Transcripts
1. CourseTrailer: Hi everyone, Welcome
to complete me. During the guy in this course, you will learn to create
amazing God coach ai powered using MC
journey in just seconds. First, we will guide you through the Creighton
done account form a journey on accessing
the AIR generation tool. Then we will show you
how to use new churn from selecting this title
to adjusting the colors, to use different mediums. Using the journey is easy, but it is tricky
at the same time, there's older their job title
called prompt engineering. So you should know
the magical words in order to get what you want. That's why we will
start with the prompts, discuss about the majority of origins and their difference, upscaling features, constraints, as well as
how to use MC journey. Journey has a very useful
platform where you can search for other people
jobs and get inspires. You will have access to
this platform and learn to use it after this
course, weakness, journey, you can generate a wide
range of AIR pieces from abstract compositions all the way to the realistic portraits. You can even use it to create custom art pieces based
on your own photographs. Rather you are a
professional artists are just starting out. This tutorial is designed
to be easy to follow, reach clear instructions
and helpful tips along the way with the power of
AI at your fingertips, the possibilities are endless. So when you are ready, join us for this course
on mute journey.
2. Quick Start: My journey, you can create stunning artworks
in just seconds. All of these are created
with my journey. And let's see how
we can do that. If you don't already
have a Discord account, you should create
one since it is the only way to use
MIT journey right now, Dan go to MIT journey.com and click Sign-in, accept
the invitation. And during the Discord server, once you are in, you will see these newbies
channels on the left. All you need to do
is to select one of them and type slash. Imagine. Now it is time to write the
keywords of what you want. In my scenario, I will say
Iron Man learns how to code. You could also change
lots of things here using the parameters
like aspect ratio. For this example, I'm going
to use the default settings. Okay, now we have four different image and the buttons just underneath your buttons are for creating the upscaled
versions, e.g. if I press you want MIT Journey provides the first image with better resolution and v buttons are for creating
dissimilar versions. Say I like the first image I want to see more
works like that one. I click the Run button
and here you go. Now with this time
to pick the winner. And I'm gonna go with
the first one again, this time, click U1. And finally we got the
high-quality image. By the way, you can always see unusual artworks from the
mid journey of web app. All right, that was a
quick introduction for me. Journey.
3. Get Started: In order to use MC journey, you should first sign-in. And to do that, we go
to the mic journey.com. Here we see an old
school animation and just continue by clicking
the Sign In button here. This will redirect us to
the discord application. If you already signed into
discord and it is open, you can join the
magenta channel. Otherwise, fill out this form
and sign into this court. If you don't have
a Discord account, please sign up to this court, since it is the only way
to use my journey for now. Alright, then we are inside
of the meat journey channel. Here I suggest you to read all of the guidelines and rules. In this tutorial, you will get used to most of the features. To start off, we need
to find an EBS channel. There are several options which are changing all the time. E.g. right now we have
newbies start te ne be 60.90. Just select one of
them and we can now use our prompts
inside of this chat. This is a public
channel and you can see the feed is
updating by itself. That's because other people
are also using this channel. And we can see the
outcome of the requests. Let's go ahead and
write our first prompt. To do that, you
should type slash, imagine, and click Enter
to accept the suggestion. And now we have the field
here for writing the prompt. I'm going to say Iron Man learns how to code with a computer. Alright, now, the MIT
journey about takes the job and we can
see its process here. It already started and
completed the 30% of the job. We can also check and see the current version
is off to jobs. When the job is finished. This possibilities
from the feed and the finished version will
appear at the bottom. Scroll down, and here we go. Journey just created for alternatives for
coding Iron Man. At this point, you
might be asking for the buttons under the image. By default, meet journey
is not producing high resolution outcomes
for the first prompt. In order to be efficient, it provides four versions
with low resolution, and then you have chance to upscale the one you liked most. You can do that by using these
you buttons, u1 up skills. The first image you to
upscale is the second image. And so we buttons
on the other hand, are for creating similar
outcomes for that job. E.g. if I press B for MIT, journey will create similar
versions of the image. Let's have a closer look at the image and
select one of them. I like the first one most. And I want to see some
similar versions of it. So I click the Run button. And down here you can see
a new job is created. And it says making
variations of image one. Let's wait for it. There you go. I really liked the first one because there are code blocks on the air and also the phase
is much more realistic. Okay, we can click
on the open in browser button here
and get the image. But I also want to
upscale this image so we can download with better
resolution there. So this time collector
U1 button here, which will create a new job. And after awhile, we
are done with this job. Now we can open this
in a browser and use this image for any project. As you can see, it is obscured
and has a bigger size. Don't worry about losing
the outcome or the URL. Because an automatic profile is being created by MIT
journey for you. And you now have a dashboard that you can manage your jobs. Go to MIT journey.com again, and you will see a different
UI will appear this time. From here, you can see
all of your past works, like the one we just created. All you can toggle to upskill
and see the finished jobs. You just want it for upscaling. If you open one of them, you can make many actions, including saving the
image to your computer.
4. Midjourney Bot: The contents journeys
producing is really great, but there's other
little problem. So firstly, this free, it is updating
really frequently, so it is hard to follow. And it is, it seems
like crowded and complicated because everyone is working from this channel. And also maybe you want to keep your
works private, right? So maybe I don't want to share my prompts and art boards
with the whole community. So what we can do is to write, to meet journey boat with
direct message so that it is, It's not going to
be public anymore. Okay, So click on
majority vote and view profile and
then send message. This will bring us to the direct message to
the maternal boats. And we can also prompt, we can also write our prompts
in this channel like this. So let me say this time. What should we say? I will try highly Queen lying
with a motor cycle. As you can see again, our job is approved
and waiting to start. And also I don't see any
jumps from other people. And I just see my own prompt, which is great, which is
simple, which is linked. Right now the job is completed. And we can also use the buttons dislike in
the maternity channel. So we all know we have
all of the opportunities. I'm features here as well. And this is a private space. So I will try to use here
done in the public place. And from now on, I will use this direct
message opportunity. By the way, results
are already here. So we can select one
of them and continue. Alright, one more
thing I want to mention here is to subscribing. So right now, in order to use magenta boat
in a private message, now you should subscribe
to read journey. It was not like this before. So they are changing the model, the business model, and its climate change
in the future again. So it's hard to say if you
can use it free or not. I'm about right now,
you should subscribe. So to do that you can
write Subscribe Collins, journey bolt, and we'll
send you a personal link. It says Do not share
it with anyone. You can just click and see
the subscription plants. Let me bring my Chrome here. So you have opportunities to
build yearly and monthly. Also, I am using
basic plan right now. And one of the most
important difference, maybe at the most important one, is the limit of your durations. Alright, so it is
obvious that you have more generations
on upgraded versions. With free trial
currently you can generate 15 jobs per month. And in the basic plan, it's 200 jobs per month. And with the standard plan, it says 15 h of
first-generation. So what that means with Journey uses GPU power to
process each prompt. When you purchase a
subscription to meet journey. We are purchasing time
on these GPUs, e.g. 15 hour of GPU time, you will have in standard plan. You have our first generations
in proper that our job, the MIT journal about processes, takes about 1 min of GPU time to finish
creating on an image and upscaling and image using non-standard aspect ratios
may take more time. Okay, These numbers are
for fast gyrations. The other option
is relaxed molt. You will be able to
use relaxed mode on standard plan and propylene. And once you upgrade
to these plants, you can create an
unlimited number of image. Each Mart in the relaxed mode. But relaxed smart
is really slow. Jobs and relaxed mode
are placed in a queue to be processed as GPUs
becomes available. And wait times for
relaxed mode or dynamic, but generally range
0-10 min per job. Other than that,
you will have three conquered first jobs in
basic and standard plan, while 12th conquered
first jobs in Pro plan. And also you will have stilled
image malt in per plant, which is a private way
to generate image. Because by default, all of
your jobs and works are public and can be seen in
your book journal profile. I miss you are in Pro
plan on using Tilt lot.
5. Prompts: Until now, we were
using basic prompts, which includes just a
single description. This might even be a
single word or immersion. And advanced prompt
will look like this. You can provide one or
more imagery resources. And the materials
always go at the front of the prompt.
Description again. And you can also pass some parameters with the
virus enters you can control many things such as aspect ratio and model versions. And parameters will always
be at the end of the prompt. Description is the
heart of the prompt. So you should I explain
all your needs? You might ask, this seems
like a long prompt. What is the ideal length? Well, there is no specific
character count on this, but my journey basically says, give detailed but
not too long prompt. You should describe all
the details you want because everything else
will be randomized. Currently meet journey doesn't understand grammar or
sentence structure. So it is always
better to describe what you want instead
of what you don't want. If you want the
forest myth journey to not include something, you can use a parameter for that and I will
explain it later on. Details are the key of using
my journey effectively. Here are some examples from
MIT journey documentation. Let's say we're big cat
as the subject than the median will specify the
artistic style of the image. Here you can see some examples. In terms of color. You have a wide
variety of options. Some examples, again, are
taken from my journey here. And you can do your own colors. Notice that these are really
specific color types, and I strongly
suggest take notes here so that you can use
for your own prompt. And finally, here we see
some environment examples. This is not limited
by anything and the environment can be anywhere
that you can describe. Okay, so now it is time to put this knowledge
into practice. Let's go back to here we are
back in the Discord again. And I'm going to type. Imagine. This time. I will choose subject as CAT. So acute Yeti. And for medium, I want an illustration. Illustration in Arctic forest. I want to be in forest. Moving forward. I want a neon lighting style. And finally, maybe we can
try green tint color, dislike in the example. We can define more
and more details, but this is enough for now. So we are here testing
many details, right? In terms of lighting, coloring and why you want the meat do and
also the subject. Also, I added motion here. So let's see the output. Alright, my journey,
I finished the job. And here are the results. I think they really look cute. And these are really great
looking illustrations. Alright, so let's try
changing something. So this time, maybe we
can say imagine again. So I will keep the rest. Same to see the difference. So again, again to
deity illustration in Arctic forest, Neil 19. And this time let's say, colorful instead of
green tinted color. So it is really obvious
the image are green, and this time we want
other colors as well. So dot-com should be colorful. So let's hit Enter
and see the result. Alright, so that is weird that our first image
looks like the older ones, but the royals are
colorful intestine. So this is, this is how we
can change the details.
6. Midjourney Versions: Milk journey routinely releases new models to
improve the product. Different models have different characteristics
and you can choose the version you want using
diversion parameter. So e.g. you can write dash dash five
at the end of your prompt. And that will produce
image using version five. If you want to see or default version and change
your default settings, you can use the slice settings
command in this court. This will show you
the preferences and you can make changes
using these buttons. You can try older versions
by switching here. And using the newest model
will give you flexibility. Especially about upscaling. We just talked about
subscription plans. If your plan is
supporting these models, you can disable the public
mode and work privately or change the relaxed mode
in order to save GPU times. There is also a popular version which is called Neijing model. This is a collaboration between MIT journey and spell brush. And the purpose of
this model is to produce anemia and
illustrative styles. It is good for actions shut and character focused in
which in general, and in order to use it, you should enable it here. Or you can use dash,
dash g parameter. It is important to note
that different versions may have different
performance and results. So it's worth experimenting
with different versions to see which one works best
for your specific use case. Let's choose ninja model and
try the same prompt again. These are the results
of the museum model. Finally, if you change the
default model from settings, don't forget to change it
back because it will use that model for the rest of your work until you
change it again.
7. Upscale: As you already know, we have upscaling buttons and
once we click on them, meet journey increases
the image size, vile smoothing and refining details that aren't multiple
upscale models available. And you can see the
difference on the table. Right now the default
version is four and the default upscale
produces 1024. By 1024 units. You can go up to 2048 by
2048 with Beta up scalar. If you don't want
too much details, but still a beak size, you can use light up security. You should specify this value
of writing the prompt and then the upscale
their buttons will know that you want
a light upscale. Let me show you the example. I use the same prompt with applied parameter and the
results are like this. When I click you for this
is the upscaled version and this is still tend to
end the four by 102.4. I honestly, I don't see much
difference on this one. I also want to talk
about Beta up scalar, since it is the
only one currently we can get better resolutions. In order to have 2048. By 2048 image, you can use dash, dash up Beta parameter. Remember this will take
more time to produce and costume or GPU time
on your subscription.
8. Platform: Before going further,
I want to talk more about the meat
journey platform. As you already know
in the homepage, we can see some of
the past works. If you want to see
all of your jobs, you can click on
the Archive here. Let's go back and select
one of the image. We can see the prompt down here. If you scroll a bit, you will see some
more information and even the parent of the job. This is useful because it's also possible to revive an old job. And I will explain
that later on. Okay, moving forward,
you can also see similar verbs from
other people down here. Let's select this cute one. And here you can like to image. And again see related enrich
if you want to explore more. But what I really want to
mention here is the prompt. When you separate the verge
with commas in the prompt, you will see they are
categorized in the platform. So if you click here, you will see some
image similar to that individual prompt, e.g. these are happy
atmosphere pyramids, and these are different artists and all of the verbs
in that style. So this is a really good tool to explore some new styles
and make research. If I again select the first
phrase and say search more, it will redirect to
the Explorer page. And you can see that
phrase in the search box. This is really,
really helpful tool for learning the prompts
from the community. But remember, all your works
are also public to displace. We could also come this page by clicking the explorer
button here. And this is just like a
normal search engine. Which means if I surge, say Rihanna here, I can
find some MIT journey. We handle bucks. Because those are asked me journey of course, but this is much faster. And please be careful here. Although I couldn't
find any written rules about commercial usage of
other people's smokes. In the Discord community. They are saying that
you should always ask before using
someone else work. I believe this will be
regulated in the future. Besides, you don't have to
download the exact image. You can just copy the prompt
and generate your own inch. The results will
not be the same, but it will be very similar.
9. Commands: I will not go into full
command list in this video, and you can check from the
mid journey documentation. Instead, let's quickly
recap the most used ones. Firstly, I want to
mention the info command. You can simply write slash info, and that will show
you some information about your account, such as your visibility mode, remaining time, and cute jobs. I use this account only
for this tutorial, so I still have
more than 200 min, which means approximately
200 emails. We already talked
about the settings command that will show
you the settings, an option to change them. You can switch to the
relaxed mode like this if it is available
for your subscription. Likewise, you can switch
back to the fast mode. And notice here it
says you can purchase more fast hours
from the website. Moving forward, I want to
show you the blend command. It allows you to upload two to five MHz and merge
them into a new image. This is actually same with using the imagined comment and
giving image inputs. But it is more useful in
this way to upload photos, especially on the mobile
phones in terms of UI UX. Also, you can not write
texts prompts in blend mode. It only accepts, emits and
meet journey recommends uploading the image that are the same aspect ratio
for the best results. Here are the two
pictures I found from the maternal platform and I will use them for
the blend example. While you can already
imagine that portrait. So does meet journey. And that's because this is a
great example for blending. The blending job
is finished now, and this is the outcome. Journey does a great job, understanding the image concepts and merge them in
a meaningful way. We can achieve this by
giving clear purposes. This is only one
approach of using Blend tool and there are
much more use cases. Next up is the show command. We use the show command
to revive my job. In order to do that, you need the job ID
of that project. Let's assume we want to revive the job we did at the
beginning of this tutorial. If you remember, we asked
for recording Ironman, and now I need that jump again. Firstly, I should
find the job ID. You can find it
from the platform, go to archive and find
that job. From here. You can copy it, the job ID and paste it here
into this court. After awhile, the jump is Revit. And here again. Now we can again use the upskill
or version buttons.
10. Parameters: In MIT journey,
we use parameters using dash dash at the
end of the prompt. Again, you can check
the full list of available parameters
from MIT journal Docs. Let me explain my
favorite haunts. First, aspect ratio,
we can use dash, dash aspect or dash dash AR
to decide the aspect ratio. And aspect ratio is the
width to height ratio of an image that is expressed as two numbers
separated by a column, such as 16 to nine or four to treat the default aspect
ratio is one to one. Good News series,
we can now create any aspect ratio image
with the version five. With the default version, it is possible to make
16 to nine units, which is typically for desktop. And the opposite for mobile
apps like Instagram. Right now, modernisation
aspect ratios greater than two our experimental and may produce
unpredictable results. But I believe this will be
fixed in the near future. Let's quickly see some examples. This time I will write a
longer prompt and maybe I can use this outcome for
the thumbnail of the video. I made the research on the root journey platform and find some concepts that I liked. Here is the prompt I'm going to write on at the end of it, I will just say dash, dash AR 16 to nine. As you can see, this is a landscape image
which is great for computer wallpapers or
YouTube thumbnails, etc. You could also use the same
prompt for our portrait mode. So this time I'm going
to say nine to 16. Ina, as you can see again, we can use these images
for Instagram, e.g. okay, So this is the
aspect ratio parameter and it is quite simple. And I believe all of
the limits will be gone in the future
in this topic. Next, the quality parameter. This parameter
change how much time is spent generating on which higher-quality
settings take longer to process and
produce more details. It accepts the values
25%, 50%, or 100%. The default setting
is already one, so you cannot increase
the quality more. You can only decrease it. This is great for exploring
new ideas and testing. E.g. if we use dash, dash to that 25, we will be forex faster
and use less GPU minutes. This is the comparison. I think given the worst
quality is enough. And by the way,
this parameter only influences the initial
image generation. So you can still upscale this
image with high-quality. Alright, for the next parameter, we have a new background. And this is not
without annual reason. That's because we will
be talking about tile. Tile is a powerful feature
of meat journey for creating patterns,
wallpapers, and textures. Here you can see some test cases of mind before finding this one. To create patterns like this, you can just write a prompt like normal and the tile
parameter at the end. Notice that title is not working with version
for currently. So we also need to
say dash, dash b5. Let's create a pattern from
knives nc.com. There we go. And really good textures can be created like this
for game developers. You can also use these patterns to blend with other image. E.g. I'm going to say, imagine copied the
link of this image. Then also go all the way
up and copy this one, the one we generated before. So I'm going to paste here and the result will be the
blend of two emits. And again, myth journey
did a great job. At this point. I must say it is not acting that
accurate always. I am only including the project
that it performed well. Many unsuccessful attempts
happened during this tutorial. So if you can't get
what you want in the first attempt,
don't get disappointed. The last parameter I want to
mention is dash, dash video. As you can guess, this parameter allow us to export the video. It creates a short movie of
your job being generated. Let's see it in action. I want to say Japanese
botanic garden. And this parameter
is not working with version 4.5 right now. So I'm gonna say version three and I add the
video parameter. After the job is completed, we cannot see any video content. That's because we should
first go and add a reaction. Select the envelope emoji
to get the video export. And after dots down here you can see the generation
process video. With this video we can see each step and eat second
during degeneration.
11. CourseProject: Congratulations on
finishing the cars. Alright guys, now
it is time to take your knowledge and create
your own projects. I encourage you to create a course project and
share with us to do that. First, decide on the MIT journal
version you want to use. Then if you need to
specify a resolution considered the
upscaling features you learned on this course. You will always have a
look at the platform and see other people
works to get inspired. Pick ME doom and style
for your own project. And finally, use at
least one parameter in your project for practicing. After you have the results, please share with us for
comments and feedbacks. Thank you for
watching the course. I hope you enjoyed and find
rarely on these tutorials. See you on the next one.
12. Comparison of AI Art Generators: As a generator AIR test, you should know the
alternatives for Adobe Firefly. There are many other
tools in industry, but in this video, I want to compare the most popular ones. Namely, we will be talking
about firefly, meet journey, stable diffusion, and Dalley, there's known best peak. Each tool has advantage
in different areas. The best way to figure out which one works best for you is to try them out and see which fits your workflow of the bust. And I will try to
explain the differences, pros and cons according
to my research. Let me introduce the tools
first, starting with Dolly. This AI tool was launched
by OpenAI in 2021. I need quickly gained attention
due to its ability to generate unique image from
textual descriptions. The software's name is
a mix of the names of animated robot Pixar
character volley, and the Spanish surrealist
artists Salvador Dali. Delhi is the first molar. Meet your new,
launched in July 2022. This art generator was
designed by a team of AI enthusiast who aim to make AIR accessible
for everyone. My journey is currently
only accessible through at this court boat on their
official discord server. You can direct message
to the vote or invite the bot to a
third party server. Stable diffusion is a
deep-learning text or image model released
in August 2022. It's unique feature
is the capability to gradually evolve the
artwork in real time, allowing the artist to observe the creation process and make
alterations at any stage. It was developed by the
startups table DEI in collaboration with a number of academic researchers and
non-profit organizations. Last but not least, Adobe launched firefly
in March 2023, releasing two beta text frame
which allows users to input a natural language
problems and receive an image similar to other
generative AI in which models, but there are more
features in Firefly. At launch, the firefly beta is only available
through the website. But Adobe stated it will
soon be integrated across its apps and it is already
integrated with Photoshop. All of these models are
trained using the dataset. Now Lee was developed with
something called clipped. It is by OpenAI. Clip is a separate model
that was trained on 400 million pairs of images with texts captions
scrapped from the Internet. Make journey uses many
datasets like Liam, which was scraped
from the web as well. In Forbes interview, majority of founder David halts has admitted that meet journey did
not receive consent for the hundreds of millions of
images used to train the AI. Stable diffusion
is no different. It again used pairs
of images and captions taken from layer
on five before training, which includes 5 billion
payers of image and text. The dataset was created by Leon, a German non-profit which
receives funding from stable it to as many time status
in this course, adobe Firefly trained on
a dataset of Adobe Stock. They are saying that
they did not train on any Creative Cloud
subscribers personal content. Next up, capabilities. In terms of capabilities, there is no official way
of comparing these tools, besides they are better
in different areas. So I'm going to share my own thoughts based
on my own experience. So I decided to say the
key features of each. Delhi is the first mover and still good for
photo-realistic Emirates. Also have API access, which lets you create new enrich in any number of
different styles as specified by the users prompt and include
in your projects. Meet journey, in my opinion, is the best in terms
of image quality. As you might guess
from its popularity, the best outputs comes
from MIT journey. For example, MIT journey is the only model that can
handle hand generation. While others still struggles
with hand drawings. You can define different
art styles and mediums. However, MIT journal
always creates image in Ralin meet journeys style. And you can easily differentiate the image generated
with MIT journey. Somehow it generates
aesthetically pleasing results. There is no API or standalone platform for
meat journey at the moment. On the way to use it is
through this chord up. Stable diffusion is best
for developers since it is open source and can be run
locally on your own computer, which means your
system capabilities is important as well. For example, tangible bite
or more RAM is recommended. On the positive side,
stable diffusion basically gives you more
options and control. You can set the number of steps, the initial seed and
the prompt strength. Lastly, Adobe Firefly is really competitive when it
comes to image quality. But there is more
than text to image, like text effects
and January to fill. Generator field is one of the most revolutionary
thing about firefly. Since it opens a
new door in AI art, it will be integrated in
Creative Cloud programs, like it is already
been in Photoshop. But I also asked the firefly
team if it's going to be a standalone platform and that's where it is not
clear at the moment. They said they are still testing and considering
the roadmap. Alright, that was the
wrap-up of capabilities. And please consider that an
ecosystem is changing fast. By the time you
watch this video, some things may have
already changed. Lastly, let's see the pricing
models of these tools. Dolly gives you free 25 credits. There's a rolling monthly, 15 free credit allowance. If you need more, it
costs $15 per 115. Let's meet journey is a subscription service
which prices starting at $8 per month annually
and $10 per month monthly. Stable diffusion is free
to use on the web app. And you can run the tool
without paint as well. But if you want
premium advantage that our packet starts from $9. Adobe firefly can be accessed for free using debate access. To use Photoshop beta, however, you must be a Creative
Cloud member with a subscription or trial
that includes Photoshop. All right guys, that was the comparison of
AIR generators. Thank you for watching
Yeti Learn, Design | Code | Animate
