Face Swap GRAPHIC NOVEL Characters from Midjourney

📙 Midjourney COURSE mastersofmidjourney.com Beginners to Advanced
🚀 FREE Midjourney Cheat Sheet tokenizedhq.com/freebies/mj-c...
🔗 FREE Supporting Material tokenizedhq.com/freebies/vide... Folder: 2023-08-12 - FaceSwapping Graphic Novels with Stable Diffusion
🔗 FREE Promptalot Extension promptalot.com/extension
📺 Tutorials by @sebastiankamph 📺
How to Install AUTOMATIC1111 • How to Install Stable ...
How to Install & Use ROOP • This Face Swapper is M...
-
📰 AI Newsletter 👉 Your Inbox tokenizedhq.com/newsletters/ai
🐤 Follow me on Twitter / chrisheidorn
💼 Follow me on LinkedIn / christianheidorn
📸 Follow me on Instagram / christianheidorn
💬 Join the Tokenized AI Discord tokenizedhq.com/invite/discord
-
📺 Recommended Related Videos & Playlists 📺
Commands, Syntax & Parameters: • Commands, Syntax & Par...
Prompting Methods & Experiments: • Prompting Methods & Ex...
Character Design Series: • Character Design with ...
Storytelling: • Storytelling with Midj...
Watch Me Prompt: • Watch Me Prompt in Mid...
This Brand Does Not Exist: • This Brand Does Not Exist
Midjourney for Merch: • Make Money with Midjou...
-
A few weeks ago I showed you how to use InsightFace Swap to achieve perfect facial consistency with your characters.
And while this definitely was a game changer for many of you, this method still had its weaknesses.
You see, unfortunately, InsightFace Swap doesn't work very well with art styles that are less realistic.
This is particularly true for certain types of illustration that are very common in graphic novels.
The faces swap just fine but they always look out of place because the facial features end up looking too realistic.
But luckily, there's a workaround that works surprisingly well and it's also dead simple.
If you use Stable Diffusion a lot, then in all likelihood you are aware of the "Roop" extension for AUTOMATIC 1111.
It uses the exact same model as InsightFace Swap and allows you to face-swap your images entirely for free.
That's also why Roop has the exact same issue whenever you try to face-swap an image that uses a graphic novel style.
But guess what? It only takes one simple additional step to fix this and absolutely NO ONE is talking about this.
So that's why, today, I'll show you exactly how to do it.
-
⏰ Timestamps ⏰
00:00 Face Swap for Graphic Novels
02:37 Setting Up Stable Diffusion
03:28 Preparing Our Image Assets
05:56 Swapping the Face
10:24 Fixing the Style
#midjourney #midjourneyart #midjourneyv5.2 #storytelling #storyboarding #faceswap #graphicnovels #aiart

[@sebastiankamph]

Thanks for the shoutout! Here's something for your fans: I got my wife a fridge for her birthday. I can’t wait to see her face light up when she opens it. Hit me with your best dad jokes, Heidorn fans!

[@TokenizedAI]

🤣

[@emx9669]

Please continue exploring and sharing your knowledge regardless if it’s MJ or SD. You explain things so well that I may even start tinkering with SD more

[@quietlyworking]

Really helpful and super fun! Thanks🙏

[@DigitalAscensionArt]

Chris, I told you that you were the man back in February. You always prove me right. This is more evidence. I am amazed sometimes at how technical stable diffusion can be at times. I've spent the last 4 months working on mastering it and I'm so glad I did it. Because now when I use it to modify my MidJourney art I get stellar results. Both MJ and SD have their strengths and weaknesses. And using both apps circumvents the weaknesses and allows you to focus on the art that you have in mind, not just what AI gives you. MJ to me is great with color, composition, and texture. SD is great with some of those things and simply good with others. But where SD shines is with control, consistency, and fixing and adding details. People can avoid SD if they want. Not everyone needs that specificity with their art. But if you are looking to level up your art, combining these apps is the way to go.

[@wholeness]

Game changer again!

[@shelleyplum2286]

you are awesome..ty so much for all the great videos and info!!

[@marineris]

Awesome! This tutorial rushed in just in time for me. It might be of help to nail the style on my second graphic novel! Thank you, Chris. Keep on with your great tutorials.

[@maxwellcoleshow]

Awesome awesome awesome. I appreciate this so much. 🙌🏽💣💥

[@andreamuller8468]

Wow megasupercool 😊thank's

[@NexilisCreations]

Going out of our comfort zone I see, thanks for uploading! :)

[@HellNahAI]

This is a big help! I upload all of my creations onto KZbin. Advice is welcome 🙏🏾 😎

[@ESheridan]

Hello! You are genious! Thank you so much! I tried to find the solution how photo realistic face after roop to make for comics and you to help me this! You are genious! 👍👍👍👍

[@OriBengal]

Bro, knowing how to learn and experiment is key... You've got that down... It doesn't matter what tool you use.... So yeah- everything starts off a little shaky. You're doing great, and you're going to be a master in SD in no time. Cheering you on.

[@JaiTV]

Thanks for doing this. It's far too complicated for my taste. But many folks will love it.

[@leodistefano7424]

...speaking of graphic novels and comics, I did some tests on Midjourney asking to describe the image designed by one of the greatest Italian designers, Magnus... well he didn't recognize it. I also did some tests on other platforms and I haven't found any styles influenced by Italian comic artists...they don't have them in their databases. Among other things, there isn't even a way to communicate it to him... hello, congratulations for everything and good work

[@stormjack]

LOL! "Why pay attention if your primary objective is to complain?" Thanks for doing these videos.

[@WhatHappenedToThem]

I have followed the directions to a T for both you and the other guy you recommended and it will not load the ROOP. It keeps saying "HTTP Error 401: Unauthorized" when trying to load the 'onnxruntime'. I have even googled it and took different steps and it seems that the file location is set up on a password protection now. Maybe too many people were trying to access it lol.

[@mire2official]

love the video. Is your lip ok? try some blistex it helps with lip repair.

[@TokenizedAI]

Funny you noticed that. My lip is actually fine. I have no idea why, but it looks like that in all the videos. Maybe the lighting?

[@morpheus2573]

Thanks for a great tutorial. However, I agree with you... compared to MJ, SD is "way too nerdy." The way things are going, someone will soon integrate all of the functionality of SD into a single program with a user friendly UI without having to deal with LORAS, extensions, and all the other work arounds that turn my pea-like brain into a pretzel... x 10 if you're a Mac user. 🖖

[@sherpya]

to keep style, did you tried with remix + variations?

[@TokenizedAI]

Have you ever used a face swapping tool before?

[@sherpya]

@@TokenizedAI yes, I used roop as in your video, without the img2img style fix

[@chariots8x230]

Thank you for going more in depth on creating consistent characters for graphic novels & comics. These types of styles is that I want to use AI for. Btw, can the face swap also be used for more cartoony styles, such as manga or anime? Also, can we do a face swap and apply a facial expression to our character at the same time?

[@TokenizedAI]

Try it out

[@footy7407]

Send me a video which is best plan for Midjourny

[@footy7407]

✨

[@user-rt6nk9sc4y]

this app can create Two people are talking so that when they talk their mouths can move and how to do it ?

[@guerrerosdelamotivacion]

Hello, could you dedicate a video about the proportions of midjourney 3:2, 16:9, 23:9... And what measurements in centimeters do they correspond to when creating a poster and printing it? thank you

[@TokenizedAI]

I already have a video on aspect ratio

[@stormjack]

I'm working on a project where I need to reproduce a room from multiple angles. Could be via Midjourney or Stable Diffusion or both. Have you found a process for doing this?

[@TokenizedAI]

MJ and SD are image generators, not 3D modelling applications. They don't have spatial awareness. What you need is a 3D model of your room in a 3D modelling software. Neither MJ nor SD can do that.

[@stormjack]

@@TokenizedAI I see. Thank you!

[@ScaryStoriesNYC]

I've never gotten any of that 1111 stuff to work, I'm an artist, not a coder LOL. i don't think you should be as worried of people who dislike MidJourney since they are not your audience. People who don't like anything installed on their computer are your audience. Once there is a way to do this using a website or Discord, I will watch your video about THAT but I am FINISHED installing things I can't get to work in order to do AI images.

[@ScaryStoriesNYC]

I am NOT the one who gave you a downvote, though, I would never do that to you. I'm giving you an upvote even though I don't want to use this. Thank you for all your hard work.

[@sherpya]

anyway most of the realistic style is due the face restoration

[@marguelipemarguelipe6595]

🥰

[@TopicTome]

Would you say this works better for graphic novels compared to INSIGHTFACE?

[@TokenizedAI]

That's kind of the point of the entire video, don't you think? 😁

[@puja1985]

Oh no, not stable diffusion! As a visual learner and old-timer in traditional art, I have my limits when it comes to AI learning curves and workflows. There is a line I need to draw when it starts to hinder my creative capabilities. I tend to get stressed out if it requires too much technical stuff. The more user-friendly the program is, the happier I am in creating art, and the happier my art will be without sacrificing my sanity. I'm sure there are a lot of alternative ways to create similar results, even if it requires a little bit of traditional manual technique. When it comes to creating a series of consistent character illustrations, I see that AI still produces stiff results. Unless you're a geek who really likes coding, you're probably going to create a more flexible and precise result as specifically intended.

[@CrhystyannDavidBernalChaves]

Welcome to the dark side of STABLE DIFFUSION hehe (greetings from Bogota - Colombia )

[@skaska8966]

Why you just don't use insightswap in discord.. simple I think

[@TokenizedAI]

Because it doesn't properly mimic graphic novel styles. Sounds like you didn't watch the video.

[@skaska8966]

@@TokenizedAI nope.. this is your first vid I watched... Just I found you on KZbin suggestion.. would you please pt me to your vid?

[@skaska8966]

@@TokenizedAI ok and about this video.. let's consider that this is better method.. you showed that we can do swap for 1 face.. what about if I want to swap 2 faces at the same time where my generated picture contains 2 faces... How will work your method? Thx in advance for your answer.

[@TokenizedAI]

I'm curious. Why would someone comment on a video without actually watching it? This is a mystery to me. 😅

[@skaska8966]

@@TokenizedAI maybe we are confused I mean this my first video I've watched in your channel...but I watched your video and maybe I missed the point where you said that insightswap is not good for memic the comic style.. here we go more clearly.

[@coldfires1234345]

am I the only one here who thinks this is both extremely helpful and insanely exhausting? please stop monologuing. "so i decided to.." "i then, i did..", "i", "i" , "i", I wish I had the ability to cut these videos down just to the important details. not to mention every single time i follow this guy down a rabbit hole, I end up downloading and purchasing more and more just to solve problems. making art SHOULD NOT be this complicated.

[@TokenizedAI]

I think you should be directing your frustrations at Midjourney, not the people who are trying to help you.