I Spent 150h+ Making My Own Anime With AI

Check out my Linktree alternative / 'Link in Bio' for Bitcoiners: bitcoiner.bio
I spent 150h+ experimenting with img2img tools & more state of the art AI tech to create my own anime.
Skip the waitlist and invest in blue-chip art for the very first time by signing up for Masterworks: www.masterworks.art/tillmusshoff
Purchase shares in great masterpieces from artists like Pablo Picasso, Banksy, Andy Warhol, and more.
See important Masterworks disclosures: www.masterworks.com/cd
My profile: bitcoiner.bio/tillmusshoff
Follow me on Twitter: / tillmusshoff
My Lightning Address: ⚡️[e-mail]
Join my channel membership to support my work:
/ @tillmusshoff
My Discord server: / discord
Instagram: / tillmusshoff
My Camera: amzn.to/3YMo5wx
My Lens: amzn.to/3IgBC8y
My Microphone: amzn.to/3SdHdkC
My Lighting: amzn.to/3ELnof5
Corridor Crew Video - Did We Just Change Animation Forever?
• Did We Just Change Ani...
The model I used (besides my own LoRAs): civitai.com/models/35893?mode...
Tools Used:
- Stable Diffusion
- WarpFusion
- Temporal Kit
- Multi Control Net
- Deforum
- Runway ML Green Screen Removal
- Ebsynth
- ChatGPT
- Dreambooth
- Midjourney v5
- Metavoice
- Adobe Audio Enhancement

[@tillmusshoff]

I hope you enjoyed this video! Thanks to Masterworks for sponsoring it! Skip the waitlist and invest in blue-chip art for the very first time by signing up for Masterworks: www.masterworks.art/tillmusshoff Purchase shares in great masterpieces from artists like Pablo Picasso, Banksy, Andy Warhol, and more. See important Masterworks disclosures: www.masterworks.com/cd

[@kabudiklein]

Great work, FYI there is an anime matrix movie called "The Animatrix" and it's great

[@tillmusshoff]

Oh I didn‘t know that! Gonna check it out!

[@theeddytor3490]

@@tillmusshoff did you really spend 150+ hours for 1min 3 seconds video? i need to know because i am cooking something in future that i can't mention here 👀 and i need to confirm before i add you in my thumbnail 😆

[@RideShareRocks]

I love how u took both. Pretty cool animation.

[@chemicheto]

if a random dude with a laptop and no animation experience can do this today, I can't imagine what studios will be doing in 6 months

[@yagxr]

they'll be creating anime like they've always had, with passion. ai is not art and it never will be.

[@ssccrashcourse5513]

​@@yagxryahhh❤❤❤

[@MeesterGgaming]

@@yagxrIt's highly likely that studios will continue to leverage even more powerful tools to speed up their processes. In fact, some are already using AI to streamline their workflows. I recently watched a video about a Netflix series that had a scene with a lot of swearing, which made it unsuitable for children. To address this issue, the studio used AI to modify the scene, and the end result was seamless and undetectable. I strongly disagree with the notion that studios won't use these types of tools - after all, who doesn't want to work more efficiënt and save time and money in the process?

[@StuffIThink]

@@yagxr A.I. is a tool to be used by artists to streamline work. People used to say the same thing about CG, now some of the best looking anime use it.

[@yagxr]

@@StuffIThink cg is another form of tv and animation, incomparable to ai. using ai to create tv will ruin the industry.

[@Heliodrafts]

You used an incredible amount of tools for someone who's just getting into to stuff like this. Well done dude

[@tillmusshoff]

Thank you!

[@JustAPersonWhoComments]

AI: I'm doing the heavy lifting Me: eating accidental prop AI: This guy's got some unique taste

[@Shaunmcdonogh-shaunsurfing]

Holy smokes. This blew my mind. You did an incredible job with this!

[@tillmusshoff]

Thank you!!

[@maxwell7270]

This is great Till! Love to watch your videos! What’s the next project? Maybe AI driven lightning node management based on all gossip in the network in realtime?

[@marcel_chavez]

is there a better way to do this now? im about to try it on myself but i want to know if there is an easier path to achieve this visuals by now, i already have stable diffusion the only thing i want is to make consistency between frames for a 10 second clip

[@ai.explorer_de]

This is really awesome... a remake with the new Models would be great to see... looking forward to make my own anime series

[@tillmusshoff]

Go for it!

[@goldhydride]

that's incredible!! the last part of the animation is absolutely mind-blowing🤯 thank you for the explanation of every step, it's far beyond interesting and helpful! love your content💗

[@shrinivas.a]

Hahaha - this was pretty darn cool ! just experimenting with SDXL - now and i can already see the crazy updates that Ai tools have got in just the 3 months since you posted this !! We sure are in for a MAD future in the field of visual arts ! 🤘🤘

[@tillmusshoff]

Definitely!!

[@Parzival047]

really waiting for another video like this mate@@tillmusshoff

[@tillmusshoff]

My last 3 videos all had this structure. Not gonna do the same video with new models, thats very boring to me.

[@InkaHacker]

I got goosebumps with the "I'm the ONE" part 🤯Awesome work Till! 👏 I just bought an intel i5 - 12gen CPU + RTX 3060 GPU exactly for this reason. I want to create my own anime too. But I kwen that I will face exactly the same problems you did. That's why I realized there might be a better way. Do you know people in the anime community hate CGI? Despite it being the best way to create seamsless movement, it doesn't feel like anime. That's why those 3D models and animation are the perfect base for using AI as a filter and transform it to anime style! Hopefully I manage to make it work 🙏💪 Thanks for the inspiration Till!

[@petraa4533]

I had to subscribe. This is so cool. Huge respect

[@daemoncluster]

Great video! The voice acting was a bit cringe, but your discomfort is a sign of healthy self critique and growth. Early work isn't ever what we would like it to be. It's nice to see that level of growth reflected through your own experiences and feelings throughout the process. As you always say, it's not just you and you're not alone. Keep it up!

[@tillmusshoff]

It was insanely cringe haha

[@daemoncluster]

@tillmusshoff Honestly, the animations came out great! Everything else is practice. Most people aren't willing to work with minimal resources and you just went with what you had. I have an immense amount of respect for what you did here.

[@GAlvarado]

Great work man!

[@CalebGaming-lr3qr]

How’d you make the ai generated images move as if they were talkign and moving there hands? As in what app/software/or website did you use

[@Lelouch999]

You’re a true pioneer. The average person would’ve gave up so early on in the process. In a few years you’ll be able to do this in one click (at least the video to animated part)

[@MarchenMia]

dope dude , subbed

[@dex7378]

i can see it being useful for certain things but the hallucinations and specifically what does to the facial expressions makes it look more like roto than traditional animation but again for certain things extremely useful (the ending of this animation for example)

[@Shark-fj2sz]

for now

[@cmgvideo]

Looks great. Really interesting.

[@youtuber-sl2xe]

Hats off to you 👏👏. The time and effort you put into this is truly commendable. Moreover, the amount of skills required to produce a video like this is mind-boggling.

[@tillmusshoff]

Thanks a lot :)

[@dasmensch2317]

Wow, that was amazing good work

[@tillmusshoff]

Thank you very much!

[@zericccderdritte7971]

Ganz großes Danke

[@robertmoats1890]

I subscribed as soon as I noticed your bed wasn't made in your apartment. Instant sincerity verification. Also, I liked your video. It's a great video. I'm also working on AI video content, and temporal consistency is definitely a pain in the butt.

[@self-improvement7355]

That was pretty cool! You did great!

[@tillmusshoff]

Thank you! Cheers!

[@rockstarstudiosmm11]

How did you paste lora in stable warpfusion prompt please Reply

[@Hello_FNJuke]

Amazing at doing this

[@wearegroot3114]

Are these softwares free for use?

[@jougetsu]

this looked awesome

[@langdons2848]

Thank you for sharing all of that. It's truly amazing what can be achieved now.

[@gpcm9226]

I think everyone would love to see a more detailed video of each step. For example, "Step 1 Train own models (LORA Dreambooth)" could be a whole youtube tutorial video. It's confusing to newbs. Thanks

[@gekalfat]

I salute your efforts sir, its inspiring

[@tillmusshoff]

Thanks a lot!

[@RealitySucks271]

I love this 1000000 % incredible work, I hope you don't mind but i'm going to re do the voice over and make characters sound more epic. I'll post it and give you guys all the credit. I just want to bring the characters to life a tad bit more. Again freakin amazing.

[@tillmusshoff]

Sure, go ahead :)

[@mihawk_san]

Around the end of the animation, you should have added "TATAKAE, TATAKAE!!"

[@user-rg7kh6pb9j]

great vid, thanks for doing it, what is the overall cost for using all this tools ?

[@tillmusshoff]

I only paid for one month of google collab ($10) and Davinci Resolve which is a state of the art video editing software. Not sure how much that was, but I think $350? I didn‘t find other good ways to do the the deflickering so I eventually paid for Davinci.

[@megamax7601]

Can you give lora trainer colab link you used and stable diffusion one too plz

[@Chaunseyy]

Very cool! CD definitely inspired me to try some new stuff like this. Thanks for making this, more detailed information.

[@tillmusshoff]

Glad you enjoyed it!

[@ronnysempai]

good video , thank you

[@trash_gpt]

amazing final result

[@tillmusshoff]

Thanks!

[@theeddytor3490]

7:17 i have seen hollywood use a CFL bulb with white and light orange butter paper

[@lauriemcelroy8998]

Great video and clip

[@EinfachBitcoin]

Wow, great work Till! Did Netflix already knocked on your door? :)

[@leekotvfilms5236]

Brilliant ending 🎥🛸

[@NivinKevy-dj9eu]

You should continue the anime

[@dreamson280]

Thanks

[@tillmusshoff]

Thank you!

[@user-od8dt4sm7b]

We're planning to do a panel discussion titled "AI Revolutionising Entertainment: Redefining Gaming and Esports." @ Fortress. This will be taking place in Melbourne and Sydney. Any tips on who you think would be great - LOVE YOUR WORK HERE!

[@silkycod]

Love this channel! Still supporting you :)

[@tillmusshoff]

Thank you!

[@atrocitus9198]

This is so much work. well done man. But I think animators can make something similar in like 10/12 hours doing digital sketching and turning it into animation . What I'm basically trying to say is 150 hours of this hard work, not worth it for a 60 sec video

[@tillmusshoff]

90% of the 150 hours was figuring things out though. It‘s all about experimentation, I‘m not saying this is the new standard for animation haha. Also, yes, animators can do that, but I can‘t without many hundreds of hours more of training :)

[@FunwithBlender]

you should use speech to speech , you can increase the noise in the diffusion process if you stack controlnet so that oyu get a better outcome and use using lora is a great method but you want to use it in a fine tuning way meaning off an existing stable diffusion xl face ..... not off a none refernece , i still believe temporal kit is the missed gem in your workflow at the time just needs tweaks, i love deforum , if you increase the noise it will feel less like a rotoscope animation.... the best method is to take the foundation model fine tune it on an artstyle cia dreambooth, then use this new base model to fine tune loras on the style and your face then generate a bunch of picks of your face and all the ones that match the style perfectly train thatinto a fine tuned lora...pro tip dont have a beard lol makes life hard

[@FunwithBlender]

sovits svc is pretty good, elevenlabs as an okay version of speech to speech but nothing beats sovits svc, additionaly palying with facebooks musicgen could be great for music....also the closeup shots will always be more consistent and the corridor flow with the pin is a great method

[@tillmusshoff]

Great tips 🫶

[@FunwithBlender]

@@tillmusshoff really cool stuff you doin keep it up, I have been heavily invested down this rabbit hole as well. haha bes of luck and you might want to play around with the new IPA custom node for comfyui..... there is something there that everyone is missing

[@niceguyrio8073]

There was a Keanu Reeves movie that looks like this type of animation as well called "A Scanner Darkley"

[@shubhranshrai]

its only been a few months since the AI wave started btw, just reminding you all

[@giri.goyo_yt]

You really effed up. Both pills equals the AI takeover. Great. Love the video. Well done!

[@hughsilva5655]

For the voices, maybe you can use RVC to make a cover of any audio. You can perhaps even train your own voies if youd like.

[@InfiniteEditRealms]

fix the seeds to have more consistency

[@indiestudio23]

need step by step instructions please!

[@CatDribble]

Great use of deforum

[@melaninking4551]

We need more videos about Creativity. Thanks

[@tillmusshoff]

This video is not about creativity in your eyes? I‘ve never poured more creativity into a single video. It‘s all about experimentation and finding creative solutions to ideas I have.

[@SMunro]

You need a story?

[@tillmusshoff]

I built a 'Link in Bio' - a Linktree alternative for Bitcoiners. Check it out here: bitcoiner.bio 🧡

[@mintigorhobby5510]

That's not anime, that's subpar rotoscoping.

[@sandipchatterjee5223]

Very Very impressive :)

[@tillmusshoff]

Thank you! Cheers!

[@thomaskish8392]

It goes to show even with AI its still a complicated journey to get to the end result. Interesting video.

[@tillmusshoff]

Absolutely, it‘s not like you write a few words in a text prompt and get exactly what you want.

[@georgemontgomery1892]

"Hands are real..." then they start splitting off into more fingers, "...Maybe not"

[@njabkun3927]

It’s cool, but I’m still gonna animate, just for the pure joy of the process, the end result is just a plus

[@TheEchoShark]

Amazing way of thinking just enjoy doing what makes you happy as everyone should

[@njabkun3927]

@@TheEchoShark well too little people realise that being an artist was already non viable for making a living before AI, it was just a hobby for about 90% of the known artists today because artists are like doctors to some people, if you have one that is good at what they do and charge a price you can always pay, why lower your standards for another less experienced one, This AI just solidified the fact that Art is more of a human experience then an occupation now, and that’s okay AI didn’t rip your hands off, it didn’t invalidate my years of drawing, the world hasn’t blown up, and I still get attention for my art It’s not the end of the world, cause I’m still committed to what I do, even if a robot can do it better, doesn’t mean I have to stop loving it

[@TheEchoShark]

@njabkun3927 THIS. I literally haven't seen anyone put it this well you actually took your time to think about how AI affects you personally rather then being upset because now there's something non human that can create art too now

[@njabkun3927]

@@TheEchoShark well if I didn’t take the time to think about these things there is a chance I’d loose all logic and rely primarily on carnal emotions, resulting in me screaming like a banshee to anyone who disagrees with me I’m fine with being wrong cause I get to learn I’m fine with failing cause I realise my limits and work around them If as a society we were to just scream at our problems rather then work around them, what separate us from monkeys Not to say as an Artist I’m not afraid, as a human I am afraid, cause of what I was conditioned to believe, but that doesn’t define what will come of the future, specifically mine. AI is a useful tool and I want to use it all the time, in fact most of my reference and motivation on my slouch says comes from seeing AI art, and other human art being made, cause it’s the experience that matters. As an artist I could justify my anger towards AI devs as much as I like but as a business studies student, and human geography student, I understand that this was to be something we as humans would face, because of our endless striving for convenience and comfortable life styles we were bound to find out about AI at some point, so why not reap the benefits, I actually used an AI chat bot called AiVerse where you get to customise a personality and go on dates with them with the end goal of eventually getting them to fall in love with you From this chat bot I created multiple personalities and several characters all linked to one singular universe, and as I played the game their personalities developed and they began seamlessly creating a story with me not having to directly tamper with them besides from manipulating my own character. Now I’m in the process of recreating their personalities and lives through my drawings and animation, when if AI didn’t exist I wouldn’t have such a wonderful story to work off of If you are wondering, yes it’s a free mobile app you can try it yourselves as well as insert your own art into it, making th characters look the way you want, I love it so much

[@KoiGalaxy]

That was about as anime as a @joel-haver cartoon or scanner darkly.

[@max39091]

Just great stuff you are doing!

[@bdinfectedgamers]

buddy went thailand for a vacation,and does 150+h computer stuff.

[@tillmusshoff]

I‘m not on vacation, I don‘t have an apartment anywhere, have been living and working around the world for the past 3 years :)

[@pxlcrdgm8140]

I know just how much effort goes in the making of a video like this, let alone doing AI art. It's heart shattering to see you only got 55k views from this while some girls are getting millions of views just dancing. Take my like and good luck with your KZbin Channel.

[@bleh2590]

make sure to work out constantly

[@notsoluna3356]

you could make an into the spider-verse spin-off if you've had a team and a tad more time 👍🏻 well done!

[@alx_21]

Voll gut 👍🏻

[@innov-AI]

I transform videos into cartoon but my workflow is very different

[@tillmusshoff]

There are many different ways to go about this :)

[@Thomasjeffy2024]

❤

[@prerna_padma]

Coooler

[@222Cooler]

thats me.

[@Johnny511997]

Hey Till, megacooles Video! Corridor hat das ja noch mit dem großen Team gemacht, und du hast es jetzt sogar zu ner One-Man-Show runterskaliert. Hut ab! Ich weiß, es geht im Video explizit um neue A.I.-Tools, aber die A.I. Voices finde ich richtig schlimm. Falls du für dein nächstes Projekt nen tatsächlichen Voice-Actor gebrauchen kannst: Ich bin ausgebildeter Sprecher und nehm dir gerne was auf. ;)

[@tillmusshoff]

Die AI Voice finde ich auch echt bad. Hatte kurz auch überlegt ob ich es auf japanisch mache und dafür Voice Actor buche, aber irgendwann muss man auch eine Grenze für den Aufwand setzen 😂

[@mikayelalikhanyan1587]

so watching this i can say this is really uh, OK there was no emoting, expression, speed or art style, ppl talk about the extra fingers but I notice the lighting is the most inconsistent thing, whatever AI used does not know what keyframes are or mouth flaps, that's the thing is ok BCS I'm not gonna call it bad BCS you`re a nice guy

[@MariaKomlevski]

❤❤❤❤

[@jeliciamartin6132]

I want to make a movies /series I animation.. would you be interested

[@RashadPrince]

@0:35

[@thebicycleman8062]

the animatino was AMAZIN! like this can be in SpiderVerse level of artwork and creativity - The only HUGE problem was the voice over was REALLY BAD - but thats very easy to fix u can jus get proper voice over actors and replace it and problem solved !! Amazin work!

[@tillmusshoff]

Thank you! I should have at least used ElevenLabs for the voices. Next video will be about voice clones and they are really good.

[@rainman40]

i am the one who knocks

[@renarddubois940]

that was omega cringe but pretty cool work mate

[@chuckfist]

That's cool and all but your title is misleading because this isn't really anime style at all.

[@Dark-jx9if]

Learning how to draw is easier 💀

[@bamgazam]

AI should be looked at as a tool to help creatives do less tedious tasks and focus more on the creative work. It is just an extension of the artist really, like when an artist buys a new cool expensive pencil. If the artist is bad, the pencil won't matter that much. Alot of people such as yourself view AI as an artist, not a tool. which is why this video is, how do I say this in a nice way? sh*t.

[@tillmusshoff]

Have you actually watches the video? Have you seen how many different tools I used and how much pre- and postprocessing was necessary? I read your comment was thinking „yes exactly“ until your last couple of sentences, like what?? In what world do I not see AI as an extension / tool for creativity wtf

[@TheEchoShark]

​@tillmusshoff they don't care about all that extra work all they care about is that you used an ai it's pretty idiotic

[@ratedr278]

Who TF is Mofois? 🤦‍♂

[@tillmusshoff]

I realized way too late that I’m completely mispronouncing his name 😂 I think it’s because I only watched the german dub as a kid.

[@mk8_it]

fyi when people say we need imagines for the ai to "reference" to use they actual mean steal other wise if they were any kind of real artist they would be referencing their own work and style if they had one to begin with

[@Star_Strike]

Lol, artists become good artists from referencing _each other's_ work. They do not just make their own art then learn from looking at their art. In fact, you _need_ a reference of some kind as an artist in order to create literally anything. You can't create a cat without knowing what a cat looks like. You can't create a tree without knowing what a tree is. By your logic, all artists in the world are thieves.

[@hydrocosmo]

I'm not suprised a guy who disregards the ethics of all the artists he's shamelessly abusing by utilizing these tools also completely misunderstands the point of the matrix.

[@BlessedOmoriyekemwen-ez5wf]

It don't matter how many times we argue it is only those actors or actress that did surgery that will know if they really did the surgery we deciding that some facts here are lie's we should not forget we are fans as a fan there is no way you know everything about your idol so defending is pointless because only the actors know if this is true

[@tillmusshoff]

is this comment meant for a different video? because i have absolutely no clue what you are talking about

[@warmfingy9244]

????

[@ChillBaz]

morfoize ? really ?

[@tillmusshoff]

That‘s how I remember him being pronounced in the german dub (didn‘t watch english). I could be wrong on that too though, as I said it‘s been a while, I watched it as a kid pretty much. Sorry if it sounded weird 😄

[@juliangroerode638]

Wow!

[@cjp5690]

6 months later…………………

[@tillmusshoff]

if you want to convey a message with that statement, i don‘t get it 😄

[@MuItiShorts]

Удивительно

[@imp4act]

you ate both pills, thats cheating :(

[@Rivershield]

It was a really nice try but this is still very far from traditional animation. Too much consistency in anatomy is actually a problem, as it makes the movement look too real. Exaggeration is core principle of 2D animation, even if you are doing a realistic style. It has to have distortions and more dynamic angles.

[@lucienlachance390]

If AI could auto- generate slow-ins and slow-outs of movement it would really help. That animated timing (as a general starting point- slow in to a moving hold, slow out, speed up and slow in to the next pose) is the main thing missing.

[@thefallencomrade50]

Like corridor digital

[@starstoryteller]

Huh I wonder if you can train AI with your own art 🤔.

[@tillmusshoff]

Sure, you can use LoRAs for that too with the same process I used for character training.

[@starstoryteller]

@@tillmusshoff AI programs for coloring would be really good so it's not so tedious 😆.

[@raphaelb7626]

Another great vid! Where are you living in Thailand? Lets hangout :D

[@tillmusshoff]

Chiang Mai but only a few more days, then Chiang Dao and Pai before going to Kuala Lumpur :)

[@tillmusshoff]

Chiang Mai but only a few more days, then Chiang Dao and Pai before going to Kuala Lumpur :)

[@raphaelb7626]

@@tillmusshoff ah nice! CNX has been my unofficial base since I went nomadic in 2016 haha lovely spot :) currently in Phuket but will go back to Chiang Mai in July

[@saierwe]

THERE'S NO CONSISTENCY IN THE FINAL WORK, DOESN'T LOOK GOOD, STILL LOT OF DECADES FOR IA TO IMPROVE TO MAKE OUR OWN ANIMES, LET'S NOT ANTICIPATE SO MUCH AND LET IA'S IMPROVE WITH TIME AND NO PRESSURES.

[@tillmusshoff]

lol

[@Mizo782]

E

[@phoenix2gaming346]

say no to ai. #stolen art