Transcripts
1. Promo video: Welcome to this course
where I'll teach you how you can make
use of Midjourney. I'm faisal and I will be your
instructor for this course. I'm an entrepreneur with
passion for online teaching. Midjourney is a Ai tool
that can help you generate images that you may need for your project or to match
any of your requirements. You just have to tell Midjourney what kind
of image you need. And Midjourney will do the job of giving you the right image. In this course, we will learn
how you can use Midjourney. We will begin by covering
the basics like what is Midjourney and how you can set up your
Midjourney account. Then we will understand what are Midjourney prompts and how you can create your own prompts to get the desired results. We will then understand how you can Tweak your prompts to get a better control over the image that is being generated
by Midjourney. This is a complete practical
hands-on course where we would be taking a
look at lots of examples throughout the course, you will learn
different techniques to fine-tune your prompts so that you can get the image
that match your requirements. By the end of this course, you are going to have a
complete understanding of how you can make use of Midjourney to elevate your creativity
and productivity. This course is designed for
learners of all levels. So whether you are
working professional or whether you are a
student at university, or if you are someone who
is just looking to explore this AIB stool Midjourney to generate some amazing images, then this course can
help you with that, with the knowledge
and insight that you will gain from this
particular course, you can use it to elevate your productivity and
creativity at your daily work. So let's get started and I'll
see you inside the course. Thank you.
2. Getting Started With Midjourney: Hello everybody. So let us start exploring what is Midjourney. You're on Google. I'll search for Midjourney. And you will be shown
this particular website, which is the first website
in the social results. So this is the official
website for Midjourney. So I'm going to click on this particular link here,
which is midjourney.com. And you'll be taken to this particular landing
page over here. Now, this is the official
website of Midjourney. So what is Midjourney? Midjourney is nothing but a generative AI software program
and it generates images. Now how does it generate images? So these images are generated from natural language
of the humans. So Midjourney, you
have to give an input, and this can be a description about the image that you want. For example, if I
want an image of a cat besides a coffee table. So you can give
this as an input. You can specify
different things and as much of details as
you need in the image. And Midjourney will create an image for you based
on your description. So what are you doing here
is you are giving input to this particular software program in a natural language of humans. And this is the language in
which we humans communicate. This is taken by Midjourney and the output is a generated image. And the input that you
gave is known as prompts. Here on the website, you can read more
about Midjourney and understand like the team
members and what it is. So it's an independent
research lab. Midjourney is the name of the company and it is also
the name of the software. So you can go to this and you can understand the
team and everything. So you can also see over here, like it's so very small team, so it's 11 full-time staff. So yeah, this is
about Midjourney. Now, you can click on showcase and check out different
images that have been generated with this particular
software you can explore. And this will give
you a just about the ability of this
particular software. You will know what ability
the software has. Alright? So you can do some
exploration. We're here. So now here on this page, I'm going to hit Sign-in. Now, the moment you hit Sign-in, you will notice that you
are being redirected to another website,
which is discord.com. So this is a separate app which is being used by
Midjourney to function. Now what is Discord? I have a tab open way
here called discord.com. So discord.com is a place where you can chat,
create communities. It's like a chatting
application. And you have different
communities that you can join. You can do one-to-one chats, or you can do
community chatting. Also. This is essentially
instant messaging tool with some social abilities. So you're on Discord, you have Discord servers
without you can create. So we'll take a look
at this one shortly. You will be redirected
to discord.com. If you click Sign-in
on Midjourney. And if you don't have an
account like a Discord account, you can create an
account by hitting register or if you
haven't account, you can login to discard. Okay. I'm just going to
create an account on new account because
I wanted to show you the fresh
onboarding as well. So I'll fill in this information
and I'll hit Continue. Now, once you have
signed in or sign up, whichever way you choose
in the earlier step, you will be shown this
particular page over here, which is in authorisation beach. So like I said, Midjourney operates with
the help of discourse, and discourse is a
separate application. Now within this course, you have a concept of boards. You can have bought and you
can add different boards to your chat and you
can work with them. So Midjourney is a bot
that exist on Discord. So it's asking me that an external application
Midjourney board wants to access my
account and I'm giving access to all this
inflammation, which is fine. So I'll say authorize. So just click
Authorize over here. Okay, so you might get
an error if you don't have a verified Discord account. Okay. So I'm just going
to go to my email. So since I've just
created the account, this account is
not yet verified. So I'll just verify
this account by going to my email and just
hitting the verified link. So I have verified my email. So I'm going to hit dismiss and I'm going to sign-in
again into Midjourney. So I'll just go through
the process again because my e-mail was not verified and it feels so I'm
just authorizing it again. And you will see the
homepage like so. And you will see a sidebar
on the left-hand side and option to purchase a plan on the right-hand
side over here. Sure you have a link like join the discord
to start creating. So to generate an image, you will have to join
the discord channel. Now if you click on
this link over here, like join the discord
to start creating, you will be given an invite to the Midjourney
discord channel. You can see there are
quite a few members here. I'll accept the invite. And it will take me to the actual Midjourney channel or actual Midjourney Server. You can see over here, the official Server
for Midjourney. Alright. Here you're going to see
everything that is happening. So sure on the home, you
can see like everything that has been happening
across different channels. You have different
channels over here. And if you're a beginner, you have to go to this
newbies channel like so. Here you are going to see what images like
everybody is creating. So you can see like they've
given this as a prompt. And you can see this is the output that
they are getting here. Okay? You can see this is
the image that has been generated by someone. So you can just play
around with what is happening in the
world of Midjourney. So yeah, this is about discord
and how you can set up your discard account to get
started with Midjourney. So here you're
going to see all of this chat and new prompts are
coming every now and then. This is very active. And you have one more new BCE, Rome as well. Here, one more. So you can see over here, these are all the other
images that have been generated by users
around the world.
3. Plans and Pricing Options with Midjourney: So let us try doing something
in this particular channel. So what we're going
to do is we are going to try writing a prompt. So I'm going to see
the message box here. I'm going to see slash. And I'm going to say, imagine, this is the prompt
that you need to use. This is a command, I should
say, forward slash, imagine. And here are you agreeing to see this change in dot text
box sort of thing? What this word prompt. Now here you need to write in the description of your image
that you wish to create. Let us say I enter
astronaut going to space and I just hit Enter. So now when I hit Enter, you will see a message that terms and conditions have
not been accepted by you. So I need to accept the
Terms of Service or you can go to this particular link and read more about the same. Just accept this. Okay? So you see a
message over here that due to extreme demand, we can't provide a
free trial right now. I'm not subscribed user with
this particular account. And Midjourney
offers a free trial. But right now as I'm
recording this video, they are not giving free trial because of the
extreme demand they have. So what do we need to do
is we need to subscribe so you can enter this
command slash subscribe. So I'll have to say
slash subscribe. And you'll say Enter. So you will see a link to go
to the subscription here. So you can see open the beach below to pick your
plan and subscribe. So you can click on this like and it is going to show you a pop-up that
you are leaving Discord. I'll say visit. And it's asking me for some verification and I'll
be taken to midjourney.com. This is one way of going
to the payments beach. Alternatively,
what you can do is within your Midjourney
account, within the login. Here you have an
option to purchase a plan so you can
even click on this. So the moment you click on that, you are going to be shown the subscription page over here. You can see different
subscription options here, like the basic plan. You have standard plan
and the blue line. And this is yearly billing. You can have monthly
billing to sure. You can see what's how it works. So here you have
limited generations, like around 200 a month. Here you have general
commercial terms. You have access to
member gallery, and you have three
conquering fast jobs. And you also have, like in the standard plan, you have 15 h of
first-generation. You are being billed as to how many r's are
first-generation you are using. So if you cross 15 h, it is going to generate
images a little slow. Alright? So this is how the plan works. You can subscribe to
any of the plants. You can subscribe to a
10-dollar basic plan as well, which it should be fine. And if you have a free
trial or if Midjourney is offering free trial when you're watching this
video, it's well in good. You can even use the
free trial motions. So freedom is enough to
learn about Midjourney. This is about the plans and subscriptions that
Midjourney has to offer us
4. Creating Your Own Server: So we have signed up for
the Discord account. And you have this
Midjourney silver open. So you have joined us
Midjourney silver. Now here you can interact. You can start typing
in the command. So in order to
generate an image, have to see forward
slash, like so. You're going to see
different options here. Okay? So put forward slash, you can say imagine. You can specify your
description over here. So this way you can generate
the image over here. But the thing is, we won't be
generating images this way. So, sure, in this newbies channel,
there's a lot of noise. So everybody is generating
their own images over here. Okay, so there are lots of images that are being generated. Instead, what I would
recommend is you should create your own server and get the bot added to
that particular server. And you should work over there. So there you will have your own images only
and not a noise. Like you can see over here. There are different
images that are coming from the entire world. So if you do that, if you add the boat
to Your Own Server, you will only see
what images you are generating and not the
image from everyone else. So sure, on the left-hand side, you can see this plus icon. On the left hand sidebar. You can see ADA. So I'll just see at a Sovo and it'll
show me some options here. I will say create my own. And I'll say for
me and my friends. And I'll say vessels Server, you can give whatever
name you want. And I'll just hit Create. So the server is created. Now you will have this general
channel appear over here. Now what I would recommend
you is you should go to the Midjourney channel. You should click on the
Midjourney part over here. So we need to add this
board was Server. Now the moment you click on
the Midjourney board here, you're going to see
this option here, act to the Server. So you can see at to Server. And it will ask for
some authorization. And it will also give
you a drop-down, which will show the list
of servers that you have. I'll select the server that I
have created here, like so. And I'll say continue. And it is going to ask
me like for the axis. And I'm going to authorize. It's asking me for
some verification whether I'm a human,
which is fine. So I'll verify and I'll
go to my Discord server. So here you can see like the Midjourney board has
been added to my silver. This is how you can set
up Your Own Server and add the Midjourney board
to that particular. So now you will only have
you working on Midjourney. If you want some inspiration and you want to see what
others are doing, you can of course,
switch over here to this particular channel
over here or Server. And you can see what's
happening around the word. Alright, so we'll switch
back to our Own Server, and that's how you can
get the ball rolling.
5. Let's Create Our First Image: So let us start creating
our first image. So what we're going
to do is we have headword to the server
that we have created. And here I have a channel where I'm going to
write a Prompt. I'm going to see slash. Imagine. Let's say
I wish to generate an image of a cat that is
peeping outside the window. So I will see a cat beeping
outside the window. And I'll just hit Enter. This will take a little bit of time to generate the image. And you can see the
progress also over here. So you can see it's generated
15 per cent and it'll take some time to finish the
Generation for this image here. So here now you can see the
Image Generation is complete. And if you click on this, you can see the image data
Midjourney has generated. So you have different variations of cat peeping out of a window. Okay, Now if you take a look
at the output over here, you will see some buttons, which is U1, U2, U3, U4, and this one is V1, V2, V3, and V4. So what does this mean? Sure, we have images
that are numbered. So this is the first image. We have second image. This is a third one, and
this is the fourth one. U1 means upscale one, like the upscale First Image. And V1 means generate a
variation of the image one. So we have these four
buttons to upskill any of these four
images, if you like. And V1, V2, V3, and V4 are to generate the variations of any of the images that
you have over here. So if I want to upscale, Let's see U3, I
can click on this. Then you have a
button over here, which is a Refresh
button or redo button. If you click on that button, it will regenerate the image. So it will just send
that command again. And now you will have like the Command going again and you will see the Image
Generation happening again. So the Image Generation
is done now and you can see it generated a
different image right now. And whichever image you like, let's say I like from
the previous option, I like the third one
so I can see U3. And what it will do
is it will upskill this image over here
with the best quality. And like you can see here, you can vary this. You have some options here. You can zoom out to X, zoom out 1.5 X, or you can have a
custom Zoom setting. And you can right-click
on this and you can, or you can just
click on this and you can save the
image like this. Or alternatively, what I would recommend is you
open this in browser. So it opens the full
image in the browser. And you can see the level of details that it has given you. So you have this class
which looks a bit foggy. There are some few
drops of water and you have very nice
details of the cat. So you can save the
image from here and save it onto your computer
for future usage. So this is how you can generate the first set of image
using Midjourney. So let us also explore
these options here. Okay? So we have very strong
and very sadly, if you wish to
generate variation of this image with some strong
elements to be varied, you can use the very strong or if you want the
variation to be subtle, you can use this option. If you click on this,
like I clicked on very strong zygote
this as an output. So you can see the option
applied was variation strong. And you can see this image. There was a little bit of
variation that was applied. So you can see the
glass changed. Here. You can see the reflection. You see everything is forked out and not that clear outside. Here you can see some
marks on the glass. There are some changes
also to the cat. Okay, you can see
some differences in the lake for Style and all. This is radiation strong. Now, if you see zoom out, you can even zoom out
over here from the image. You can zoom out Dweck's
one point phi X, or you can specify a
custom zoom over here. Okay? If you click on Custom Zoom, you will be shown like
what zoom level you want. Okay? So I did zoom out 1.5 X and
I got this result over here. So let's same image was zoomed out 1.5 X, you can
see over here. And then the third option
that we had is burnout. So you have pan out left, right, up and down. So I clicked on pan out up. And if I scroll down, you will see like over here, the image band up a little bit. You can see over here, you didn't see much of a
difference over here. So this will see
original image here. So you can see the difference
of this image here, okay? So these are the
different options that you can play around with after you have
upscaled an image. Okay? So that's how you can
work with Image creation. With Midjourney
6. Importance of being specific and concise: Let us talk about
the importance of conciseness and being specific with the Midjourney prompts. Sure. We have added a Prompt or cat
peeping outside the window, and we have a result over here. So here in this prompt, I have not added
a lot of details. I've just added like a cat that is peeping
out of the window. And I got this result here. But there will be times where in I know what I exactly want. So let's say with this image, I was experimenting and seeing the results to
get some inspiration. But if there are scenarios where I know exactly what I want, it would be beneficial if I give Midjourney all the inputs. And this is where being
specific and concise matters. So being specific and
concise with your prompts helps you get better
results with Midjourney. See, Midjourney is just
a software program. So if you're looking
for something, you need to help Midjourney understand
what you're looking for. The most specific you
are in your prompt. That is more likely
that the image that is generated by Midjourney is going to match
your expectations. For example, if you're looking
to generate a landscape, or if you need a
picture of a Mountains, instead of saying landscape, you could see or be specific here like I want
mountains in it. And if you want sunset, you can specify
that to this helps Midjourney avoid generating
irrelevant images. And if your prompt
is a bit weak, Midjourney may
generate an image that is not matching
your expectations. So for example, here I have just specified a cat beeping
outside the window. So instead, I can
just copy this. So I'll copy this prompt here. And I'll write one more prompt. So I'll say imagine here, imagine a cat peeping
outside the window. And here I'll say
looking at City skyline, I'll specify that
it's night over here. Okay? So it is looking at
City skyline at night. Now I can also specify some more details
about the cat, okay, so I can see the cat is
black with green eyes. I'm going a bit into
a lot more details. And I'm also saying that the city sky line is
lit up with lights. And I want to specify
the mood as well. So I'll say the image
has dark and moody phi. Let us give this as
a prompt and let us see what Midjourney
has to offer us. So here I'm going into
absolutes specifics. I have specified the kind
of cat that I am imagining. I'm also specifying that the city skyline is
lit up with lights. So there is a cat that is
peeping outside the window and the city skyline
is lit up with light. And the image has
dark and moody feel. So we'll wait for this
generation to complete. You can see over here what
Midjourney generated. So you can see over here, these are the four
images that we got. The cat is black, green eyes. And it Exactly took the details
that we had given to it. The city is lit up. You can see like all
the lights outside. And it has a dark
and moody feel. This is a kind of image that I had asked Midjourney
to generate. If you need some
changes or if you want to see some
more variations, you can of course, use these options here.
Feel free to do that. And you can even regenerate
the image originally it all the options by
pressing this button here. But the point I'm trying
to highlight over here is you need to be specific and
concise with your prompts. So this can help you save
time with Image Generation. And some tips for writing specific and concise prompts
would be to use keywords. Use as many keywords that can help Midjourney understand
what you're looking for. For example, here I wanted
to have a picture of cat. I specified cat, and I specified the keywords like
skyline is lit up. Okay, Let up with lights. If you want to add
some more details like whiskers and for you can add
all those details as well. Also feel free to
add objectives. So here I've added
objectives like moody is specified that
the image needs to be dark and moody fee, you can add adjective
like sober and so on. So being specific with
your prompts matter a lot to Midjourney or
working with any software. So Ai is just a computer
program and you need to communicate your thoughts well to this program to
get the best results. So this is about conciseness and being specific
with Midjourney
7. Midjourney Settings New: Hey there, welcome. So let us start talking about the settings that you
have with Midjourney. So to access Settings,
you have a command, so you have forward
slash and you can type in Settings over here. And you can see the
description you view NHS, your post new Settings. So this allows you to view and adjust Settings
with Midjourney. You can see all the settings that exists appear over here. You can see the first
option is of the model, so it's a dropdown. So this keeps changing. So since now we have many
models for Midjourney. This is coming in
as a drop-down, but earlier it was being
shown as buttons over here. Okay. So right now, the latest version or the latest model that we
have for Midjourney is 5.2. So the AI software that
you use have models. And these are the
models that operate behind the scenes to give you
the results that you want. The latest version
right now is 5.2. So you can think of this
as software versions. So the latest version of
the model will be the best. You had 5.25, 0.1, you then have Nietzsche
model as well. So Neijing model is a model that allows you to generate
anime style images. Then you have more versions
of Nietzsche model as well. So you can just take a look at different models that exist. So I'll keep myself
on model 5.2, which is the latest. And whenever any latest
model is released, this automatically changes
to that latest version. Then you have the style
options over here. So whenever you generate
any image with Midjourney, those images will have a
bit of artistic flavor. They will look
little bit artistic. And that is because of this
Style Medium over here. So if you want to have raw images without
any artistic look, you can switch to
roll mode over here. You can change to
Style, stylize low. You can even change and go
for Ohio and over here, okay, So it completely
depends on your preference. Then here you have
public and remixed mode. So public means all your images will be visible by everybody, so they are public. And remix mode allows you to generate the variations
of your images. After they have been generated. Then you have high radiation
mode and low variation mode. So whenever you generate
images with Midjourney, the four images
that are generated have the amount of variations. So this controls the amount of variations that
these images have. So if you want to switch
on low variation mode, you can click on this. Then here at the bottom you have the speed at which images
are being generated. So you can even control that. So you have a turbo
mode, a fast mode, relaxed mode, and you have an option to
reset your settings. So fast mode means the images
will be generated faster. Turbo mode will make sure that your images are generated
even more faster. And relaxed mode, we'll take a bit of time to
generate images. Now with this account, I'm on a Pro plan. And if you go to my
account details here, I'm on a basic plan. And as you can see,
within my plan, I have some number of free
hours of four-stars included. So I have 3 h and 20 min
of fast ours included. And I can buy more fast dots. So you need to make sure you
use this option carefully. If you don't need
images quickly, you can switch to relax mode. So this will take some time
to generate the images. And this will also
make sure that the fast ours from your
account or not being used up. So you can say fast as
foursome important projects. So this is about the settings that you have with Midjourney. You can play around
with the settings and customize them to
what works for you.
8. Modifying Prompt to work with Different Styles: Hello everybody. Let us see how you
can use prompts to control the style of images that are being generated
with Midjourney. So let us see. I'll say, imagine,
let's say I wish to generate an image of a
lion sitting on rock. So I'll say prompt line, sitting on a rock in a jungle. And I'll give this
as an input to the Midjourney pot and let us see what it has to
generate for us. So this is the output
that Midjourney Give me. What I can do is I can rewrite the prompt and get some
different results as well. So here in this result, you can see I have four lines. There is little bit of
difference in the surroundings, as you can see, not much. And that is a difference in the line that we
are seeing also. So here you can
see this line has some red kind of a shadow or
a light falling on itself. And this one has
something different. So yeah, so what I can now see to Midjourney over here is, I can say imagine. And here I can say, I need an image for a lion
sitting on a rock in a jungle. And let's say I need this
image to be short on canon D, and it has to be shot
with a wide-angle lens. So let's say I can say short with a wide-angle lens, like so. And I can press Enter. So this is the output
that I got over here. And you can see this
image looks much more real as if it's
short on camera. So you can specify the camera and you can get the
output this way as well. You can even modify these Style by writing the type of
style that you wish to. So for example, I can say imagine and I can paste the
same prompt here. Okay? Now, instead of like
adding detail about canon, I can just see the image
has to be a cartoony image. So I want this to be
a cartoony kind of image which looks
funny and playful. So playful and funny are the keywords that I'm
giving along with cartoony. And let us give this as an input and let us see what
output we get. So when I gave this as a
response, we got an image. And image looks somewhat similar over here to the first
one that we had generated. So here we had not specified
any of these tiles, but this is what we
had got as the output. When we said we need line
sitting on a rock in a jungle. Then when we scroll down, like if we see line sitting on a rock and jungle
and the image has to be cartoony, playful and funny. It gave us this. So this image is not
clearly what we wanted. So let me like modify
this Prompt over here. So I'll say slash. Imagine. I'll say a
cartoon style image of line setting in a
jungle, playful and funny. So let us give this as a prompt and let us see the output. So I have the output
generated here. And you can see this image
is what is carton style. It looks playful and funny. So you can take a
look at this image. So this one, I like this one, but these two at the bottom
ones are better ones. So you can generate
variations of these. If you're looking
for cartoonists, take Style Images and you
can play around with this. Okay? So, so, yeah, this
is the impact of specifying the Styles on image. You can see like
the keywords and the objectives that you
give in as an input here can actually have an impact on the output
that we generate. Sure, these two prompts
looked similar, but here, cartoony Image,
playful and funny. We're in the end. So I just rephrase this prompt. So does it actually
the same prompt? What I have rewarded it. And I've made it
much more concise. And I've given that as an input. And as an output. You got this as the result. And we generated the
variations here. So the variation is 78%
generated right now. And you should have a
variation of V4 right now. So you can see the preview
of how it looks like. It's 93% done and it's
done 100 per cent. So this is brilliant. And if you're working
on something now that requires some artistic images
or cartoony style images. You can definitely check
out these ones seal. So now what I did
over here is when we said we need an image
to be shot on Canon. I modified this prompt a bit. Okay, So here if you go here. Okay, So after the cartoon one, I give this prompt. Okay, I've given this
prompt and I said, taken on Canon ADD, I need line sitting on Iraq
and jungle wide-angle lens. I got this as an
input and does a slightly different from
what we had generated. Oh Leo. And this looks much
more realistic, so I liked this second one here. But in the line is
looking to the left. And this one is good because there is also some
light coming in from the top. Even this fourth
one is a good one. So depending on
your requirement, you can modify the prompt. I have one more version. So here what I have done
in this prompt is I have specified like a cannon, use S and R6. So this is the model of
camera that I've specified. And here I have
specified the lens. So this is 70 MM lengths. And you can see I have
specified like the same prompt. It should look like I have
realistic light and image. If you click on this, this
looks much more realistic. Okay? So it has some light. As for the description, and it has given us four
variations of the same. So depending on what
you specify over here, the style changes and it tries to match
your requirements. And you can play around
with this a bit. And depending on
your requirement, you can specify what exactly
you need from Midjourney
9. Controlling Aspect Ratio: Hello everybody. Welcome. So let us start talking
about the aspect ratios. So when you generate an
image with Midjourney, you get an image in a
particular dimension. Now let's say you have
some specific requirements as to you need an image in
a particular aspect ratio. Now how can you do that?
So you can do that with the help of the
Aspect Ratio parameter. So what I'm going to do is I'm going to copy this prompt here. This is a Prompt using which
I generated a few images. And let's say I wish to get the output of this prompt in
a particular Aspect Ratio. So I'm going to pass
in the same prompt. And I'm going to see
hyphen two times. And I'm going to specify Aspect. So you can specify Aspect and then you can specify
the Aspect Ratio. So let's say I want this to
be 16 colon nine format. And I'll press Enter. So we'll wait for Midjourney
to give us the result. But this particular input
is going to give us the output in 16 is 29 format. You can even use a
variation of Aspect. So you can say
instead of Aspect, you can see er. This also stands
for Aspect Ratio. And it gives you the image in
16 by nine format as well. Okay, So let me
show this to you. So I'm going to say
imagine AR 16 is tonight. So the Image Generation
is complete. And as you can see, we have the image in the
Aspect Ratio of 16 is 29. Okay? If you click on this, you can see like it covers a lot more details because
now it has lot more area. So this is one version
and then this is the one with the ER parameter, okay? So you can see this also
works in the same way. So you can either specify
hyphen, hyphen, ER, or Aspect, and then you
can add the Aspect ratio. Now if you upskill
any of these images, so I have upscaled the
folder image here. And if I scroll down, you will see the upscale
version of image. Okay? So I can open
this in browser. And you can see over here, like this is 16 is to nine. You can save this
and you can use it in whatever form you want. So you can try and
play around with different aspect ratios that
match your requirement. For example, there are different aspect
ratios like forced to 52 is 23327 is 24, and so on. So you can play around with
different aspect ratio. And if you're working
on a project, you can specify the Aspect
Ratio that you want. So this is about
the Aspect Ratio. So Midjourney allows us to control the Aspect Ratio of
the images that we generate. And we can use this as
per our requirements
10. What is Chaos: Images that have been generated as a result
of the prompt. Now, how do I control the
variation of the image grid? So the output that
Midjourney gives to me is a image grid wet for images. And these images have a
certain degree of variation. And then, as you can see, so Midjourney allows
its users to control the variation of the images that is generated as a
result of the prompt. And you can do this with the
help of Chaos parameter. So I can see slash, imagine. I will say lion sitting
on a rock in a jungle. So this is the exact same
prompt I've typed in. And I can see hyphen
two times Chaos. And I need to specify the Chaos value followed
by this parameter. Now what is this Chaos value? So this Chaos value
dictates the amount of variation that the
initial image grid has. So you can pass in any
value from zero to hundred. So zero means a
lower Chaos value and 100 means a
higher Chaos value. So by higher Chaos, it means that the
four images are going to have a high
degree of variation. That is what this means. As an example, let us see, we generate four images. So I say Chaos 25, I say Chaos 90. And let us generate one more image with
hyphens, see parameters. So Chaos has an alternative
name called hyphen, hyphen C. So instead of Chaos, you
can also specify see, and you can specify
the value here. So let us generate three images and let us
see what this looks like. Alright, so the results are out. And here I have this Prompt Align sitting
on a rock in a jungle, Chaos value of 25. So you can see here, this image has little bit of variation
than the original one. So the original one,
if you take a look, had little variation
like very, very little. So this was not much, okay? Because the default
Chaos value is zero. So if you're not specifying
the Chaos as the parameter, the value is set to zero. But once I set a value here, you can see that there
is some degree of variation because I have
specified 25 for this one. Now, if you scroll down, I have one more
variation generated, which is like the
same prompt lying sitting on a rock in a jungle. But the Chaos value is 90. And since 90 is closer to 100, this has a higher
degree of variation. You can see the variation
amount over here. Like this one is like
he's on the rock and there are like there is a sky and a cloud
behind this line. This one is about to road. I believe he has its
mouth open or something. And then you have trees
covering in the background. This one has trees
and rewards as well. This one has a slightly
different look with the R4. So these are highly
varied images. So if you're not sure about what kind of images you need and if you're just
experimenting, you can have Chaos value, like a custom Chaos value that you can mass with your prompt. And you can see and get
some ideas over here. So this can be used as
a brainstorming tool if you're not sure in which
direction you wish to go. Now I have one more
variation over here that is generated. Width hyphen,
hyphen C parameter. So this was hyphen hyphen Chaos. Chaos can also be
written as hyphen, hyphen. See, over here. You can see the variation
and here I'm passing in 100. So here I pass 90, and here you have 100. Now, you can see the amount of variation that these
images also have. Remember, having a
lower Chaos value will get you similar images. And having a higher
Chaos value will give you images that are highly
varied in the grid. That is Chaos for you.
11. Excluding Objects From Your Images: Hello everybody,
Welcome to this class. So let us start talking
about how can you remove anything from an image? So here I have an
image generated of a lion sitting on a
rock in a jungle. You can see there are quite a few things that have
been included in this image. Now let us say, I don't want the greenery or the trees
appear in the background. Okay. So how do I tell Midjourney
that I don't want trees or any sort of
greenery in the background. So one way is you craft
a prompt and see, I need an image of
Lyon sitting on a rock and don't include trees. So what do you think
Midjourney will generate? Midjourney is a Ai program. And what do you
include in the prompt? It just takes it. So for example, here I have a prompt which says line
sitting on a rock in a jungle. Don't give me reason the Image. But if you see over you, you can see trees
in the background. So Midjourney does not understand the commands
given this way. So if you give any
negative commands of what to include and
what not to include, Midjourney won't
understand this. So by default, Midjourney will
consider any word that you include in the prompt as something that you wish
to see in the image. So the question is, how
do you manage this thin? So you can manage this with
the help of no parameters. I can see here. So let me say imagine and
now let me copy this. So Alliance sitting on
a rock in a jungle. Here, I can just
see hyphen know, and I can see trees. So after no parameter, I can specify any object
that I wish to exclude. And if I wish to exclude
multiple objects, I can specify the list
with the help of commas. So I can say trees. I can see sun. I can say, let's see. I can see greenery. Whatever I wish to
exclude from the prompt. I can specify it with the help
of comma separated values. And these comma-separated
values can be a list of keywords that are Objects that you don't wish to have
in the final image. So this is really helpful if you are trying to
generate an image that match a specific
requirements and they do not have things
that you don't need. So let me show you some example prompts that I
have written here. You can see here I have line sitting on a
rock in a jungle. And the prompt that I've given is with the parameter no trees. But if you see over here, it has removed the trees, but it has greenery over here. It's still okay. So what I did over you is
I added one more prompt as an experiment wherein I add
three as a singular form. So here it was trees. Here it has three. So you can see the result
that I got because of this. You still see green
Objects in the background. Now, what I did is
I told Midjourney, I said line sitting on
a rock in the jungle. I won't know tree
leaves or trunks. So you can see what
it gave me this one. The first result is what I, what exactly matches
my requirements. So it does not have any tree and it rungs or any
such green object. Here also you can see this
one also matches quite a bit, but you can still see
some trees over here. This one is pretty good too. Now, what I did is I gave
one more input over here. This one was about trees, leaves and drunks, and Midjourney did a good
job at this one. But then later on I
iterated on this one. And I said line sitting
on a rock in jungle, no tree leaves, tree trunks. And I also added
greenery over here. So here I didn't have greenery. Instead of trunks eyes. I went more specific. I said tree trunks. And you can see this is like what exactly
matches my requirements. So all four images match
my requirement over here. Now, there is no greenery. There is no, there are no trunks that are appearing in the image. And these are trunks of the
trees that I'm talking about. And there is, there
are no leaves or anything like no
greenery at all. So everything is removed and all the three objects
are also removed. This is like more of a very rock style images that I have
generated like very rocky. He's in a jungle now. So what will you show instead
of trees or and greenery? So Midjourney was
smart enough to add rocks and mountains
behind the scenes. And you can see, this is more of a rustic look as you
can see over here. So this way you can
work with Midjourney to exclude Objects
From Your Images. And this will take
some practice. So I created some variations of the prompts to show you what
exactly I'm doing here. You can see this was
just a general prompt. And then I added the
new parameter trees. So of course there
are no trees here. So if you look from
Midjourney perspective, Okay, us telling me to remove trees, but there are no trees. So it just has greenery. But let's say I don't
want greenery also. I need to be much
more specific there. Okay, so in this one I
said no tree leaves, trunks also to
remove everything. But here I went
even more specific NSAID exclude tree trunks
as well as greenery. And you can see the
image was perfect. So does it, how you can
make use of no parameter? Just keep in mind. You cannot specify
know parameters, a language like
we did over here. So line sitting on
a rock in a jungle. Don't give me trees
in the image. This won't work. So it will show you
the image of trees because you mentioned three
keyword and the prompt. So Midjourney takes any keyword that is specifying the prompt. And if you want, remove
that thing from the image, you have to make use
of no parameter. And no manometer is really
helpful as you can see, if you want to
exclude something. No parameter to the rescue
12. Stylize and Style Parameters: So let us start talking
about how you can control the stylization of images
with Midjourney prompts. So Midjourney is trained to generate images
that are artistic. You can control how artistic
you want the images to be. So for example, you have this prompt here and
you can see this is the default result that is being generated with
the help of Midjourney. Now, there is a parameter
that I can pass. So I can see imagine lion
sitting on a rock in a jungle. And I can see stylize. And I can pass an integer
value from zero to thousand. Now a lower value in this range indicates low stylization. Now what does this mean? So when you have images
with low stylization, it generates the images that
match exactly your prompt. Sure, Midjourney would try
to be less creative on its own and it will try to give you what exactly you
are looking for. But if you have a higher value, enter with this
particular parameter, Midjourney will be a bit
more creative and images will be much more artistic. So let me show you
what I mean over here. So here I have generated an image of lions sitting
on a rock and jungle. Your, I had an image
without the parameter. So here, this is the
default image and this is how the
output look like. Now what happens if I add
the stylize parameter? So if I see stylize ten, you can see I'm getting
exactly what I need. Okay? So this is exactly the line setting on the rock in a jungle. Now, if I make stylize zero, it gets even accurate. So Midjourney tries
to bring the results much more accurate
to what I'm saying. What happens if I increase this? So I took this to 50 and
this is a result that I got. I took it to 100, and this is what I got. 100 is the default. I'll show you what I mean
by default right now. But for now, this is
what 100 Give me, and this is what
thousand gave me. So it thousand Midjourney tries to be a bit more creative. It tries to add more
artistic element to the image and it deviates a little bit from what
we're actually seeing. So if you want Midjourney
to be a bit more creative and come up with
some of it's own thoughts. You can increase the
stylize value, okay? And this particular value
is also part of Settings. So you can access
the settings by saying forward slash Settings. And you will see
Settings this way. So here you can see this
is the stylize, low, stylize, medium,
high, and very high. So these buttons
have some values. So when you say stylize low, it means the stylize
value of 50. When you say stylize
medium, it means 100. Stylize high is essentially
to 50 and very high is 750. Okay? So depending on what you need, you can select from here. Now, default is medium
selected over here, but if you want to
overwrite this setting, we can override by passing
it as a parameter. So if you just want
to experiment with this option for a
couple of images, you don't need to access
setting and change it, but you can override the Settings with the
help of this parameter. And that is why this
parameter exist. Then you also have one
more mode over here, which is the wrong mole. And you can get images that
are not at all artistic. So here you are getting
a little bit of artistic field in Your Images. But if you say row, so let me show you
what I mean over here. Here, I'll get this prompt. Okay? I'll say imagine. And I'll give this prompt, and I'll say Style. So this is not stylize, but this is Style,
and I'll say row. Now when you give this as an
input to the images that are generated are having
less artistic look. So you can see the image that is being
generated over here. So with this row option,
with Style parameter, this images appear with a very
less or no artistic look. So they look like as if you have actually taken a
picture of the lion. But with the other images, there is a bit of artistic
look, as you can see. Alright, so they
look a bit artistic. So this is what rho
parameter helps with. So if you're willing to have some raw images for
your requirement, you can make use of this
parameter and see rho over here. Or you can even control it from the settings
like I showed you. This is how you can work
with this Parameters.
13. Course wrap up: All right, so this
brings us to the end of this transformative journey
with the Ai to Midjourney. We have learned a lot about Midjourney and how
you can make use of Midjourney to generate some amazing images
with the help of AI. We began with the
basics of what is Midjourney and how you can
set up your own account. We created Our First
Image and then we went through the
process of tweaking the prompt to get the
relevant results and how different inputs given to Midjourney can result
in different responses. Now remember, the
exploration does not end over here
through power of Midjourney lies in adapting it to different kinds
of creative tasks. I'm excited to see how you're
going to use Midjourney to make yourself better and
productive at your work. Remember, creativity
and innovation is all about adapting to new
tools and technologies. And Ai is now a significant
part of the Creative Toolkit. Thank you for being engaged learner
throughout the course. I hope this course has not only equipped you with new skills, but also inspired you to think about creativity in
a whole new light. With this class, you will
find a project that you can complete and share happy
learning and all the best
Faisal Memon, Product | Engineer | Entrepreneur
