Transcripts
1. INTRODUCTION: The Tisa death you're about
to witness is completely generated using Google
Omni. Let's take a look. The contents there.
It's always been there. You just didn't have
the right eyes, whether you're 250,000 miles from home or right at your desk. See it, describe it. Watch AI write the rest, even when the world
around you is on fire. What if you could
point at your world, and Google Omni would
turn it into a caption, a script, a post right now? Google Omni. Your content
at the speed of thought. No more camera, no
more crew, just you. Transform yourself into alien, do some action, et cetera, et cetera, et cetera. This is content creation
on a whole new level. Want to learn exactly
how I did this and more. Join the class and let's build
something in syn together.
2. Flow Agent Interface | Gemini Omni: Oh. Welcome back, everyone. So this is a brand new interface of low platform by Google. Now, if you have any projects right here on the
left side, as you can see, now we have a separate
category for images, videos, character
scenes and uploads. We also have AI tools. Now, if you select images, it will only shows
you the images. If you select videos, it will only display videos. And if you select characters, it will only shows
you the character. If you select scenes, it will
only shows you the scene. And if you select uploads, it will shows you all the media that you have uploaded
to the flow platform. Now let's say if you want to generate an image or if you want to use the new feature of the flow platform,
which is agent. Now, this is how agent work. For example, you want to
create a six panel storyboard. Now, first of all, we
need to select agent. And once you select that, we can type any prompt. So I'm going to click here
and I'm going to delete the avatar because right now the avatar is not
working with the agent. We can click here
as you can see, we can see all the details. We can select aspect ratio. We can select number
of variation. We can select image model, and if you are generating video, you can also customize your
video generation settings. Now, once you happy,
you can click here on the save button or you can
click here to go back. As you can see, we
have this prompt. I'm going to click here,
and now as you can see, it is generating my prompt. Now we are waiting for
the final as you can see, it asks us some question. So now we can approve this, and now it is generating
six panel storybook. And now have a look we
have six panel storyboard. Now we can also do
some advanced thing like we can add
this to the prompt. We can select agent. And now I'm going to type, give me three different
camera shot of this scene. And now I'm going
to click here to generate unlike the nanobanana or the normal flow feature, the agent will generate
three variation, three images, three
different camera angle. Now it ask you some questions. If you don't want to
have this message, you can check this approve, do not ask again, or you don't check it and you have to
approve it every single time. I'm going to approve
it, and as you can see, it will take zero credits
and check this out. We are generating three
different camera angle. This is the benefit of using the agent feature
of low platform. And now check this out. We have three different camera
angles of the same scene.
3. Avatar Creation | Gemini Omni: Okay Welcome back, everyone. So first of all, to
create your Avatar, you need to go to
geminigogle.com. Once you are here,
you need to create a new chat and you have
to select 3.5 flash. Now we can click here. We can go to more Upload
and we can click here. If you have already done your Avatar and you
want to delete this, you want to create a new one. So don't worry. We can
simply go to these settings. Once you go to settings, we can go to Avatar and we can
simply delete this Avatar. Now we are here, so
I'm going to delete this and I'm going to click
here at a Selfi Video. Or we can simply close this. We can click here. Try now, or you can create a new chat. You can click here.
You can click here and you can select Avatar. Once you select this,
we can click here. Try now. Now we
have this QR code. You can scan this with
your mobile phone. So let's go to the mobile phone. We are here in my mobile phone. I'm going to scan this QR code, so go to the camera app. Now once you go to
the camera app, you can simply click
here onto this icon. It will scan this QR code. We have this link. We can
simply go to this link, sign in with your Gmail account. If you have already
signed in with your Gmail account,
click here Agree. And now, as you can see,
we have this setup. So I'm going to take my screen now follow the
instruction from the camera app. Once you complete the process, it will upload your Avatar. Now, once you complete
all the process, as you can see, we
got our Avatar. Now we can click here
to use the Avatar, or we can create a new chat. We can simply click here. We can go to more Upload
and select Avatar. Now we can also use the
avatar that we created using the Gemini and we can use that in
the flow platform. So create a new project. Once you create the
project on the left side, we have the characters. We can simply click here. And if you click here
on the plus icon, as you can see, we have the character that
we just created. We have our Avatar. We can also create
the new one by just clicking on
this upload media. You can select your character. I'm going to select this one. I'm going to add
this to the prompt. I'm going to describe
my character and go to the next step. Now, once we have the character, we can simply select the
voice for the character. Hi there. We're
introducing a new feature. I'm going to select this one, and I'm going to add this to the character or we can
simply create a new voice. So I just added
this information. I'm going to also select
the voice. Hi there. Hi there. We're introducing a new feature where I
can read this loud. I'm going to select this one. And I'm going to
save as a new voice. Now, as you can see, we have
this float baby head voice, so we can select this and we
can add to the character. Now, this part of the character is optional,
so we can skip that. Now we can also select the character and
we can click here. I'm going to name my character and simply click
here on Dunn button. Now I'm going to click here and I'm going
to add my character. You can also click here. So you can see your character. You can also click here to see your Avatar because if
you click here on A, you can accidentally select
one of these options. This is the image.
This is also an image. If you look closely right
here in the description, we have the image image, and right here we
have the character, and for this one,
we have Avatar. Now let's select the
character, add to the prompt. We can click here,
select the video. We can click here OX. We can select VO 3.1 it or
we can select Omniflash. It will take 25 credits. Now we can click here on Got it. As you can see, we can
generate ten second clip, and it will take 30 credits. And right here, as you can see, we have selected
Omniflash model. I'm going to also add a tiny detail and let's
generate this video. Now, if you got this
message, don't worry, the video is died in the queue, and now we are
generating the video. So the video is ready. Let's preview the clip. I found some error. The hand is not floating, but other than that, the video is looking awesome. Now with my character, as you can see, we have
some failed prompt. For the prompt, I use Cloud AI, and this is my prompt. Person wear black
sunglasses in Benton style. Then I write person
transforming into four arms. Four arms is a band
and alien, this one. For this prompt, if we
go up a little bit, I use this prompt. Person wear black sunglasses
running away from Alien arm single continuous
shot camera follow person from front cinematic
Chase enhanced prompt. And I just copied this prompt, and I first created an image because I don't
want to waste my credit, so I visualize my clip
before I generate the clip. So with the prompt, this is our clip
with the avatar. As you can see, the avatar
stay consistent in Hall video. We need backup now.
They are everywhere. We also have the
audio with my voice. If we go back, this is the
forearm test with the image, and with the video, this is with the forearm. Check this out. Now, this is an animated
version because as you can see, if we go down, as you can see, I type person transforming
into forearms, as you can see, this
is animated style. So in my next
prompt, what I did, I type another alien, person transforming
into accelerate, Bentanlien realistic. And as you can see, this is photo realistic transformation. And I run some test. This is my test, as you can
see with the nanobanana. I visualize my video first, and then I generated this clip. With Avatar, this is what
we got. Check this out. Isn't this awesome? This is eight second clip. Let's run it again and hadle it. Isn't this awesome? So this is how we can use and create our Avatar using the
new Google Omni.
4. How to Create AI Teaser Full Work Flow : H. Welcome back, everyone. Now you saw the teaser, and I'm going to show you how I generated the teaser
using Google Omni. So for the script, go to the Cloud AI, and I have typed this prompt. Person doing AI content
creation with Google Omni, give me teaser script, and we have the script. As you can see, this is the exact script
that I have used. And after that, I type now person in
different places like moon Mars Battlefield with flying bullets and
explosion nearby, et cetera, working
with follow camera from front and
perform whole script. I want the person to
perform this whole script, and now we have whole module. As you can see, this is a
prompt and this is the script. The contents there,
it's always been there. And if you want to
go to the next, you can click here for the hook. Now we have the battlefield
and we have the ending. I have used this whole script. So let me just go to the floor. Before I have generated videos, I have generated images so I can visualize my image first, and then I can generate videos
without losing any credit. Now, if you go to the
video, you can see, this is my script, and this is my prompt. And if I play my clip, contents there. It's
always been there. You just didn't have
the right eyes, whether you're 250,000 miles from home or right at your desk. So this was my first clip. After that, I have
generated several clips. So if you don't get the
desired result that you want, try to regenerate your clip. Most of the people
make this mistake. They make a first generation, and if the first
generation is bad, they thought the prompt is bad. No, you have to generate
multiple generation, then you will get to conclusion if the
prompt is good or bad. That is the reason why I
have generated images first, and then I have
generated several clip. After I got all my clips that I have generated
using Google Om, now for editing the teaser, I have used DaVinci resolve, but you can also use Capcut. Now if you don't want to use complicated software
like DaVinci resolve, you want to use a
simple version. You can use Capcut. It's very easy to use. You can cut your clip,
merge them together, go to the captions, and you can generate capture. Now, for my teaser, as you can see, I
use the first clip, and for this part, the lip sync was off, so I use multiple clips
of the same prompt. That's the reason why you should generate multiple clips
of the same prompt. And that's how I have generated this entire desi
using Google Omni.
5. Create Alien Vlogs: Oh. Welcome back, everyone. Now I'm going to
show you how I have generated these Vlogs
using Google Omni. For the prompt, I have used
loud eye. This is my prompt. Person is traveling
different planets and taking selfie with
that planet citizen. Enhanced prompt for
traveling commercial, mobile camera
quality Vlog style. So we got the prompt as
you can see, after that, I have typed star from
Earth person say he's going to travel all planets and take selfie with
that planet alien. Take off from Earth
all the way to the Mars log handheld
camera angle. And as you can see,
we have five prompts. I have tested this prompt, I have generated images first, so I can visualize my videos. Now, after this prompt, I got an idea. When the person
returned to Earth, there is no one in Earth. So I have typed enhanced prompt, logging person alone
on daily routine. And as you can see, the Cloud
A gave me these options. So I have selected this, and I have generated
this prompt. Now I have used
all these prompts. I have generated images, and now let me show you
the log with Google Omni. So this is my first log where I start from
Earth. Check this out. I'm visiting every planet.
Time to leave Earth. Wow, that G force is real. We are floating in space. Mars day one. Say hello. Now we have this log where we have multiple planets
with multiple aliens, and as you can see, I divided my prompt with different
shots like short six. If you scroll down,
we have shot seven, we have short eight,
we have short nine, and if you go to the
Cloud AI, I have type, then person travel to
four other planets and take selfie
with their alien. Take this out. These
are the prompt, and I just copied these prompts exactly
from start to the end. And here we have the result. Check out this view of Jupiter. The rings are
absolutely massive. This storm is absolutely wild. We finally made it, my friend. Now we have the
second variation. Day one same prompt. It is freezing. It's a bit windy on Neptune,
but we're good. The rings here are just
absolutely insane. Kepler is beautiful,
feels just like home. Now I have used one of my prompt from the start
and check this out. I mean, check out the quality and my interaction with
absolutely incredible. Greetings from Europa. We have this video? Oh, my God, you guys won't believe what is
happening right now. This is absolutely incredible. Look at him posing for the vlog. We also have this wall. This is just incredible. Now I also ask Loudy now person travel to Mars doing a
Vulg Wes Anderson style. So we have Wes Anderson style
and check out the result. Mars is remarkably symmetrical. The grand Martian lobby
is quite splendid. The local diner serves
an adequate milkshake. These canyons are
exceptionally flat. Wish you were here
or perhaps not. Now we have the whole
story where the person is returning to Earth.
Check this out. It is sad that he's leaving
all the species behind. And going back to the Earth, and now when the person come
to Earth, there is no one. He is alone. Have a look. We also have second variation
of the same prompt. And as you can see, we
have glitch in this video. Now the best part of generating multiple variations of the
same prompt is that you can use a little bit part of the variation and you can
merge two variation together. So, for example, I can trim this part where he is
looking at his phone, and I can place this part where he's
looking at his phone, and I can place them right here. So we can cut from here and we can add where he was
looking at his phone. We can merge it
between this part, and then as you can see, we
will have the whole sequence. Now, when he is alone, this is the daily routine in
the morning or character. Got to keep moving, you know? Got to keep moving out. And now have a look. We
This is final transmission. Where the person is
posting picture. And at the end, we have
this little message. He uploaded all 4,000 photos. Nobody like them. We also have this variation
of the same prompt. Hmm. And this is how I generate this alien encounter
blog using Google Omni.
6. Professional Daily Vlogs: Oh. Welcome back, everyone. Now I'm going to show
you how we can create interesting vlog
using Gemini Omni. So as you can see,
I have my avatar, and I am using
gemini.google.com. Person wear black shades
in whole sequence, making world biggest cake ever. We are finally building the biggest cake the
world has ever seen. This is pure sugar and cream. I mean, check out the cinematic. Everything looks fantastic. Now to do the proper log, I use Cloud AI, and I ask for some ideas. Give me 50 ideas of log, person driving
different kind of car, person making world rackets, person doing unusual things
z are, et cetera, pro. And as you can see,
we have 50 prompts. Now if you don't like
any of this prompt or if you run out of the
prompt and you want more, then you can type give me
50 World rackets idea, or you can type give me 50 more. And as you can see, we
have several more prompts. And after that, I ask for
more and check this out. Person doing a log, person driving candy
car gear, visible, make of candy, gummy
beer car, ego car, and I type this aspect ratio, which is vertical
and I also mention my quality which is raw
mobile camera quality. Guys, welcome back to my
crazy candy carve log. Check out this gummy
bear gear shift. We are actually driving a
car made of lego and sweets. Now I have this
version where I just type person drive
car made of jelly. Instead of multiple elements, I just mentioned one
element, which is jelly. And we have 960 ratio, raw mobile camera
quality, and check it. Hey, guys, welcome
back to another vlog. Yes, this car is
completely made of jelly. I cannot believe how
smooth this ride is. Have a look. Now I
have this vlog where I'm doing a vlog person destroy
the world's biggest cake, 960 raw mobile camera quality. And I also mention my
style this time, Mr. Beast style. And check this out. We are destroying the
world's biggest cake. Oh, take that giant cake. Subscribe for more craziness. Have a look. Isn't this crazy? Now we have this
where I just type this prom which I got
from the cloudy eye. Person drives a car
from each decade, starting with a 90 30s Ford up to the 24 IV,
ranking them live. Check the 30s Ford
is pure muscle. 70s torque is unreal. 90s precision is king. But this 2024 IV takes it all
now I also have this log. If I click here, as you can see, person attempts to break the
world record for stacking sandwich ingredient
layer by layer in 60 seconds
without it topping. Full attempt with
official counter, and I also mentioned
blur all other faces. No female 90 60 raw
mobile camera quality Mr. Beast style. Check
this out the Gemini, blur all the faces. We don't have any female. Stack this massive sandwich and break the world record.
Look at that stack. We've got to keep moving fast. It's starting to
lean, steady now. We did it. Let's go.
New World Record. Now we are here
in flow platform. And as you can see, this time I'm using my picture,
which is this one. And if I click here, this is my prompt 20 totos lineup for an official ten
meter race filmed in real time for however
long it takes, time lapse and real time
cuts, crowd sharing, blood, all other faces, no female 90 60 ombile
camera quality. And this time, I mentioned
fast paced style, add motion graphics of
the turtle race start, add word by word animated
subtitle and check this out. Welcome to the official
ten meter Tortoise Race. This is pure unerated
speed right here. And he crosses the line. Now with the same prompt, this time, I just
edit this line. Add motion graphics of turtle to show their
progress. And. Welcome to the official
ten meter Turtle Race. We have the motion graphics.
This is pure speed. Now in this version, I just changed my style
to cinematic style. Welcome have official
ten meter tortoise race. They are moving at
a blistering pace. This is nail biting action fool. Now we have another log. Beekeepers cover themselves
in 10,000 plus Bs, forming a living suit. This is my basic prompt, and we have all the
other details like in the previous prompt we had I
bring myself in 10,000 Bs. The vibration right now
is absolutely insane. It feels like a massive
heavy blanket of pure heat, and weirdly enough, I
feel perfectly calm. By the way, all of these
prompts are from Cloud AI. Now we have the same
prompt where we have the turtle races and the
ultimate tortoise race. They have been racing
for five whole hours. He actually did it. If you generate more videos
with the same prompt, you will get
different variations. That is why we are having
different kinds of variation. Now we have this log, balloon artists
twist dogs, swords, hats, and flower as
fast as possible. This is my basic prompt, and we have all the
details like we have Mr. Be style and what by what animated
caption, et cetera. I'm making balloon animals
as fast as possible. I need to get this army
approved right now. Wow. That is an
incredible balloon dog. Behold, the ultimate
army of balloon animals. Isn't this crazy? Now we have this prawn weigh the person
drive tank in a city, and we have this detail, which is same as
the previous prawn. I bought a real military tank. What on Earth is that? We are taking this beast
to a drive through. Hey, man, can I get a
double cheeseburger? Now we have this version where the person adds rubber bands
on top of the watermelon. We are at 450 rubber bands. It is vibrating, guys. It's gonna blow.
Three, two, one. Now we have the same prompt, but this time we have
been here for 30 minutes. This watermelon
is about to blow. Look at the rind cracking. It's gonna pop. There it
is. There it is. Yes. How to generate these types
of video with your picture. Without Avatar, all you need
to do is just click here. You can add your Avatar right here and you can click here. You can select Omniflah instead of video light
fast or quality, you can select Omniflash. It will take 50 credits. Once you select all the details, you can also select
your duration. It will also take less credit
if you lower your duration. If you want to generate
for 10 seconds, you can select that and you can type any
prom that you want.
7. AI Animation Video | Gemini Omni: Oh. Welcome back, everyone. Now I'm going to show
you how we can make different types of animations
using Gemini Omni. First, I have this prompt
person fighting robots, shooting them with
gun, stop motion tile. And as you can see, we have
the stop motion style usual. Now, sometime you will
get this type of glitch. But if you regenerate your clip, you will get better result. Now we have this clip, and this is my prompt a
person stars motorbike, bike is not working person. First, he get frustrated, then he starts
screaming angrily, then he smash and break bike with a bat stop
motion animation. Huh. Why won't you just work? Now with the same
prompt, this time, I just changed my style to
detail 80s anime style. C check out the result. Why won't you work? Now I have this version where I just tie
persons star Bike, and this time, I did
not tie motorbike, and we have this result. Hmm. Now we have this one. Mm. And we have this version with detailed
three D Exa style. Next, we have this
animation person sitting in a car,
person driving car, a mysterious van chasing
him from behind, person wear black
shades, 80s anime style. Where did this guy come from? You're not catching
me that easily. And now with the same
prompt with stop motion animation,
check this out. Next, we have this clip, person
fighting mushroom zombies in hospital with gun wear
black shades claymation. Yes. Now we have this clip person outside
in snow feeling cold. Snow on person shoulder, foggy breath, stop motion. And next, we have
this one person, John Wick Style
fight, stop motion. Check this out. Now, if
you have a storyboard, as you can see, I have
this storyboard picture, and I type show entire story in 10 seconds starting
from top left to bottom and right in order
cinematic and check this out. It animate my entire storyboard. Now, this time with the
storyboard and with a character reference
with the same prompt, at the end, I just
type photo realistic. Check out the result. Now with the storyboard prompt, I just type Pizza style, and this is what we got. By the way, if you want to
get the storyboard prompt, you can simply go to
Cloud AI and you can ask for six panel storyboard
animation style, and you can type your story. Now, next, we have
this one person in extreme cold
weather stop motion. The gemina is really good at mimicking the
stop motion style. That is why I have several
examples of stop motion. Now I have this detailed
prompt, and at the end, I have flat vector illustration
style and check this out. This is a bizarre example. Next, we have this
clip person doing a Vlog Wes Anderson Pixar style. This is my first official
broadcast from the field. I have meticulously organized all of my primary instruments. One can only hope that the
weather remains favorable. Next, with the same prompt, I disdain my style from
Pixar to stop motion. This is my daily dispatch. Notice the perfect symmetry. One must remain precise
until tomorrow then. Next, I have this example, person wear black shades
in a stormy weather, heavy raining clay mission. So the clay mission is
my animation style. Have a look. Next in Gemini, I use my avatar, and I type in a stormy weather, heavy raining clay machine,
and check this out. This storm is getting
worse by the second. I need to find
some shelter fast. Now I regenerate my clip, and this time, I have this one. I cannot see anything
through this downpour. This whole valley
will be flooded soon. Now with my avatar,
I typed this brom. A huge centipede appear
from behind person slice it into two pieces with
sword stop motion. You are already dead. So this is how we can generate animations using Gemini Omni.
8. Editing Videos | Gemini Omni: Welcome back, everyone. Now let me show you how we can edit videos using Gemini Omni. So we have a symbol
clip where I just look at my hand and
I imagine a cube, and then I touch the cube. The cube spread into
small pieces of cube. Then I grab one cube in the air. Now to edit the video,
this is my prompt. A hologram cube appear
on person hand, person double tap on the cube, and it spread around the person
with multiple mini cubes. Person then grab a cube
with his two fingers. Now to edit your videos,
what you need to do? You have to click here. You have to I'm going
to select the upload. You can add your
video to the prompt. You can select Omniflash and we can also
select the video. We can type a your video
is longer than ten second, then you can click here. It will trim your
video automatically or you can manually trim
your video like this. If you want to trim
the first part, you can trim like this. Now with this prompt,
check out the result. Now, as you can see, we have so many glitches in the
video, but don't worry. If you regenerate your clip, you will get your result. Now we have this second
variation with the same error. And now check out
the final result. This is my final variation of
this video and have a look. Isn't this amazing? Now, another important
thing is that if your video is in
vertical and you, for example, accidentally select the horizontal
resolution, then you will get
this type of error. As you can see, we have two hands and we don't know
what's going on in the video. So let me show you how
you can mess this up. So for example, you can add your video to the prompt,
and as you can see, my video is in vertical, but the resolution
that I have selected, the aspect ratio is horizontal. So if you want to
edit your video, you can simply just
change the vertical, and then you can
edit your video. You can also look at the
icon, as you can see, the icon is now in vertical, we have the horizontal icon. So this way, you
can tell either you are editing a vertical
video or horizontal. Now I have this clip where I am swinging and I imagine a SWAT. And this is my prom to edit
this video person swinging a metal swat person
outfit changes every second from
samurai to night armor, et cetera, and check
out the final result. Isn't this amazing? Now we have this clip where I just
type person driving car, a mysterious van
chasing him from behind person wear black shades,
and check this out. Now, as you can see, this
is a bad generation, but if you regenerate, you will get your
desired result. Now I have another variation
where I type this prom. Person with black shades, a huge centipede
appear from behind person slice it into
two pieces with SWOT, multi camera shot,
and check this out. We don't have any SWOT, we have the centipede. After that, the video
stop because we have five second duration video. We have another variation. This time, as you can see, we have a little bit
improve now we have this variation with
the same prompt that I showed you before. So as you can see, we have
this improved version. We have a little bit error, but as you can see, we have
this improved version. So if you regenerate your clip, you will get better result. Now, I have this detailed
prompt with the same video, and I use Cloud AI to
generate this prompt. Simply, I just paste
my previous prompt, which I just showed you earlier and I just type enhanced prompt. And the Cloud AI
enhanced prompt, and I use that prompt. And as you can see, we
have this improved result. Now I have this clip
where I imagine extreme cold temperature
outside in the snow weather. And now check this out person
in extreme cold weather. As you can see, my face
is turning into blue. Now with the same video, I just add some words in my
prom person outside in snow, feeling cold snow on person, shoulder, foggy breath,
and check this out. We have complete
background replacement, and as you can see, we have snow on the
shoulder on the person, and we have this foggy breath. I have the second variation. Check this out. So if
you regenerate again, I'm seeing again and again. If you regenerate your clip, you will get better result. Now I have this clip
where I'm acting like I hold a gun and I'm shooting
bad guys, zombies, et cetera. Now, this is a prompt that
I use to edit my video. Person fighting
mushroom zombies in hospital with gun
wear black shades. And as you can see, we have complete background replacement. We have mostro zombies. We have a little bit adder, but as you can see,
the job is done. We have a background
replacement. We have the zombies,
and I'm holding a gun, and we have the black shades. Now, I have another version where I tie person
fighting robots, shooting them with gun
and check this out. Now we have this clip where
I imagine that I am in heavy rain weather and the
rain is falling onto me, and I type this person wear black shades in a stormy
weather heavy raining. And check this out. We have complete
background replacement. We have the rain on
top of the person. We have the black
shades as well. Now with the same prompt, this is my second variation. And as you can see, this time, we have this improved version. I mean checkout the water on top of now I have this clip which
I shot with my smartphone. And as you can see,
I'm imagining that I'm trying to float a camera that
is placed on the ground, and I typed this prom
to edit this video, a DSLR camera plays on ground person trying to make
camera float in the air, then cameras start floating
and check this out. I mean, this is not
looking like the AI. This looks like a real practical
effect. Check this out. One more thing, the Gemini omni nailed the performers of
removing the background. So as you can see in this
play I have this background. And I use a simple
prompt change background to a professional studio, and as you can see, guys. When it nailed the performance. It changed the background, even it completely changed the lighting on me.
Check this out. In this video, I have this
red light on this side, and check this out. It matches the background
light with the subject. I have another clip where, as you can see, I'm
sitting in my studio. We have a different light. I'm moving my head. I'm
doing this hand movement. And I typed this from
replace background to professional studio
and check this out. This isn't even
look like the AI. This looks like a
real life video of me shot in my studio. Have this version where I type chain background
every second, professional studio,
apocalypse, et cetera. And as you can see, we are switching in different
backgrounds. And unlike other
tools, as you can see, guys, it removes the
background with high accuracy. We don't have any fringes at
the edges of the subject. We don't even have the soft edge on this here because
most of the time, if you try to remove
your background, you will get a soft
edge at the here. But as you can see,
the hair is thick. We can even see the
background through this here, even with this chair. So the gemini omni nailed the performance at
replacing the background. Not only this, you can
also add reference, and you can edit that
reference in your video. So as you can see, we have
this, again, the hand video, and I use this hologram
picture and I type hologram appear on person
hand and check this out. I have this version where I type hand control fire
and check this out, the fire casting
light onto my hand, which make it real. Now we have this version
where I type hand transform into multiple species every
second and check this out. We have a little bit error, but as you can see, it matches the light
according to the scene. Now we have this
clip where I pretend to drive a car and
someone hit me, then I look back
and then I drive. So I type this person
driving car side view angle. A car hit him from behind cinematic depth of
now the reason why I type side view
angle because in my previous version in
my previous variation, the Gemini omni switch from
different camera angles. So I don't want to
switch any angle. I just want a single
continuous chot. So that is why I typed
this view angle. A car hit him from behind cinematic depth of
field and have a look. It completely replaced the
background and check this out. Now, it doesn't
matter if your video is 1 second long,
2 seconds long, three second long or
five to ten second long, it will take equal credits to edit your videos
using Gemini Omni. So I try to bypass this and what I did in my
video editing software, you can use Capcut or DaVinci
or any other software. I merge three clips together. As you can see, we first clip where I'm
showing the glasses, we have the second,
which is a wallet. And third, we have the box. So I want to replace the
box with a smartphone. So this is what I did
and it actually worked. And you can also do that
to save the credits. First change glasses
every second, then change wallet to
luxury premium gold wallet, then change box to a smartphone. And if I shows you the final
result, check this out. And now we have the smartphone.
Isn't this amazing? Now we have this variation Have a look. Now with this video,
I actually did the same by combining
multiple clips. I first, we have that, second, and we have
the third clip. And I type replace hairbrush to alien insect person
interacting with, then change camera
cap to gold coin, then change marker to a knife. And as you can see, it actually
done a pretty good job. So this way, you can edit your multiple clips
in one single prompt. We have this final variation, and I really like this one. So that is the reason
why I'm telling you to regenerate your clip if you don't get
your desired result. As you can see,
this version is far more better than the
previous version. And that's how we can edit
videos using Gemini Omni.
9. Create Motion Graphics: Wow. Welcome back, everyone. Now I'm going to show
you how we can create awesome motion graphics
videos that you see on YouTube completely
with Gemini Omni. Now we are going to use
Cloud AI for the prompts, and we can type this
type of prompt. I want to create prompt
for motion graphics. Give me simple three
D camera parallax with ten second duration. Now you can explain
your style right here. After that, you can
type your topic. You can type any
topic that you want. This clip, what I did, I want to make video on animal
versus mammal. So I type with ten
second duration, about explain what is difference between
animal versus mammal. Give me simple one
master prompt. Now, this is where
the magic happen. If you type, give me one
single master prompt, I will give you one single. Instead of giving you
multiple prompts, it will give you
one single prompt that you can simply
copy it from here. And as you can see,
this is a very long, detailed and simple prompt
for ten second duration, we even have the
timeline what will happen inside in
between this timeline. So I use this exact prompt, as you can see
this is my prompt, and this is what we
got from Gemini Omni. What is the difference
between animals and mammals? While there are many
kinds of animals, mammals are a specific group. All mammals are animals. Now you can also
change your style. So I typed this time
a ti flat graphics about explain what
is the difference between animal and mammal. And I got this prom
with a scene breakdown, so I copied this prompt. And as you can see, this
is the prom that I used. Animals are multicolor. Mammals have fur and milk. All mammals are animals. Isn't this amazing? We just paste this prompt, and we got this amazing flat two d vector
animation as well. Now, after you start your chat, you don't have to
explain everything from scratch. You can just type. Now give me with three
D low poly style with ten second duration, and this is the
prompt that I got. So I use this prompt, and this is the
low poly version. Now, again, I type
a different style. Now, give me Japanese ink style
with ten second duration. We have these visuals
and we have the prompt. So I copied this prompt. Now, if you don't get
a scene breakdown, then you can also ask for scene
breakdown of your prompt. With ink style, this
is what we got. Now, this time, I ask for text typographic style
with ten second duration. We got the prompt, and
this is what we got. Now we have this
version where I ask him to now give me three
D motion graphics, and this is what we got. All mammals are part
of the animal kingdom. But not all animals are mammals. Now with this formula, you can just simply change your style and you will
get amazing result. Now we have this paper
flip motion graphic style. We have, again, three D motion
graphics style explainer. We also have a style, which is SAS style, which is a very famous
YouTube style right now. And as you can see, we have this amazing smooth animation. Now in this clip, I
mentioned vox style, which is a very
famous YouTube game. And What is the
different a man god. Animals are any living creature that isn't a plant or bacteria. Mammals are just
one specific class, one specific class of animals. So if you have a
mustyle then you can also mention that
style in your prompt. Not only this, you
can use your avatar, and you can ask for motion graphics on top
of your explainer video. So you can see, I type person wear black
shades in sequence 20 turtos lineup for an
official ten meter race. And after that, I have some technical detail
like nine, 60 ratio, row camera, quality,
documentary style, motion graphics of the turtle
to show their progress. And as tortoises lineup
for the official race. We have the turtle
motion graphics. Unbelievable. Tortoise
number seven takes the lead. Now, I have this version
where we have this pm person standing in a classroom
explaining difference between animal and mammal. Netflix interviewed
style, add infographic. Add word by word
cinematic subset. All mammals as we classified as Notion into the
category of mammals. Mammals are defined
by specific traits like having hair and
milk production. Now I have this prompt person
standing in a classroom explain difference
between animal versus mammal documentary style, add in for graphics, add word by word subs. All mammals are animals, but not all animals are mammals. You see, mammals are
just one specific class. They have warm blood
and they produce milk. That's what sets them apart. Now you can also take your prompt and you
can build a story. Now a type, now a
person is going to another empty class and setting up his new setter
for another lecture. Testing the audio for
the next lecture. Now I type no person
is giving lecture on AI information in
interviewed star. Artificial intelligence
is changing how we work. It processes data
faster than now. Now the reason why we don't have any global essential
understands for it. That is why we got
this simple video. Now, another crazy thing
that you can do with this style is that you
can paste any reference. So for example, I have
this reference which I copy it from
Pinterest and I type, now give me this style
ten second duration. So the Cloud AI gave me a
prom to mimic this style, and I paste the exact prom
that I got from Cloud AI, and this is the
result that we got. Now to take this video
to another level, we can customize this video. Let me show you how.
Now this is what I did. I took the screenshot and I add that screenshot
to my prompt, and I use Nano Brana
too and I type change Spotify to faisan and
change color to red. So I customize the entire
design and make it my after I customize
the entire design, I upload that design to Claudi, and I type this simple prompt. Give me with this style ten second duration
of animal and mammal. So we got this prompt, so I copied this prompt, and I paste it right here, and this is what we got. So this is how you can
do motion graphics in a professional level
using Gemini Omni.
10. Adding VFX With AI: Welcome back, everyone. Now I'm going to show
you how we can add VFX using Gemini Omni. So you can use Cloud
AI for the prompt. I have my picture and
I have type prompt. Give me ten VFX ideas,
including person. After that, I type, give me another ten
with action sequence, and this is a prompt that I
used to generate this clip. With my Avatar, select
it, check this out. Isn't this crazy? Now
I got the same prompt. This prompt is also generated by Cloudy eye
and check this out. I will not let this plane
crash. Everyone is safe now. Now I have this example where I type person
sitting on a balcony, watching a plant grow in a pot, time lab and check this out. Let's see how much you grow today. I knew you could do it. Now I have this example where
I type a huge centipede appear from behind person slice it into two
pieces with sword. And I also tie
multi camera shot. You really shouldn't have
sneak up on me like that. What an absolute
waste of my time. Now I also have
this example where I transform into fore
arm from Bent en. For this, I also use Cloud Eye, and I use this prompt person transform into forearm
bent and alien realistic. And this is a prompt that
I used with my Avatar. Check this out. Then in the same chat, I type person transform
into Diamond Head, bent and alien
realistic cinemat, and I copied this entire prompt, and this is what we got. Mm. Next, I want to transform
myself into wild mud, and this is the result of that. Mm. So this is how we can create
VFX using Gemini Omni.
Faizan Amjed, Video Editor,Filmmaker & Youtuber
