more three D or less, you know, there are
inference combinations. And then if you like
the way it looks, but just want you reshuffle
the scattering pattern, simply move slightly one of these parameters
simply, you know, we can change their origin point or the direction
or their petition, slightly, you know, completely reshuffle earthing when
you move all these. Then let's call down here
to the physics here, and here we can choose
exactly what it's at. So we can choose, for
example, no gravity. So the points will just
float frontally, right? Or we can change the
variance or the sizes. You know, you can
see how quickly I can make this become
unique for my needs. You know, randomness,
is constit makes those slow down after the
explosion like this, right? So it can become very cold
very fast with this, you know? So let me just not use
too much rotation, right? Okay. And so, yeah, so very simple to use. You can apply a
text so many things back layer, front layer. But let's suppose
that should able to apply this on a
custom image, right? So to do that, simply
precompose this guy. So as always, we precompose it. We know leave mode,
shatter this. Okay? So now, if we go inside here, double click, we simply
disable the dummy, and finally we import one image. Here and let's stretch
it full screen. Alright, let's go outside. And who, it's really
working, obviously. Okay? And finally, let's suppose
that you want to be able to scatter a specific
area of the image, right? So there's a way
very precise way, which is to create a solid here, and we call this scatter
shatter mat, okay? And in this solid,
we're gonna make a simple circle mask like this, right, and then we're going
to feather it slightly. And we're going to
hide now the matting, and we go back to the
shatter effect here. And so we need to scroll down, there is a gradient setting. At that point, all
you need to do is just simply place link the gradient layer to the shatter map that
we just created, okay? Then you change these
to effects and masks, and finally place
threshold to 1%. Fundamentally, of, I
will read the white as the only point where the shatter will
leach you happen from, right? So let me show you. So if you go here
now, as you can see, will only scatter from
that specific area. And you can change
that, you know, pull perfectly matching Wow. We can now scatter
all that area. You know, we can break
a specific point like this and obviously
reacts to the mass. So if you make it bigger,
it will be bigger, like this, you know, very
handy, you know, like this, you can create very interesting
effects where you kind of destroy all the specific part of an image, you
know, like this. And at that point, you can
go the shatter settings here and make the
depth a little bigger. In this case, for
example, more three D, and just have fun with
everything else. Alright.
54. Grading (Simple Routine): A very common routine
that you may use when you deal with
footage is the following. They just import this footage
into timeline, right? So first of all, if the footage has been recorded with some distortion lenses, unless you want to
keep that on purpose, we need to reduce
or eliminate that. So let's go to compensation. Here and this allows you
remove the distortion. So let's just increase
this reverse diagonal, and let's just straighten
the lines back a little bit. You see that now
we're recovering the actual lines of the reality. You know, it's not disordered
anymore that much, right? All right. Okay. So after that, a sanded routine for color correction is used,
first of all, a filter. So uenary color
is where it's at. I recommend using
this because it's kind of a mini effect
that contains everything you need for color correction. You know, it's very versatile. The most important ones are
the basic section here. In case came in case
you use log footage, you need to input your camera specific
input filter here, right? In this case, I
didn't shoot log, so I didn't choose this filter. And you can do the first
tweaks or right from here, you can make the footage kind
of warmer, colder, right? But before even
that, I recommend already choosing a
main creative filter. So you go here, creative, and you choose from one
of the available filters. You see? So there are so
many, and fundamentally, there are cold ones
in this area here, and then there are
warm ones down here. And then black and white
ones here. All right. So and then you can also
install filters for Apha fat. So these are called
lots, L, UT, right. You can find those online or YouTubers share those on times. You know, there
are so many ways. That's set one of my favorite
ones is blue intent. So if you apply this, Okay. And the filter fundamentally
remaps the colors to a specific out color. You
know, that's what it does. And as you can see it
does many things at once. It makes it instantly more
cinematic, a little better. The colors become more vivid and use the
intensity of the filter. So you can use more less, right? So in this case, I like
to use a lot of this, so it just increases
a little more. Is one, 50%. From here, then go back to the basic correction and just start to tweak the footage
in case it's needed. Only if it's needed. You
don't need to do it always. So maybe just a little bit of contrast here or maybe
the temperature, we can make it a little
more warm or cold, right? Those are the main
choices. And let's return. Like this. Let me show
you a couple more things. Curves I never use
kind of dangerous. Color weirs also never use it. It's kind of these two
things are very dangerous. Secondary color cushion
can be interesting because lows you shoot target specific
color with a key here, you sample a specific color, for example, the sky. So we go here, click on the sky. So this is similar
to the key light. Fundamentally, samples a color, and then you can check
what is detecting here, and then you refine
or change the color, all of that detected area. So if you go here
to temperature, for example, you see
what's happening. So it's kind of allowing us
to change in weird ways, only specific colors, but use this only if necessary, right? So the quickest way
you already finish the color correction
is simply going here, apply a filter, and maybe
took the basic correction, you know, the ID footage
should be already clean. That's the point.
You should learn how to shoot proper footage, and that way, you need to do simple basic
corrections here. And then just leave
it for the fun part. That's the point. All right. So at that point, we can
apply a classic vignette. That's a very important effect that every single project has, especially if it's cinematic. This allows you to kind of smooth the edges of
the screen like this. You kind of
concentrate the focus at the center of the screen, right, which is where
the action happens. So you can smooth these
two values like this, you know, very important stuff. And then between the
lumetri and the vignette, we can apply a grain effect. So we go here, just
search ad grain, and then apply just
apply it between. Between the lumentr
and the vignette here. Alright, and then
just switch final, choose one of the line random
presets, doesn't matter. And then just hoot the
intensity of this. So this makes add some particular some imperfections
to the frogen purpose. So let's control it
from this amount here. So let's maybe 0.4. Right? And so that makes
it subtly more detailed. That's the point, you know? And then, very
important on any piece of footage that is cinematic, you need to use the letterbox. Letterbox are those black
bars that movies have. To simulate that in an
easy way, you can do this. Create a solid above
everything here. Let's call this letterbox. Okay. And then here on this solid, we use one fact
called jars here. And then simply just reduce height and
width is 0% like this. And then just Wow, you see what's happening just tweak the percentage from here. You know, you can
tweak how narrow the letterbox can be, you know? So there's no standard choice. You can ly do your costume
letterbox, you know? It's a completely
artistic choice. And this can also function as a way to compensate one thing. If you remember at
the beginning, we use the ans compensation effect, right, which is very important to remove
the fish eye, right? So if we scale the
footage down, okay? And we enable resize here, max two X Okay? And we put compensation
after the lumetr. Wow, we see fundamentally
some information that we are losing right now because of the lans
compensation, you know? So there is a way to record it, which is to simply scale
the footage as much as possible until the horizontal
borders touch the screen. So we squeeze it,
okay, around here. Alright. Okay, so that's one way to preserve as much
information as possible. But as you can see, there's also distortion here happening on
the top and bottom, right? So to hide that, we can simply use the letterbox. So we keep the
letterbox at the top, and we increase the percentage until we can cover the
artifact completely, you know, around here.
And that's it, right? So that way, we preserve the widest point of view
on the footage, right? While also straightening
lines back to normal, right? And also, we use the letter
box that we would have used anyway to keep the footage more cinematic, right?
That's the point. And, yeah, that's that's one of the tricks
that I always use. And that way, even if
you use a single camera, you can make your
footage very cinematic. Let me show you, you know, that can look
beautiful, you know, to tell a story, you know, like those and that's suit. In case the car
space that you're working with is not standard, you can change it
anytime by going here on the A BPC, right? And then just change
it from here. Assume color Gamma
here, right? Here. Okay. And, okay. And then in case you
have HDR footage, you can manage the dynamic range tweaking from here from
this specific effect. So apply commander HDR compender to the to the footage
that you're working with. So this allows you
tweak the highlights, right, bring those
back in range, right? So if your footage is HDR, not compressed, this effect will shine with that
type of footage. And finally, if you
need to analyze the colors of your footage
in a more scientific way. Well, in that case, go to
window and then scopes here, Domtar scopes, and then you can keep this here
close to your footage. And fundamentally, doesn't
matter where you will go. I will show you the balance
of the RGB channels or more. So the one I recommend the most, the beginner friendly one
is here, right click. Parade. Parade is the most easy to understand.
Let me say all this. So because it kind of splits
the three channels RGB, and you can see the amplitude of each from in the visible
spectrum from the matter, you know, so wherever
frame you are on, it will show you the representation
of that broken down. So you can see you can play
around with the facts here on the Lumtri you understand how every single thing here
influences the graph. You know, you can squeeze, see what's happening,
for example, here. Or the contrast, obviously
is going to change the difference between the
highest and lowest points. So if I compress, you
see what's happening? Now it's expanding.
Now it's gonna compress all the colors,
you know, like this. You see what's happening?
And, yeah, yada, yada. So just play around
with these and see carefully what
is happening here. You can see that there
is a slight prevalence of blue compared to the
rest in this footage. And so, but the truth is that after you develop your
own eye for your footage, you will not need these
anymore damach, you know? So you will just close those or keep this, you know,
close the whole time. And because you will
develop your own eye. It's a little bit like
playing your instrument. You know, you will kind
of develop your own ear, and then you can just
play without sheets.
55. Video Outpaint: One of the most
powerful tools that you can use is video out painting. Fundamentally, you
start with the footage that is cropped for some reason, and then you can simply recreate contextual content
around it to kind of expand the frame back
of the footage, right? So you have more information created with AI in
a contextual way. It's crazy, right?
So this video was generated starting just
from this simple close up. And as you can see,
you can go back to create information that was not there in the
first place, right? And so this is called
video out painting. The way to use it is to
use I'm going to send you this tutorial here is
a tool called runaway. So you follow this tutorial
and you can use it. So you're probably
wondering which cases this makes sense use. Well, if you remember, when we did the
stabilization exercise here, that's one case where
this makes sense, right? Because, for example,
when you use heavy stabilization
for time remapping, you may lose a lot of information around the frame
in order to stabilize it. So if you go to the cube here, and if you remember right now, we're forced to zoom
in a lot, right? And we are using some rudimentary method to
tile the footage, right? We simply using tiling. But what if we can simply create actual information
around the frame? That way, there will
be no artifacts, and we can avoid using reptile like this and
simply keep it more zoomed out while staying super zoomed in, super
stabilized, right? That's the point. So that's
one application of this. So you fundamentally import
just the video like this, you know, we just keep
it crop like this. You export this video and
then expand it, right? And then you reimport it here, and then you just
go out of here. And that way, you
can simply zoom out more and keep the civilization intact. You know,
that's the point.
56. Audio Reactive Effects: To make audio reactive stuff, in our refactor, you
have three main ways. Alright? So let's just import
something that has audio. First of all, we have
already here this amazing MP four,
which contains audio. Let me show you here, boom, boom, boom.
Alright, let's see. Here's the volume of those Look. Okay, so let's just choose
the work area to analyze. First of all, let's start from the beginning to this
point here. All right? That should be enough.
Okay, so then let's just right click on the file that contains
the audio, right? So let's click
keyframe assistant, and then convert
audio to keyframes. Boom. And this creates a
new layer that contains the keyframes that represent this audio waveform
fundamentally, right? So after we have that, we can simply delete left
and right channel sliders. We don't need those. Only
both channels is important. So let's call this envelop. Okay. And let's keep
this visible for now. And now let's start
to create the element that we want to animate. So we make a nice Shape layer, for example, can be anything. But in this case, let's stay simple. Let's
call this shape one. Okay. And we're going to make a simple wiggle on its position. So let's go here, create
a nice wig. We call six. And then instead of
a fixed amplitude, we're going to
link the amplitude to the actual envelope
of the audio, right? So that way in
more silent parts, is going to be controlled
at the center, and then it's going to
diverge violently from the center only when there are peaks in the audio, you know? Okay, so to do that, we
need to first assess what is the range of the
key frames of the envelope. So we simply go here on
the slider, graph editor. As you can see it
goes between zero and around 15, maybe 20. Yeah, around 15, yeah. Okay, so zero, 15. Let's remember
remember twos numbers. And then we type
just A. Alright. And here, we're going
to define what A is. So A equals both channels. Okay, let's first
type linear because we're going to rescale now
the numbers of both channels, which originally go
0-15 to a range, let's say, 1-500 pixels. Alright? Okay. So now fundamental
in more silent parts, it's gonna be very
controlled, very still. Let's make the shape layer very small now so we can
see what it's doing, okay? We enable mochbre. Let's let me show you
what is happening. Let's see if that's
working already. Ooh. You see what's
happening? So, only when there is a high
peak in the audio, the shape layer becomes
crazier, right? Only when there is a peak. And then returns at decentra, right? You see what's happening? Okay. So now, such you make
the layer a little cooler. So we apply our default glow. We already know the
spoomEls that's it. Okay? And so that's one thing. Let's apply the same type
of wiggle to the rotation. So it becomes more
interesting Wiggle here, copy expression, paste to the rotation, change slightly the wiggle
rotation to something else. So let's start from 0.1, and that should
be it. Let's see. All right. Okay, so that's
fundamentally the first way to use audio reactivity
in Afterefact. Alright? Let's go
with the second way. So we use we make a
copy of shape one, so we have Shape two now. So on shape two, instead, we're going to use we're gonna disable for now the wiggles. So, boom, boom. So it got stays there, so
we see what's happening. And we apply one effect called
waveform, audio waveform. And we apply it before the glow. Okay, so let's zoom in. And so as you can see by
default makes a line like this. So we can link now this effect to the
source of the audio. So in this case,
beatbox MP four. So not anymore, the keyframes, but the actual audio
file straight. Boom Alright, so let me
show you what's happening. As you can see, it's ready. Reacting. Let's make the
lines a little higher. Alright, you see
what's happening Poof. Let's change the color. Okay. Just white white for now. And now, why not telling this waveform to be exactly on the
outlines of this layer? We can do it by using a mask. So since this is ready a square, we can simply make a
rectangle tool here. Double click. Now, the mask is ready there in
the right place, and then we can link
it to the path. You know, path mask one. Woof. That's it.
Right. That's it. That's it. Okay, let's tweak slightly the
aesthetics of this. Let's increase maybe the lines. You can have fun
with these stings, experiment a little bit. And that's already
the second way to make audio reactivity. So let's actually re
enable the wiggles. So it also has its own
path, right? Strike. And finally, shape three is simply the same type of
effect, but different. So on shape three, we
lead audio waveform, and we apply instead spectrum. Here, it's called spectrum, it works the same
way as waveform. So you simply link it to
the beatbox file, okay? And simply for now, again, disable the wiggles to
see what it's doing. Okay? And we use the path which at this point
is ready, ready to go. We need to increase the
max height of the lines, so we'll never see
those. Exactly. Change the colors, change
the start and frequency. This depends on the audio input. So in this case, probably we
need to decrease the range 100-1 thousand. Exactly, right? So this effect normally use
for or reactors, right? Okay, and let's make it thick a little
higher poof, right? Okay, let's keep it
this way for now. And let's re enable finally
our amazing wiggles. And finally, as in gas, we are going to put
everything back together. Let's see the three squares now with their own
independent wiggles, but they are all connected to the amplitude of the beat box to diverge from the century. Let's also make them
interact with maybe with some nice landing mode.
Alright, let's see. Wo That's Sith. Have fun.
57. Audio effects???! (literally): You may be surprised
to know that Aftereffx has some audio effects
notation, actually. So changing the volume is just not the only
thing you can do, actually, you know,
in theory, yeah, you can do some simple
fades like this, you know, go from minus 192 to any number to do
fade outs, fade ins. But actually, if
you go to effXs, you have an entire category of effts that you can do here. So I'm going to show you
the most useful ones. So, for example, applying
a reverb can be useful. So if we do this. You
see what happens? So it's simulating a
room for this audio. To preview audio instantly
without rendering, you can just press the
dot on the number pad. In case you don't
have number pad, you need to disable all the eyes on all the layers to have the
same behavior like this. So disable all the eyes. So we'll just press space bar. And just previous
the audio, you know. Okay, and then let me show
you the settings of this. So okay, play with the reverb. You can make a longer reverb, select this, increase
these numbers. Okay. And then the
most important thing is dry versus wet. Fundamentally, dry is the
original unprocessed audio. Wet is the opposite. So by doing two keyframes, one of this and one of this,
you can do a cross fade, which in zero effect to 100% effect or just a
little bit of it, right? So that's one thing. And then let me show you a
couple more things. For example, slander
can be very important on specific edits,
specifically anime edits. This is a very used effect
on when you want to decrease the intensity of
the audio just for a second, and then return to Max power, right? Let me show
you what this does. You hear what it's doing? So it's kind of
modulating the sound and make slightly
weaker in a way, right? But that's the trick. You cannot increase this value dynamically. So let me do an example. We go here, we go and sync, for example, one of the beats. Okay, let's say
that I want to do this. Okay, for example, here. So here we increase the dry. We actually decrease the
dry from 100 to maybe 25%. At the same time, the
wet does the opposite. So let's just set this up
to do exactly the opposite. W 100 minus the dry simply to so we can use too many
keyframes. Let's see. And then you just
return to full power. So actually, Zu dus 0% and
then return to 100% dry later. Like this, and then we use
some nice viziers like this. Let me show you. Okay. And this. Something like this. Let
me see what happens. Right? So it's kind of
an important detail that you need to implement in some specific type of edits. You know. This
works wonders with drops of music. You know,
it can be very fun. And then what else? We have the delay is
kind of a reverb, so in the end, you
don't need to use it. And then all these effects
are kind of more niche. Okay, maybe sometimes it can be useful to use this instead. High, low pass this allows you to make the audio
muffled on purpose. So we can start from completely
open spectrum, so 20,000. Then we make a nice keyframe
on this parameter here, and then we can close
the spectrum gradually. Until we do, for example, 500 and will become muffled, didn't you change,
obviously to low pass. Okay. Let's see. You see what happens?
So now here then becomes muffled right
here, it's opens. And then You see that
becomes manfold here. I'm going to give you one
preset that allows you to fade out any audio
in a smooth way, right? So just install this expression. And the way it works
is you simply drag in the just precomposed
the audio track, right? And then let me show you. We just apply the expresion
It's called fade out. Default fade out here,
right? So double click. And you will see some
keyframes here up here. And now, fundamentally,
wherever you place these keyframes will make the
audio end there, you know? So let me show you and
we'll kind of make it do in a smooth way. No. So let me show you. Alright. Woof. You
see what happened? So instead of just stopping with a dry fade out or
stopping suddenly, it's gonna going
into some kind of roomy thing and then fade out in the
smoothest way possible, you know, let me
just play again. You know, and then the
audio actually finishes. I don't recommend
going to advance with audio and Afrofax because it's very
basic in the end. You know, we have very
it's kind of clunky. So this is as far as
you can go, you know? So if you need more
advanced treatment, go in a specific softer or
at least go in premiere. Premieres already better. Or if you're a producer, just do it separately
in C based or FL Studio or Ableton
Amazing Softers. Use those if you are
more advanced or B?
58. VR + 3D Elements + Lights: After fax is able to
create VR videos, so let me show you how to do it. We simply go here
composition and VR and create VR
environment, first of all. All right, let's choose
the resolution two k, 24 frames per second and
create new VR mind master, and we're going to use at node camera plus a three D null. And also, we're going to use three D logins and then create. Alright. For now, let's close this thing. Okay,
this all looks like. Fundamentally, it's a
simple square composition, and the camera represents where we are going to be
into the video initially. So we can animate that camera. And also in the video,
we're going to be able to orbit around the
objects, you know? Okay, so let me show you. For example, let's make
a simple three D solid. Let's make a grid. Let's
become the ground, right? So, three D, then we go here. Let's generate the simple grids. We already know this stuff. Alright. Just a walk look. And so we start to rotate the
ground layer by 90 degrees. Let's go here 90. Alright. And then we start
to move our point of view, we need to use this
controller here. Alright? So as in gas, if we change the
vertical position, Woof. See what is happening?
Now we are here, right? Okay, so let me
show you if I start your orbit using the Y rotation, obviously, we're going to
look around ourselves, right? So let's go actually more at
the center of the ground. So we go forward a
little bit, like this. So now we are more centered. Like this. Let me
do shark. Exactly. At this point, let's start to tile the ground. So we go here. Let's use reptile. We know this, expand like this, 3,000
pixels in all directions. Now we have more
room to move around. Why do you see what's happening?
Okay, now we're here. And what else? Let's keep it this
way for now, okay? And let's start to collocate three D elements into
this environment, right? So let's make nice text
simply for now, like this. Let's make a phrase A. Hey, let's just pull this text, three D. So, as you can see, the second view is very important to know
where we are, right? So this is at the ground. We are Let me see we are here, right? The text is here. Okay. Let me just raise it. I think I just should be
probably Yo, here, here is. Yeah. So just look around
to see where it is, right? Okay, so let's just choose
this point of view, okay. Let's just let's go here. You can also create
multiple cameras. So choose different
point of views. So for now, let's
just go by eye. Let's go here. Let's return
to one view mode, okay? Yeah, so let's go here. All right, choose Let's just choose where this guy is
gonna be. So let's go. Change the position,
twit it tweak. Okay, a little smaller
should be fine. Alright, rotate. Make a simple animation
on the camera, right? So let's suppose that we're going to go
from this position going forward all the
way to somewhere else. So we move like this right, so we scroll scroll, scroll. We move into this environment somewhere else,
right? Like this. Okay, so let's go
back here, okay? Let's duplicate the text
layer, and we push. We change the word. Let's go. Let's go here, solo. Double click. Okay. Hey, how. Alright. And we push. Obviously, this guy
somewhere else. Let's go here, scroll. You know, Legos how? And then we go forward. The camera is already
scrolling, right? So we do play this layer
again, L, push again. Okay, we go forward. How are You. Okay, and finally, so we have full control on
every single element in our point of
view, also, right? So let's go. Let me just
create the final layer here. How are you? Let's go back here. Alright here's. Okay, let's see how
much we can push it, so we arrive here probably
we can push it a little more right exactly
until here. Okay. Okay. And then finally, supply a nice glow at
the top of everything. So adjustment layer, post, make it three D and apply
one effect called VR glow. So you have also VR effects. So this glow is
specifically made for VR. Let me show you. You
see what's happening. So it makes triggers
the glow like this. And finally, we can also use some light in this
environment, right? So let's create a
couple of lights. So right click. Light. There are different
types of light. So let's go with
spotlight first. Spotlight can be very
confusing to understand. Fundamentally, it's
like a camera. A two node camera has position and also the
point of interest. So one trick to not
get confused is to simply use the
shift parent trick, so you snap it that way
exactly to the target. So shift parent to
the first word here, and then just raise
it up like this and then tweak the position. So let's see what is
happening where here. Let's see. Let's go over here,
change the anchor point. Alright, see, thank
you to the snapping. We can choose exactly. You know, which area
we're gonna lit. So let's go here. Alright.
So the word is there, the light is there,
so we need to rotate it for sure.
Let's go like this. Okay, and also tweak it or
should be a little higher. Exactly. Okay, increase the intensity of
the line for sure. So let's go here, increase,
increase, increase. Alright, here's. So we just push it in front
of the Oh, here's. Alright. Okay. So then you just tweak the cone settings of the line, have fun
with these things. Elects. Let me show you. Alright let's make it very
narrow. Ex ante Electus. Okay, and then you can let me place a second type of
light on the next word. For example, you can
use also a point light, which is fundamentally
a one node light. It's kind of a skere. So it's
a make it harder already, and we give it directly the
position of the second word. So we go a little forward here and we give it the
position of the how word. Alright, see what's happening. Now the light is
exactly on this word. So we can place it easily in a slightly different
place like this. Okay, let's change
the intensity. Maybe we should push it slightly before exactly like this. Or after, you know,
that's your choice. Okay, let's put it
somewhere here. Okay. Then let's
go forward again. And finally, let's make a third the type of
light called ambient. So the ambient says
does what it says. You know, we choose
the base level of light for the
entire environment. So let's go with ambient. Let's choose 100%.
Let me show you. You know, obviously, this defines the purpose
of odd light. So let's decrease the intensity
of the ambient light. So let's choose
different percentage. Let's go here, maybe
10%. Let's see. Alright. Okay, let's choose,
something like this. And finally, let's place this
at the top of everything. Let's place a nice
vignette. Okay. Okay, so then you start to design everything you want
here in this environment. And finally, when you're
ready to ask for it, we need to simply
go back to select that composition that we used
so far VR Master reveal. While this is selected, you go to composition, VR, create, and then this time instead
of the first option, we need to select
the second option. Create VR from current
composition, right? Make sure that our
composition is selected and simply make sure copy settingc this in case you use especially
three D plugins, you may need to keep this
and this enabled, this case, we don't need this
and simply generate and if it brings you to this weird view, it
means that I worked. So and then we can actually add one additional glow
on top of everything. At this point, it's kind
of converted back to a two D 16 by nine frame. So let's just blast the
colors a little bit. So it just becomes visible.
Just a quick test, obviously. But just to show you the gist of how this works,
let's go here. Okay. And finally, what else? Yeah. Okay, so at this point, we export this whole thing. To export VR video
from Aprovx you need to use at least to have
media encoder installed. Alright? Make sure they have
media encoder installed. So then we send
this render queue, but we invoke media encoder, so click this button. Alright? Okay. So once you're here, you simply go to the settings. So you can choose P
four where we want. So in this case, let's go
with MP four standard. And in the settings, remember to enable one
specific chat box. Go all the way down here, and there is one chat box that will fix everything for us. You know, here's BR
video, check this on. Alright, this will inject a specific meta data for YouTube you detect
this as a shatic video, and it will make
this interactive. So let me show you. So this
point we can just export. Let's call this four
and export export. Alright, let's just
jump back to YouTube. Once the video is completely
loaded and elaborated, you can simply
click on the link. Let me show you
how it looks like. Alright, so fundamentally, it starts from the
position that we last left the camera on
in After facts, right? And is gonna do exactly the
moment that we told it. And let me show you. We can navigate into the video. Whoo. You see what's happening? So people can then just Look, you know, find for details that can hide in your
environment, right? You can also use audio. You can make it very
immersive with us, you know, like this. You can also use depth of field. It's completely supporting
everything, you know, three D elements, JOB
files, OBJ files. Where we want, drag it in, you know, use
lights for shading. We we want. That's
it. Alright. And make sure that you use
high resolution, so it's going to give you a
full quality, and that's it. That's it. So have fun. Make sure that your
computer is powerful enough and make sure that
you have a GPU with Cuda. So in Va, cards usually are
the best, and that's it.
59. MOV vs MP4 (technical): Alright, so when your
project is complete, it's time to export it, right? So you're going to have
two macro choices. If after the X ft, you need to go back into
another software, for example, even after fax itself, maybe you need to preserve the full quality
one on one, right? Whereas if you need to
just go to the upload, you need to x42, maybe
a more compressed file. Um, just to have a more
manageable size, right? So, in that case, we're
going to use MP four, right? So let's go slowly
on the two cases. In this case, we're
gonna export to preserve the full
quality of all this. Alright? So, in this case, you
go this weird thing moving we need to choose the area of the timeline that
we're gonna or, right? So in this case, the
timeline is 10 seconds. I need to export the whole
thing, so I keep it this way. In case you need to for
specific parts of the timeline, simply press B and N.
That's a shortcut. Choose a area, you know, like this, you know,
B and N. Alright? You can bring up
these keys by going to here, edit,
keyboard shortcuts. And here just remap
wherever you want. Tech whatever key
you like, right? So we go to File, Export, and add to render queue. This. Here there are three main rows that
we need to regulate. Usually you don't need
to touch the first one. Best settings means that it will automatically choose
the best settings. So don't touch it. We'll use the full
quality, alright? Then to chose the actor format, we need you to click
the second row. So let's go here. Okay. And so if we need to choose
preserve the full quality, we need to choose quick time, quick time here, and then here, make sure that
animation is enabled. So don't be misled by the name, even though it says
animation actually is all around preset that
works with everything, not just animation,
even footage. So fundamentally, it
will allow you to have a slightly smaller
file without QLT loss. Alright. So let's use
animation forever. And then you have two choices. If the video has transparency, I need to preserve it, right? You need to click here, RGB and Enable also plus Alpha. Otherwise, if you just need to export the video
with RGB, right? So you need to
actually render the black as black, use RGB, right? So then copy myself to this. In case you need to
change the size, you should kind of upscale
it slightly for YouTube. You can do it from here, resize, change the size like this. And then remember to enable the audio in case your
project has audio, keep the sample rate
as the original. Don't change it. And this, the higher the better, so
through the two, it's fine. And then don't change
anything else. Make sure that
stereo and confirm. And then just choose
the final folder. Okay so we're going
to place it here. For example, test Okay, and then just press Enter
or click this Render. Wait a couple seconds,
drink some water, boom. Alright, done. So at that point, you will we can double check
if the file plays correctly. On some Max, the MOB file
doesn't play correctly, so you need to download VLC. VLC is able to play
mov files correctly, so I'm going to give
you a link right now. Okay. And o, whereas
Windows just works. Even though I recommend
using VLC anyway. So let's just recycle
these render settings. Duplicate with file name, okay? Let's go with a second case, where we need to compress
the file, right? So let's go here, second row again, change
the package to 8264. Okay? Don't watch any
other format, this ho 64. And then to choose the
actual final size, you need to go to form options. So in case you need to respect the specific file size limit, you go to VBR here and
then decrease this slider. The smaller this is, the
smaller the final file will be. These are the
megabytes per second. But also, it will
trade off the quality. For example, when I
need to import this back into a digital
audio workstation, I initially go very low, maybe 2 megabytes/second,
you know. Whereas, if this is for the final upload, the
higher the better. So 100 megabyte, 200 megabyte, 50 is very good, right? So in my experience around so A density that is 20-50
megabytes/second, that should be indistinguishable
quality, you know. But, you know, for now, if you want to preserve
the most that's possible, keep it around 50 and up. Alright, VBR means
variable bitrate. Fundamentally, it will take
slightly longer to export, but it will probably make a slightly smaller
file, more efficient. And this is the
other way around. If you don't have time, you need the fastest
export possible. Use CBR, which is
constant bit rate, but obviously we'll make
slightly bigger file. So in this case, we have time, so let's use VBR and then
just go here and confirm. Don't change anything else. And that's the thing. Remember that MP four is not able to preserve
the fourth channel. As you can see, Alpha
becomes unavailable, right? So if you need to still
use Alpha channel, because you need to
reimport the file into another software or into
the software itself, you're forced to use MOV. And finally, just keep
here everything the same, same rules for the ad you apply, and just confirm, and
then just render. Okay. And now we're going to compare the two
file sizes, right? As you can see, the MOV is 37 megabytes versus four
of the B four. Okay? So as you can see, let
me just play through. Let's see the quality
of the two files. Okay? So this is the MOV. Alright. And let's see
the second one P four. Alright. There's just a slight
difference with blacks, but it's fine because it
makes sense, you know, this is slightly
more compressed, so it kind of loses slightly
some dynamic range. But it's fundamentally
undistinguishable. Now you can take
your conclusions. Okay, the truth is that if you try to import the MOV
into YouTube YouTube, it may sometimes remain stuck
in the encoding process. So that's why using a
more universal format, P four is still better. And I'm going to show you
in a few minutes how to preserve full quality
using p four.
60. Media Encoder (background rendering): If you wonder if there's
a way to actually fully predict the file size when
you export as P four, yes, but you need to use another Adobe software
called Media Encoder. All right? Make
sure, first of all, that's the same version as
your after effects, right? So if you use After effac 25, encoder needs to be 25, 26, 27, 28, all right? Make sure that
those are matching. Okay? That's said. Let's right click
here on the MP four. Okay? And simply to use it, instead of pressing
Rando from Afrofax, we need to use this
second button, Q in A and E. Alright? So click this button. Don't worry about
the stings there. It will automatically
override those setings. Okay, so now we are here.
This is how it looks like. And we simply works
the same way. You choose the package
first, the file format. And here, actually, we have
some extra options, HS 65, which is a newer
and colder actually better than HS 64. So
we're going to use that. And here, if you go to video settings and you
scroll down slowly, you may see the slider
for the size. Here, okay? So these things we
know, but you see here, we have the prediction
of the actual file size, which is pretty handy, right? Okay. And don't use ever slowest here because
it's going to take forever. Use just one step before that higher in case
you like quality. And then just press here, play, and they will do the same
thing that we did play here. The cool thing about using
media encoder is that, first of all, is the most
stable type of render. So it kind of reduces
the possibility of crashes during the render. And it allows you to use after effects while it
renders on the other side. So keeps after
effects completely free to keep being
used at the same time. Weight is kind of dangerous. So you can do that only
if you really have a very powerful
computer with er cores, at least at least 16 cores, and also 5 gigahertz of speed, and at least 64
gigabytes of RAM. So those are kind
of the min things to really make the stunt. So having a render going there plus keep working
on another project here. But that's something
that it's possible. So it's called render
in background. That's one advantage
because if you are critical on time, you
can do stuff like that. Even though the best
option is always having a complete separate
machine for the renders, and then that way, you can keep working on your
main machine full power.
61. Export Settings: If you notice that you use the same exact settings
for the export very often. It's going to be very
boring after a while to set those up manually every
time here, here, here, right? So there is a way to make templates for these
export settings. So for example, we just go click this arrow here
on the second row here, and it's going to give you the
option to create template. So scroll all the way
down, make template. You create, first of all, the settings of your
template from here. So you set up this,
complete this up. And then let's go
this test here. Okay? And then you can make
that become the default for the movie export or in case
it's for a frame export, you change it from here. And that name becomes
immediately available. So if Kol called this test 13, it will also show it here. Movie default now
should be an option. Test one, two, three, alright? So very handy. And that way, when you set that up, it will automatically show that
template type template. Then the next time
you trigger a port. So if you do control, if
you go to the export, in this case, it chooses
a different preset that I set up, that's it. So that's how you automatically certain settings ready to go, so you can jump stretch
to the rando button. And you can also change
the naming of the file. For example, you can make it fetch automatically the name of the composition or of the
project just by clicking here, this button here and you
can create a template for which name the export
you automatically have. So for example here, I have
a template for project name. And automatically gets the name of this project file,
as you can see. And yeah, all these things are tiny things that
make you save time, time, and time, which is the most important
resource that we have.
62. Single vs Multi Frame Rendering: Um, let's go to edit preferences
and memory performance. So right now, I'm giving
the computer the permission to use about 80% of all my RAM, you know, so so
it's going to use 100 out of 127128, okay? But we're going to export
in single frame mode. So one frame at a time, right? So let's see how
fast the render is. But we need to first dump the memory because it
is already rendered. So it's completely in
ram memory right now. So let's dump it
with the shortcut. I showed you at the very
beginning of this course, so we clean the
memory, okay? Exactly. So now, we go to the port Q, and we just call
this file a single. Okay, boom. I render. Let's see how long it takes. All right, 13 seconds. Alright, so now
let's first of all, clean the cache so
we cannot cheat. So control old P, and clean. All right. Now let's go change
one simple thing. Edit. Preferences, memory, and we're going to enable
multi frame rendering. So enable this guy. And in my case, I'm going to give permission to after effects to
use up to 90% of my entire CPU and up to 75%
of my whole RAM, alright? And maybe it'll
just slightly more. Let's get just 24
for the system. Okay. So let's just go back here and attempt a new render and
don't change anything else. So let me try to
make a new file, and let's see. Woof. Alright? You see what just happened? It took 7 seconds compared
to 13 or earlier. Alright? So this is how fast the render can come
with multi frame rendering. The issue is that
it's very risky. So for short renders or when there are not too many
effects, that makes sense. And also if you have a BV CPU, so you need to have at least
eight physical cores and 5 gigahertz speed and
64 gigabytes of Ram. These are pretty much
the minimum things to evaluate using multi
frame rendering in certain conditions. But normally you keep it off. Otherwise, you may just lose more time because the
renderings may fail, and then you know, in the process of attempting
new renders every time, you may just lose
more time compared to just run a single single frame render that just runs smoothly in one
attempt, you know? So that's the point. But for short renders and for
not complex projects, multi frame render,
by all means, use it because as you
can see, it gives you a significant
boost in speed.
63. The Perfect MP4 Compression: Two so there is a way to compress a MOV file
into a las MP four. And to do that, we need to use
a software called D Coder. If you have Mac, you
need to use alternative, which is exactly
the same software. It's called hand brake. I'm going to give you the link of both right now, alright? So the way it works is we need to simply open this
software, so we go here. Okay. And we drag in
our file, first of all, so we want the one
that we just exploited with Afrofax the MOV, which is 34 megabytes
big right now, okay? So let's actually, let's
open this gear here. And we go here and
you select MP four. Okay? And then you
check on web optimized. Okay? And just copy my
settings fundamental. Then you go to sizing here and just copy exactly
all my settings here. Alright? And then simply
in case your footage has just double check that the input and output resolutions
are the same, right? If they are the same,
simply tweak it from here. So simply change one of these
dimensions in case it's four K. You can
actually artificially change the resolution like this. It will automatically
upscale it. So in my case, let's keep it as the original one on
one like this, okay? And then just go
to video filters. Make sure that
everything is off. Then go to video encoding, copy exactly my settings, you know, screenshot this. Make sure that is
everything the same. And finally, this is the
most important thing. Here, you need to type
hyphen hyphen lossless. This simple string here tells the file to use the
lossless compression, okay? And then, just go audio here and just use this setting here. And then just
confirm to confirm, simply go out of this window
here from this x here. And then you can
just press encode. You choose the final
destination of the file and wait a couple seconds. Alright, and poof, you
see what happened. Now the file is
ten times smaller. So 4 megabytes, and the
quality should be exceptional. So let's go to the file.
Let's see how it looks like. Alright. So, this is
the way to have MP four while preserving
the full quality as much as possible, right? So, and this is a
exclusive feature to this type of software. You cannot do this internally
in offre fax or encoder. So you need to use
coder or a handbrake. Okay? And that's
my preset for you. And then you can save these
settings as a preset, simply go here, save us, right. So as you can see, I
have many presets here. If you want some
general guidelines, this preset is
perfect for YouTube. Even better if you
let me show you. What I do exactly for
YouTube is to even use a MKV format because MKV actually allows you to use the same thing that
I showed you here, plus last less audio as well. So that's the plus
of using MKB format, but this is only for YouTube. The other platforms force you to use as audio compression. So you need to use the other
one I showed you, A AC. Okay. And what else? Then IG, remember that it's not very tolerant with
high quality files. So in my experience, the file needs should be
within 100 megabytes big max. You know, doesn't tolerate
bigger than that. TikTok instead is more tolerant. You can go all the way to
half gigabytes file size. Anything higher than that
is just inefficient. So instead, YouTube, you can go even to 1 gigabyte
file with this preset. Don't import MOB
files into YouTube. It's going to take forever,
and sometimes it may interrupt the process of
the encoding in YouTube. It may give you errors, right? So just use either MKV
for YouTube or MP four, Lastls for the other platforms. In case IG in case one of the other flat platforms rejects your file, simply don't Lastlss. So disable this strength and create a loss
preset, as well, okay? And just play with the sliders.
64. Render Engines (technical): You may also experience
different render times by changing the render engine. So actually, when you go
to the project window, you also see this sway shaped
icon. If you click it, you will see the options
of your computer. So if you have a GPU or just a CPU with a
graphic acceleration, you may see different
options here. So Guda is usually the fastest possible
option that you may have. In some computers, it's
usually the slowest ones, you have only the
first option here. So this kind of level one,
level two, level three, right. But sometimes you can now use level three
for some reason, becomes completely
unsustainable and unstable. So you need to downgrade
to level two x for. Normally use the highest level, you can see, one, two, three.
65. Proxy (Normal Methods): Sometimes your project
may be unmanageably slow, so I'm going to show
you four ways to increase the speed of
the previews, right? So right now we have
a four K video. Let me show you, okay? And so you can see if I put this on full quality with
just these two effects, emotional lure plus displacement over time, let me show
you what happens. You see how long it takes
to make one single frame, and I can grant you that
the CPR musing is not slow. As you can see, it's
taking some time. And the first option is to simply decrease the
quality of the preview. So you go here and you
decrease it to one quarter. Let's see. Alright. Okay,
that's already one solution. Simply do this. Right. In
case you need to escalate, you have a little even
more speed than this. Okay. Let's first of all, you know, let me just
fix this by the way. The time code is
broken. As you can see, we now start from zero. So if the time code is
broken, per control K. And change this
number here to zero. Boom, boom, boom. Alright. Now it starts
actually from zero. Alright. So even after
you put this on quarter, it will still be uh, slow. So you need to try
the second method. Press this button here
on every single layer. So this is kind of a
microcompression that, in some cases may make become the preview faster, even faster. So this has two modes. This is no compression. This is slight compression, and this is heavy compression. Heavy compression may cause
issues, how some artifacts. So don't use it. That's my recommendation. But so in most cases, just use the first mode like
this. Let's see. If we go to fall, let's see. Okay, still taking some time? Okay, I already became
slightly faster. Earlier, it was taking about 1.1 seconds per frame
in full quality. Now, we already went down
to 900 milliseconds, which is already a
micro improvement. Alright? Alright.
Okay. So, but in case, this is still not fast enough. Okay, let's go to
the third method, which is create a proxy. A proxy is really a compressed version of
one of the footage files, fundamentally, you
know, and so we can right click on the footage that is creating this issue. So let's reveal this. First
of all, to see what it is. And then just right click
on the file at this level here, create proxy movie. Okay, and first of all, let's change the folder. Change the preset here to HS
64, 5 megawatts/second here. Okay? And then change the resolution also to
a lower resolution. So just enable resize, and then let's go all the
way down to standard HD. Alright? Like this. And then let's keep everything
pretty much the same. Just copy my settings here
here here and confirm. Okay. In case you in case you
need to have transparency, you need to stay to MOV. But in case you don't
have transparency, in case it's just the video,
you can do exactly this. So just confirm, no
Audi is needed. Okay? And so let's place this file in a folder called proxies
next to your project. And also don't place it
on an external drive. Use all your files always locally when you work
on that project, right? So let's just save
the file here, we can call it proxy one. Okay, and convert. So what it's going
to do now is to convert the file to
this new version, so it's making a copy, and
then it's going to connect it automatically to the original file,
four K file, right? So we're going to see
now a second option up here here when the
conversion is completed, and we are able to toggle back and forth between the
two versions, alright? And that hopefully
is going to give us a further improvement in
the preview generations. Alright? So let's see. Okay, let's go back here and
let's see what is happening. We so to see the
check box somewhere. I think I just created the
proxy. Oh, I didn't set it. Okay, yeah. I forgot you set the proxy
as well. It's fine. So we we click on the file and set proxy and go to the file. In this case, is the V drive. So let's go here and simply double click on
the file here, proxy one. Alright, now it's
connected. Okay. As you can see, we have this checkbox. When this is on, it will use
the file right here, 720. When we check it, obviously, it's going to use default standard uncompressed version
of the footage, alright? So the point is that it will
this is all you need to do, just make sure that you
enable this, keep this on, and it will automatically
apply it on the timeline, you know, to wherever copy of
the footage is being used. So let's see if there's any additional
improvement in the speed. Yeah, still pretty
much slow, right? Still taking almost
a full second. Since it's still 1 second,
it's crazy, right? If the issue persists
at this point, why don't you escalate with
a fourth final method, which I called native proxy.
66. Proxy (Advanced Method): This method is very simple. We search you or create an object in your main
timeline, so no object. And we call it scaler. Okay? And we start you
parent everything to that scaler. So
like this, boom. Okay? And then let's change
the resolution, the actual resolution
of this project. So control K down to 720, like this. Alright? Boom. Okay. And then let's simply open the scale of the scalar
here and then type 100 divided by the
vertical resolution of the original resolution. In this case, four K, so is 21 60, right? And then multiply it by the new resolution that
we're currently using. So right now we're
720, so times 720. So to recap, 100 divided
21 60 times 720. So this formula
allows you to give the perfect number
to the scale so that it looks the
same again. Boom. Oof. Alright. Okay. And, and then, okay, now that we have this, we can scale this back up
like this. Alright. So at this point,
this is not over. We still have here the 44k file, right, which is very
annoying, very heavy. So let's precompose
the footage file. So we simply place it inside amazing leave all pre comp here. So we call this footage, okay? And inside it. So inside this
guy, double click. We disable the original. So this round the
original four K. Okay? Which is this
file here, right? So this file here now we need to uncheck the proxy
of for this method. So we finally keep it original. We don't need the
proxy of this anymore. So we can unlink it, L this. Set proxy? None.
Goodbye. All right. So it's just four K original. We disable it now. We disable. And here, as you can, guess, we import natively the
720 version of the file. So just double click
this. Okay, import the 720 and drag it
here at the beginning. So now it's perfectly
synchronized. And then just kill
it up for now. So just right click on the file, transform and fit to come. Boom. Alright. Okay. So this is the very light
version of the file, right? Remember that this is
just the 720, right? And I see in yes, now, when we go back to
the main composition, where we're going to
do our project with the effects all full
quality, right? Let's go here. So, right. Let's make sure that this
is the main composition. Let's see which one it is. Okay. So now that we're back
on the main composition, we can do our product
as regular, you know. But the point is that now, even though we are on
full quality here, we should have a
significant gain in speed. Let's see if that's true. Poof. You see how quicker
is rendering, even though we are
in full quality? It's maybe two feet, less than it's taking
it's like 75% faster. And you didn't hear
the best thing. If we can go even lower now. If we go to one quarter, now it's gonna be even faster. Wow. Almost real time, right. But then, obviously, the
resolution such to be so low that maybe a little problematic to understand what is happening. But you can you can pretty
much do it, even if you want. So now it's really let's
see, even 100 milliseconds. Right? We're y almost ten
times as fast now, alright? So and then when you're
done with the project, when all the effects
have been done, you simply go back to four K. You do the reverse of
what we did. So you go back here, footage, Renable the four K, right, and then you go back
to the main composition. You change the resolution, right here, change
the resolution of the entire project
back to four K, right? And then simply the scaler here, which for the whole time,
has been just here waiting. We just put it back to 100%. And then simply we are
ready for the full port. Obviously, now returned
to be very slow. Let's see. No, we'll
probably, over 1 second. We're already gaining ten times the speed
of render, you know? So that's the point.
When you work, speed is the most
important thing, right? And this is the ultimate way I found in my life to
have a ultimate proxy, true proxy that
gives you advantage of preview speed, right? All those things unfortunately
are iffy, right? So, this is a custom solution
that I created for you. So you're welcome.
67. Still Image Export and PNG Sequences: Exporting a frame Affax
is reasonably easy. I need to go to the frame
but you need to export out. So for example, this guy here, and we press Control hot a Alright. I will automatically bring us to the render queue for the frame. So it's really targeting
just the frame this time. And just as I showed
you, with the videos, you simply create a preset for the PNGs in case,
you know, here. But the point the first time you will need to
do this manually, and you need to also have
Photoshop installed. Go here. I feel like it will ask
you to install Photoshop, but probably by the way, the first time we'll ask you to save a PSD file somewhere. So when it does ask you that, save that file somewhere I will never be
deleted or moved. That way, from the
second export on, it will only ask you
just the PNG straight. Okay, so at this point,
that's it, we go here, PNG, just go here, select either PNG for
full quality, no loss, or if you need to compress
it, select JPAG here, JPAG and you can choose the level of
compression from here, you know, all these
things like Photoshop, you know, so in this case, I don't want to
lose any quality, so I do PNG and so
simply copy my settings. If you need to preserve
Alpha, you can do it here. In this case, we don't can resize it, all the
classic things. Just like this, save. And let's say photo one like
this, and then confirm. That's it. So now we have the file in that
folder, ready to go. Alright. Sometimes you may have a PNG sequence to import, right? Another way to export
video is a PNG sequence, so you need to break down the video into the
single frames, right? So that option is
available by doing this. Go into the render
Qu as if you're going to export a video, right? But actually, you need to
change something here. In the format, you change
to PNG sequence here. So when you instead of
exporting the video now is going to export every single frame as a separate file. And you can do that in
a JPEG or PNG format separately into that folder. So then you just save,
and see what happens, now the files are going
to have the name you give them plus the number,
the index, you know? So let me just show
you very quickly. So if I export this. So if I save now, Okay, I'm going to stop it because it doesn't
matter in this case. Okay, I'm just going
to interrupt it here. If you going in
the export folder, you will see that there
is a new folder here now. And in fact, we have every
single frame separated, you know, as a single
file, you know? So we have literally
the entire video broken down into
every single frame. And as you can see, they
have a progressive number. And then you can just
batch rename those. So just selecting all those
like this from the first, last right click
on the first one. There are different methods,
you know, to do this rename, for example, just
a simpler name, maybe just a test. Right? And it will
automatically rename everything with a
progressive index. Then let's suppose that
we want to reimport this back into Afrofax. So someone gave us a
PNG sequence, right? So to do that, we
need to go back to Afrofax. Double click. Double click here as if you're
going to import a file. Okay, and we go to the folder
where the PNG sequence is. And after the files have been
renamed with this method, I showed you, with a progressive
number, very simple. Double click on the first file, but before you do
it, double check the PNG sequence is checked on. And now it will work. So if I do this, double click on the first file
of this series, it will automatically place
those one after the other. And the cool thing is, it will interpret this
as a video already. So if I drag this
into the composition, let me show you of, it's already a video
fundamentally, you know. So I can then time remap it. For example, you
know, I can stretch it like this, and it will work. Let me show you if I put this
in solo, it's really video. Okay, let's go here. You know. So right now it's
literally playing obviously every image
after the other, as if it was a video, which in the end, videos are just a
sequence of images.
68. Real Motion Blur (Secret Trick): So normally, when we shake
something like this, so right now, there's
a simple wickle on the position,
right on this image. We can enable the default motion blurr,
which looks decent. Alright. And that's super efficient for the
CPU, that's good. But the disadvantage of the default motion
blurr is that it makes you lose some brightness
from your materials, you know, so it makes it
slightly become a little darker. And this happens the more
we shake the footage. So if we do very high movement, this may become very apparent. You know, we start to lose definitely some brightness
on the footage, right? So there's a little trick to preserve that information to preserve the true brightness. And this is especially important in hyper globe based effects. For example, if you use saber, the plug in, if you do
light saber effects, if you do those things,
this is very good for you. Okay? So let me show
you how to do this. So with the save, actually let's take a
screenshot of this. Okay. And we disable for
now the motion blur, okay? And then we simply go we just
apply a adjustment layer, right for simplicity here. And on this adjustment layer, we apply a couple of things. First of all, effect
called exposure, right? So, and you copy my settings, simply copy my settings
this way, okay? Then apply a second exposure, right, and just copy
my settings again. So, fundamentally, the
first one has gamma 0.1, and the second 110, okay? And in between these
two exposures, we put this effect
force motionbur with these specific
settings, okay? And then, oh, you can
ignore this. All right? Alright. Okay, so let's act enable this adjustment
layer, right? And finally, let's go
to the project view. You see that there is
a number here, A, BPC. This is the bit
depth of the colors. Normally, keep this
always to a bit, okay? But for this specific trick, we need to increase this to 32. And if you want to
know, technically, the reason why we do
this is that this way, by going through 32, we enable infinite numbers fundamentally to
represent colors, and this is what we need
for this trick to work, right? So let's do that. Okay, and it's okay, let me now just play to
show what is happening. Woof. You see that? Now, all the brightness is just preserved
during the motion blur. You know, we preserve all that amazing juicy
brightness information. And this is assuming, extremely important when you use
hyper glow effects, especially saber effects, right? You see that. Now we see we have so much more detail
during the motion blur. Alright? Let's compare this to the previous
version. You see that? In comparison, the
default approach with motion blur makes you lose
so much of that information. But with this, all
the luminescence of the footage is
preserved until the end. Alright? The only
disadvantage of this is it's very heavy, makes everything four
times heavier to render. So as you can see here, we have 9,100 milliseconds per
frame in full quality. So as long as we
downgrade everything, so we disable this, we
disable the 32 back to eight. Like this and also we reenable
normal mushroom blur. We are lily almost ten
times faster, right? So that's the only disadvantage. But sometimes for
specific scenes, you may need to use 32, especially if
highlights preservation is very important in
combination with mushroom blur. In case, good news, you can isolate specific
compositions, right? You can precompose just thin to have only 32
bits for that sin. Alright? You can do that and use eight bit for the
rest of the project.
69. Project Size Optimize: Okay, so let's suppose that you finished a project
and you're ready to send it to someone or you
need your archive right, it's very likely
that the signs of your project can be reduced
significantly, right? And that happens
because in the end, when we start a
project and we end it, we may not have used some of the materials that we
imported originally. That's the point. So
there is a way to automatically remove
all the things that we didn't use
in the end, alright? Without ruining the
result of the project. So, okay, in this
case, for example, let me show you we have let me show you the size
of this project. Right now we 7
megabytes big, right? Okay. And in the end, I only use this
composition only, right? So it's very likely that I can reduce the size
of this because here, there are so many compositions. I just simulated this for
you, you know, purpose. Okay, but this is something that happens in real projects. Alright, so the trick is this, you select the main composition. So the one that
contains everything, r. So select it like this. So here, and then go to file. Dependencies, reduce project.
That's what we need. So when you click this, it will automatically remove
all the things that have now been used. Alright? A CNC removed
almost 300 things. And let's check out the new signs of the project.
Alright, cross fingers. Woof. Hundred 54 kilobytes. Alright, so it's like
hundreds of times smaller. You know, that's the point. Be only careful to one
thing. When you do this? Suppose I make here, a new comp, test. Okay? Inside this comp, I have a special null
object like this. And then I make a
long distance cable, for example, here
to the null object. The rotation of this null object to the rotation of
this, you right? So now the main one here
is referring to this here. And the issue is that now test is not inside the
user comp, right? So Afrofat thinks that
I'm not using test, which means that if I
reduce the project now, I go here, user, reduce. Let me show you. It will
also delete your test, which is actually
not what we want. And thus, now it will
break obviously, it will break the expression.
You see what happened here? So when your project has this case, you
have two solutions. Just go back here. You
have two solutions. You either simply make sure that the order comps are actually inside the main comp like this, and that will already
fix this issue or simply select also those long
distance compositions as well before you reduce. So you simply select both
test and user in this case, and then that way when I reduce it will not delete
those. You know, that's it. And finally, especially after you have reduced the project, you may need to have a way to automatically
archive everything in one single folder
because sometimes real projects have many
files from different places. So it's very hard to
manually track all the files down and make a safe standalone copy of
the project, right? So to do that, automatically simply go
to file dependencies, collect, and then choose simply the folder where
everything will be saved on. So for example, here, let's call this
archive ready one. It goes save. Boom. And inside that folder, you will see simply
one project file, and then the footage, all
the footage that has been used automatically
goes into this folder. That's it, you
know? And down way, you can simply pack this
as a zip file and send to your collaborators or
simply archived on your drive. And down way, you don't worry
about some missing files. This completely eliminates
the issue of missing files. You know, After fractures, able to open versions
of project of Afrofax that come from older
versions of the software. So if you have After effect 28, you can open a project from 25 for example 26, 23,
whatever, right? But you cannot do the opposite, which means that if
your collaborator uses a older version
of your project, you need to send your project coming from a new version
of Arafax to them, then for them to be
able to open it, you need to first downgrade
your project, right? So to do that, simply
go to your project, obviously, and go
to File, Save as. And as you can see, we'll
show you a couple of options to downgrade your
project in this case, Afrofx 22, the 18
or wherever, right? So if you use Afrofax 28, you can go to 26, 25, xc. So that's the way. That way, they will
be able to open it and then send it back
to you, and that's it.
70. Dynamic Link With Premiere: Alright, so let's
suppose that you have also edit
videos in premiere. So you have here, for example, a edit right with these
nice images here moving. And let's suppose that I want
to go in Aphafax with this. Alright. So there
are three main ways. Let's go in Alphafax. The first way is to simply
double click here as if we're going to import the
material. So double click. And double click on
the project file, simply this guy, edit
project here. Boom. So really double click on the
project file of premiere. And it will tell you, Okay, which timeline do
you want to import? In this case, there's only
one choice. And that's it. So now it will treat the entire premiere
timeline as if it was a pre rendered
video like this. But this is also dynamic. It means that if I do any change in
premiere, for example, I disable the video right
now or maybe I cut this way, so I disable it like this now, obviously we detect that. And obviously here
disappears, right? There's completely
nothing, right? And then the rest of the
video is still fine. So it's completely
connected in real time. So and okay, that's
the first way. And obviously, remember
that if you do any change here in Aftereffx, it will not show up
in premiere, right? So because the
connection right now is from premiere to Afterefx. So the first result
actually is here right now. Okay. And, okay, so that's
the first way to do this. The second way is
in case you want to preserve the actual costs that you did in premiere, right? So to do that, we
need to instead, simply right click here again, Import, Import
Premiere Pro project. That's the option, right?
And then, same thing. Double click on the
project file of premiere. And as you can see, it shows you a different window
this time, confirm. And this time, it
creates the actual after effats composition because poof, it preserves the full cuts that we did in premiere, right? But be careful because
when you do this, you start to lose
the dynamic link. So here, at this point, it's kind of independent
composition altogether. So this method is
meant if you then you should stay in Afterfax
for the rest of the time, you know, and complete
the entire project here. Alright? So, because let
me show you right now, even though the second
clip is disabled, it will still show
up here, you know? So now at this point,
it is disconnected, completely now independent
in Afrofax, all right? And finally, third case, let's suppose that you want
to actually stick for longer, actually spend most of
your time in premiere. So you keep the timeline
here, you do your ds here, and you need to use Afrofax only for specific clips, right? In that case, you simply
select the target clip, for example, this
guy here, okay? And we rightly conduct clip. And then go to replace
with Afrofax here. Okay? And that way, we'll create a Aftereffex
project for this. And then in Afrofax, we have a mini composition
exactly for that clip, which is still
connected to premiere. So it means that now we can do our via fax, the
crazy stuff here. For example, let's apply
a nice displacement map. Let's destroy this image
a little bit like this. Okay, so now if we
go back to premiere, we should see the
result happening. So let's see here. Wow. Alright? So, because now this clip specifically is connected with After
effects in real time. So you do your effects there, and then you go back
here and you can continue you edit normally
in premiere, you know? So, so these are the three
main cases to use dynamic
71. Projects Inside Projects (Templates Godsend): Okay. Alright, so now let me show you a sleazy but amazing trick, especially for templates,
alright, to simplify your life. So there are some times where you have nested comps
inside each other. For example, let me show
you what is happening here. We have one comp
called word one, right, that contains
these four words. So, into each one, there
is one word, text, right? And there is a second comp that contains this next word and yada yada until word
three, word four, right? Okay. So the point is that, yeah, we can customize the text. That's it, okay? But let's suppose that we want to do
a second sentence, right? So in theory, you can use a trick I showed you
at the very beginning talking about precomps simply by duplicating the word
one, precomp, right? So if I do duplicate
only word one, now we can drag it into
the main composition here, right? So let me show you. Let's make this smaller. Word one, make it smaller. Let's keep it there, right?
Okay, let's keep it there. So now in theory, thanks to the fact that I
duplicated word one, which contains everything else. In theory, I can just
customize now the first word. Let's see, and have an
independent set of words. Let's see. No, that's not true. Only the first word now is
still independent, right? Because now if I change the second word inside word
five, let me show you. Like this, make it
a little smaller. Let me show you if we go back to the surface. You
see what happens? Now this also changed because
it's still connected. So, this means that in theory, we will need to go inside word
one and duplicate each of the comps to truly make it independent and then override
it with the old trick. As you can guess, this will
take some time, right? So in these cases, when you have nested comps, the trick with override
becomes inefficient. So you need to do something
a little more advanced. Simply go to the
surface composition. Let's delete it. Word five. And instead, we just double
click on an empty spot, and we simply report the entire
project file. We go here. This project is called
Project inception, so we double click on the
project file itself, like this. And then that way, we have a truly independent set of compositions
already, ready to go. So at that point,
we can just drag actually finally word
one, just rename it. With something different so we know what is what? Let's good. Let's call this new. Now,
if we drag this here, now this is truly independent. Let me show you if I
change this like this and then we're here like this, Yara, yaa here, like
this, like this. And finally, here. Alright. Let's go. Okay, let's go to the surface. Let's go all the way back here. Wow. You see what is happening? Now these four words
are ready to go, fully independent,
fully detached. Alright? So in these cases, you need to use the trick to just have a better
time and save time.
72. Mogrt Import / Export: Mbars are fundamentally
After facts templates that you can
use for premiere, right? So to create one, simply go in After effects, design your animation,
or wherever. So in this case,
for example, let me show you how a wiggling shape. The scales in and then goes out. So we can use it to then make a placeholder where
we can just drag in our image in premiere
and we'll apply this animation to whatever
element we want, right? So after you prepare your
animation with key frames, expressions, first of all, it should trim the composition
to the desired length. So in this case, this animation ends pretty
much here at 4 seconds. So let's go there with
the work area and we trim the work area,
right, trim here. This way, this is
the length that will appear in our facts when we drag this into the
timeline, right? Otherwise, it will be
unnecessarily long, right? So after you do
that, save, okay? And then let me show you. First of all, we need to design the interface for how it
will look like in premier. So to do that, we go to Window and then
essential graphics here. And here we can drag
in the things that want you see in premiere. So first of all, let's
select the composition. It's called main, okay? And let's also make a thumbnail
so we know what it is. We set poster time, and
then we give it a name. So let's call this scale
in scale scale in one. Okay? And then we simply drag in the parameters that we want to
regulate from premiere. So for example, we have a simple multiplicator for
all the wiggles at once. So this is very simple. So double click on
the property that you want to use in
premiere, double click, and we'll show it
down here and then simply drag it from here
to the interface here. Okay? And then here we
can just rename it. So, for example, wiggle, let's simplify it
even more as a name. Wiggle, simple, like
this. All right. And so this number by
default is going to be one, but we can also constrain the maximum range
of this slider, as you can see, for example, it's going to be 1-10 max, okay? We need to precompose. We need to precompose
this shape layer, right? So we control shift seam
to precompose it here. And we call this
container, again, like this, leave
all and confirm. Okay. Everything's still
working, still fine. And simply at that point, drag the composition into
the interface like this. Alright? So that way in premier, we would just drag
in our stuff here. Let's also disable
the default shape at this point, height. And yeah, and that should
be pretty much ready to go. These two things are enough. Let's keep this at the top. So you kind of design some
kind of mini template, yeah, you know, second
tea is pretty fun. And finally, we're ready. So we just export it
from here, export. And at this point,
you have two choices. If you need to send
this to people, you need to export it
as in the local drive. So you will just place
the actual file into that position in the folder and send that file to people,
the mogart, right? If you need to synchronize
straight to your premiere, simply choose the first option,
local templates, alright? And then don't use
anything else, just keep everything disabled,
and then just confirm. Okay, wait a couple seconds, then Let's go in
premiere. Let's see. So in premiere, you just go to the Essential
Graphics window here. If you made in some
premiere versions, you may show this as
a different name, but fundamentally
it's just here. You know, just find
it EccentralGraphics. And let me show you now if you go to the list of
templates, we should see it. Let's see. Let's scroll down. Scale scale. It's called scale
Who here's scale and one. Alright. So and then you use it, simply drag it into
your timeline here. And then you access the editing. Just click it here. You can trim it as if it
was a video like this, trim it, select, and
see what happens? When I click it, it will show the option that we designed
in Afrofax, right? So I can choose
the wiggle amount, and I can drag in now any image I want. So
let me just drag. Maybe one of these
videos should be fine. So let's drag one of these videos himself
into this container. See what happens?
Drag in like this. Boom. Alright. And then you can scale it in
case it's stretch, design it, tweak the
scaling from here, and seems fine for now. Then let's go back
to the timeline. Let's give it time
shoot update. Poof. That's it. So it's using
my custom element and then doing exactly the animation that I designed
in Artfax, right? And obviously, I can choose
now, the wiggle amount. So don't be misled by the fact that here you see
on the integer numbers. As you can see, it only looks
like it is jumping 0-1, but actually you can still
use floating numbers. So if you input 0.1, it will still update. Alright? As you can see, it's still even
though here is zero, the element is still
shaking because it's actually doing 0.1. You see that it's
moving? Alright. And yeah, so you can
control the amount. You can make a crazier
wiggle or more subtle. Let's make crazier,
for example, here. And obviously, now it'll
be out of control, yeah. So let's go maybe
a little lower 2%. And finally, let's
suppose you to import Mogart back
into Afrofat because you want to reverse and
engineer those or study or just edit those in
a more advanced way. Well, in that case, you need
to be in Afrofat obviously, and go to file and then
go to open project. And then simply go to the file
where you have the mogart. So in this case, I just
export the die here, scale in one Mgart.
Double click the file. And it will unpack it. I will tell you, Okay, where
do you want to save it? Okay, extract here. And, we have the
actual source project ready to go for our
customization at this point, as you can see it looks exactly like the one that we created. As you can see
everything is there. And I can then do a custom version of Da Mogart in a more advanced way. Alright?
73. Where to find Templates (Motion Array): Alright, so the problem with
templates is sometimes, if you don't have a sufficient proficiency with after effats, using templates,
all of a sudden may sound very overwhelming, right? Because in theory, it's amazing. You know, someone else
made the hard work and you just need to import
your custom assets to the template and complex effects will automatically be applied
to your assets, right? So the truth is that you need to have a decent understanding of how the projects work in Afro fax to be able to
use the templates, right? So let me show you how you orient yourself in an
overall way, alright? Most templates are
presented in the same way. They have a single project file, alright, and you need to
import that file first. So you double click here.
Let's go into the folder. We have some templates here. So, in this case,
is template of um, fundamentally, overlaid for transitions,
right transitions. So the way it looks
is that usually have a simple project file like this in double
clk to import it, and it will automatically
show you the full structure, the full project structure
like this, right? And usually it's
divided like this. You have the final comp where you see the final
results like this. So if I navigate, I can see, in this case, all the variations of the same thing,
you know, variations. In this case, it's just
transitions, right? Then there is a edit
folder where you can edit the colors for every single one of
these things, right? And finally, something extra folders that you don't need
to touch, usually, right? So these two folders is where you need to spend
most of your time on. Well, first of all, let
me give you my trick. The way I like to
use templates is to just drag it in like I just did. Then I simply
continue to work on my main composition
here, for example. Okay? Here, I just
do my stuff here, and then I just go to the final comp of
the template here. I just double click
on one of these, to see what is happening,
which one I like the most. For example, here, let's
see number 48, okay. And when I like one of these, I simply import this one
into my main composition. So I simply drag it. I go to the main
composition here, and then I drag, in this case, number 48 into my
timeline here. Boom. That's the first
thing I do, you know? And then I start
to customize it. So to do that, if the
template is made correctly, you can already go
to the edit folder here and then go to the
same corresponding number. So in this case, 48 here, 48, double click on 48, and you see the
controllers for that. Okay. So if I click on Control, now I see the color
change settings for the transition 48, right? And then asian gas, we need to lock this thing here. Otherwise, we will
lose track of this. So we keep this view locked. We go back to our
main composition. As you can see, it starts
to be very confusing. You have all these tabs open. I can keep main
on the very left, so it doesn't troll away. And that way, I can just
go mid transition here. And then I can customize
the colors of this. You know, let me show
you if I do this. In theory, you should react. Oof. Alright, it's working. And that's one
which you customize templates fundamentally,
you know? Then when your
search will become more advanced, more confident, you can simply go
inside that transition, that elements of the template, simply double click it and just go inside until you reach the source
elements like this. And then you may see that there
are some hidden elements. So you can show those us. And then you kind of already know what is doing what, right? So you simply reverse
engineer the template, to figure out where you can control the
colors, et cetera. But if you want to stay
basic, that's the main use. That's the main way
to use template. You drag your ferret
into your timeline, then you go to the edit
folder if there is any and then just just customize the colors from that
corresponding number, and that's it, you know. But the truth, again, is
that you need to have a decent understanding
of the software to be able to move maneuver
yourself into templates because every template can change slightly can
be very confusing. So just make a lot
of practice with the software or ask an
expert to help you, alright? One of the best
easiest ways you find amazing templates is you simply
go to this website here. Let me send you
the link. And then you simply go to video
templates, after facts. And here, just search
what you need. You know, if you need overlays, you type simply overlays
here like this. You know, overlays
for what? For social. Alright, social or transitions,
you know, like this. And we'll find where
we want, you know, like this, in download,
you can preview, and the way it works
is that you will find the project file exactly the way I showed
you fundamentally. Most of these works
the same exact way. Enable a subscription. I think it's $30 a month. So if you do this as a job, it's worth it because we kind of simplify everything
for you, right? So make sure that you have
decent fluent Affaxs so you don't lose time
figuring out how the templates work
and you'll have fun.
74. Make Custom Scripts with Chat GPT Ai: When you start to become very
advanced in Alpha factory, you can start to design
your own scripts to make crazy
stuff, for example. So suppose that we want
to be able to create a grid of solids
equally spaced and also we can choose their
size so that then we can just process
those separately as individual layers, right? So let's do that. So do that simply ask
it what you want. So for example, let me top. Just give the code
absolutely no explanations. Boom. Okay, let's see. And it will give us the
actual raw JSX code. This is fundamentally
the content of a JSX file of a script file, it looks like this when you
open it with a text note. So let's wait that
finishes baking the code. Okay, at this point, copy. And now we just save it in
a folder as a TXT file, so just a simple
note file for now, let's go here and just right
click New text document. Let's call this grid one. Text file, TXT, okay, save it, and then change the extension
of the file to JSX. So change the extension
dot JSX. Like this. All right. Okay, and
now is ready for AFX. Okay, so we make it
empty composion simply. Remember to enable
scripting in case you didn't do it at the beginning
of this course, right? So you go to preferences, and then just go to
scripting and enable our amazing chat box
for the scripts, right? So here, make sure
that this is on. And finally, go to
file Scripts run. And then let's select our
amazing file grade one. S here, Double click.
Let's see what happens. Alright, okay, we'll sell okay. Just input the stuff. So let's go solid
with. That seems fine. 100 by 100. The color is gonna be that. Okay? That's good.
Distance between solids, let's make maybe 50
pixels a little more. Okay, and obviously we choose how many rows of solids
and how many columns. Let's make ten by ten. Okay? And then confirm. Fool. See? That's it. And now we can simply connect all these amazing eyes
to a noll object, for example, like this. Connect, connect. So we can easily control
all of them at once, like those, right?
And that's it. That's it. So now
the cool thing is that obviously they
are independent. So for example, if I apply a nice wiggle on the
position one of these, let's choose ten by
100 pixels. Let's see. Okay? And then I paste
the same expression to all the other ones.
Let's say paste. Now they should
become all crazier, let's see, enable motion blur. Okay, let's make a
little more contained. Just move it to the
center of the screen. Let's see what is happening. And also, we're
applying nice Glo can never miss any project. Sorry, I can't resist. Let's go here. Default.
Glow. Alright. Woof, let's. Alright. And that's it, right? So, so With scripts, you can do combining so you don't need to
be a code expert. You know, you just need to have some basic code knowledge. If you know the expressions
in after facts, the one I showed you, it's fine. And then you can just do more crazy custom
behavior things. Just by using Chi GPT, you know, so you can explain it what you want in case there
are some bugs. You can simply go
back to the code, explain the issue has been. You can give it a
screenshot of the errors. So go back and forth until
it's fully debugged. In this case, we're
pretty lucky work the first attempt because it's a relatively simple
task in the end, but it's a redundant
task, right? This allows us to
simply create now custom grids for projects, you know, very easily, alright?
75. Protect Your Templates From Theft: All right, so the
process to protect your animations is
divided in two steps. First, we're going to delocate the source and then we're going to scramble the code, right? So first of all, this
is your animation, for example, that contains
everything, okay? And we're going to first of all, delocate it in a completely
different place. So we click on the project
panel, file scripts, run, and then run the doctor
script here, delocator. Okay, this will ask you how
many layers you want to do. Let me give you a
presa. It's pretty performance friendly
and still does the job. So let's go with 256 compensations
total with 128 keys. And let's make one null and one solid as a filler to
create more confusion. And finally, let's make every single name 64
characters long minimum. Okay, this process will
take a couple seconds, and we'll create a
custom project tree for this purpose.
Alright, here is. So at this point, first of all, we need to hide our
animation source. So there is a specific comp with a underscore on its name, so we need to search it like this underscore and double click on that special
comp like this. So first of all,
we need to shift all the animation elements
inside that special comp. So we shift all the
animation there, right? You can either interior, you can shift your
compositions also here, but I recommend placing just
the bare layers if you can. You know, just make your comp
just as a list of layers. That's the best way to
make this work, alright? So now let's let this thing
here. So we can clean it. Alright. So that's our
source of animation. This special comp
like this inadeSqare. And you can continue to
make your work here. That works, you
know, fully perfect. When you're done,
we need to then give the user the ability to customize some
parameters without ever accessing this
camp, alright? To do that, we can go to a special composition
called user. Okay. And here, there's
already one connection, one example connection
done for you, you know? So right now this layer
has the position animated. By the wiggle into this
layer on the source. You know, the source
has just a wiggle, as you can guess here, right? So there's kind of a long
distance connection, you know, and this works also
with keyframes, right? So keyframes expressions here are being read
automatically here. But you choose this
method for the reference. Don't use a direct cable. Otherwise, it will just
reveal the name of this comp. You know, instead,
use this method. You just copy this entire thing like this and then
scroll all the way down and simply
change this, this part. Just the layer number, you know, the layer number and
the source composition and the specific effect, property that you want to connect or animate,
right? That's it. You know? So here, you can just set up your sliders, right? And then just with this specific type
of reference method, you just change the
layer number and the specific property
that you need to connect. For example, let's
suppose I want to get the scale value of
the second layer here. So simply go to the user and
scroll all the way down. Actually, we copy
this code, right? Copy this entire code. We paste it into the
scale property here. Let's past it on a
new layer here, yeah. Okay? So let's paste it on the scale property,
this new layer here. Okay? And scroll all the
way down, obviously. So if you remember,
what's layer number two? So layer two and
transform scale instead, right? Scale. Like this. All right. Boom. And now it's copying exactly this
layer scale. Let's see. Nine, nine. Alright, so
it's fully connected. Let me show you 48, 48. H. Alright? That's it. So here, just set up your
sliders for the user to use, but they will never directly
reference to the source. That's the point of
all this, right? Okay. So then let's suppose
that the guy is very bad. You know, he wants you
to steal your code. Alright. Okay, so let's
make her laugh even harder. So in that case, we need to obfuscate the code even more
by adding some garbage, so they will not be able to
find that easily, right? So to do that, we need to simply select all the comps
involved in expressions. So in this case,
user and source. So we can do this in two steps. So we select first the
source composition. So let's reveal it here, reveal, select it here,
okay, at this level. And then go to the second
script I sent you. It's called scrambler.
So you go here. And then just double
glicon scrambler here. And it will tell you, Okay,
or even kind of warning. Okay, do your
experiments. Alright. Just read this
carefully. Okay, that's set for now, let's go safer. Let's go with 1,000
iterations of noise. And each one needs to be 200 characters long,
right? Very big. Okay, let's wait
a couple seconds, and now this all the elements
that have expressions or layer references will automatically make those become
anonymous fundamentally. My take about 30 seconds
to 1 minute depending on how many layers are
using expressions and are referencing to
something else, alright? But when the process is done, this is how it looks
like fundamentally every single expressions becomes so heavy that it becomes very discouraging for the
bugler to steal your coat, because as soon as they
start to open one of these, Okay, right now, we just
clicked and it's lagging. It's lagging like crazy. There's
an extreme wall of text. You see what is happening,
thousands of lines. And right now, I have issues myself to
see where the code is. I don't know where the code is. Maybe I carefully
scroll, maybe I find it. But you see, it becomes
invisible. That's the point. So it kind of adds thousands of lines of garbage in order to, um, simply hide your
precious stuff. That's the point. So same
thing on every single layer. You know, every single layer
now is completely ruined, completely destroyed, you
know, let me show you? And we used only
1,000 iterations. If you push it up to 9,000, do a trade off with
this, you know? But the point is that, okay,
right now, I just clicked. I just already taking forever to open because it's
completely blocked. You know, I don't know
what the code is. It's crazy, right? So, this layer originally was
referencing to this layer. But now there's no way
to figure out what it's doing because it's
completely destroyed, right? And the point is that all
of this is still working. Just remember to
close the expressions and the performance
will return normal. Wow. You know, that's it.
Everything is still working. So the issue starts
only as soon as you start to open the express tabs, again, you know,
but that's good. You know, we need to make the buglers life harder, alright? But when you work, it's fine. And do this only at the very end on a copy of the
project. Remember, right? Okay. When you're done,
remember to simply remove the underscore from the source animation
composition, right? Otherwise, people will be able to find it easily.
So remove it like this. Okay. And then just go
back to the user comp, open all the expressions, and fix the change in the name. So you do it here
simply removing underscore plus
here. Boom, boom. Alright, so it returns to work, and only then you
can scramble also the user composition so
we can finally hide this. Despite this may look confusing,
it's still crackable. So we can hide it by also
selecting finally user. Do and we scramble it. We go here, scripts,
run, scrambler. And yeah, let's go with 1,000. Let's go with 128 catchers. Else let's see. Alright, that's it. You know. And yeah, so that is
the full workflow. So as you can see now
it's very slow because it's completely
hidden somewhere. You know, here, you know,
there's no way to find it that easily. And, okay.
76. Plugin and Scripts Recommendations: Another amazing website that I recommend exploring
is ascript.com. So this website allows
you to download scripts, which fundamentally are mini
plugins that allow you to make complex tasks easier
and easier way, you know? So just explore this website. There are so many crazy things. For example, there's
one called Type monkey, in case you need danced
kinetic topography stuff, type this type monkey. That's one of my favorites. Okay, here, that's
another crazy plugin. Just search by category. You know, according
to what you need, go to Afrofax here and just explore the categories
according to what you need. You know, if you do a lot of
animation with characters, go to the rigging tools. If you need glitch art, go to the glitch category, et cetera, you know, there
are so many crazy things. If you work with
three D objects, or if you work with cameras, anything you can imagine
is in Digi's website. Alright, finally, if you want some recommendation for
crazy versatile plugins, I definitely recommend the
red giant trap code suite, and along with the sapphire
suite, let me show you and then also to have a super
CPOfficien motion blur, use the RSMBRal
smart motion blur. And finally, for
frame interpolation, the king is Twixter, right? So with these four
things, fundamentally, for 90% of people, with these plugins, you're
gonna be fixed Ferrara. You know, you go to have
so much versatility at your disposal and
your projects that you can do what you
want. That's the point.