Generate AI Videos With AnimateDiff & ComfyUI | Text to Animation Tutorial
Vložit
- čas přidán 7. 06. 2024
- The first 500 people to use my link will get access to one of Skillshare’s best offers: 30 days free AND 40% off your first year of Skillshare membership! skl.sh/mdmz10231
Create impressive AI animations using the Animatediff extension. this tutorial covers the installation process, important settings, and useful tips to achieve great results.
ComfyUI: bit.ly/3LM1hbN
ComfyUI manager: bit.ly/3LRqYrg
AnimateDiff: bit.ly/3RO22ol
Motion modules: bit.ly/3rvCWjG
VAE: bit.ly/3LRLM1X
Find prompts on SeaArt: bit.ly/3PMrKsc
More prompt websites: • How to Write A Prompt ?
100 Generations:
bit.ly/46AUQAe
🔗 Software & Plugins:
Winrar: www.win-rar.com/start.html?&L=0
Topaz Video AI: bit.ly/3t04Otl
⏲ Chapters:
0:00 In this video...
0:15 ComfyUI vs A1111
0:39 How to install ComfyUI
1:39 Great deal from Skillshare
3:01 The ComfyUI interface
3:15 How to install AnimateDiff
4:44 Import AnimateDiff workflow
5:12 AnimateDiff settings pt1
6:40 Prompting in AnimateDiff
7:23 AnimateDiff settings pt2
7:56 Generate animations in AnimateDiff
8:25 How to Import settings
8:52 Animation pack
9:07 How to upscale video with ComfyUI
9:40 How to upscale video with TopazLabs
10:20 I have a question
➕Positive prompt:
waist-up "female Astronaut in a Jungle" by Syd Mead, broken helmet tangerine cold color palette, muted colors, detailed, 8k
➖Negative Prompt:
Deformed, blurry, bad anatomy, bad eyes, crossed eyes, disfigured, poorly drawn face, mutation, mutated, ((extra limb)), ugly, poorly drawn hands, missing limb, blurry, floating limbs, disconnected limbs, malformed hands, blur, out of focus, long neck, long body, ((((mutated hands and fingers)))), (((out of frame))), blender, doll, cropped, low-res, ugly, disfigured, too many fingers, deformed, grainy, extra limbs, bad anatomy, deformed, extra limbs, extra fingers, mutated hands, bad proportions, blind, bad eyes, ugly eyes, noisy, text, writing, logo
Support me on Patreon:
bit.ly/2MW56A1
🎵 Where I get my Music:
bit.ly/3boTeyv
🎤 My Microphone:
amzn.to/3kuHeki
🔈 Join my Discord server:
bit.ly/3qixniz
Join me!
Instagram: / justmdmz
Tiktok: / justmdmz
Twitter: / justmdmz
Facebook: / medmehrez.bss
Website: medmehrez.com/
#stablediffusion #aianimation #animatediff
Who am I?
-----------------------------------------
My name is Mohamed Mehrez and I create videos around visual effects and filmmaking techniques. I currently focus on making tutorials in the areas of digital art, visual effects, and incorporating AI in creative projects. - Krátké a kreslené filmy
So...what other tutorials on AnimateDiff or ComfyUI would you like to see on the channel ?
Common fixes:
- for VHS_VideoCombine errors, install the custom node ComfyUI-VideoHelperSuite
- if you can't see the manager after installing, make sure you have the latest version of GIT and re-install, if it persists, retry the installation methods here: civitai.com/models/71980/comfyui-manager
- visit the AnimateDiff Discord for more technical support: bit.ly/48Vnpdu
Thank you so much for this. Imma use this tonight. I would love a tutorial using ControlNet and IPadapter, more specifically I would love to learn the best workflow for integrating images into the process.
I'm currently working on an AI production and using a still from one of the video scenes would be an amazing way of blending the live footage with the AnimateDiff visuals.
I'm using google colab but the GPU disconnects after 2-3 generations. Please help me with this? also I'm using free google colab, I haven't bought any computing units.
I can't run it on my laptop cz it's Core i3, 8 Gen, 8gb Ram, 2gb Vram hahaha
the lora stuff
Hey, what if I got a mac? It won't do winrar.
@@starhaven that sounds fun, I'll be making more tutorials on AnimateDiff for sure
Seriously, you've convinced me! I'm making the switch to comfyui. The ease of dropping workflows is a game-changer. Thanks for the insight!
Have fun!
Love your tutorials! And thanks for promoting ComfyUI it's so much fun to work with. I'll probably switch to it in my tutorials soon, even though I feel like a lot of people are still a bit scared of the node structure.
glad u like the video, have fun :)
@mickmumpitz I’m excited to see what you do! Been following your Channel as well ever since your first video 😊
I can’t wait to see your progress with this new tool. I learn so much when I watch your channel 🤓❤
More to come!
Most tutorials on ComfyUI and AI in general are self aggrandization sessions. Yours are clear, concise and super helpful. Thank you. Greets from Kenya!
Glad to help!
wow this is amazing!!
Thanks man
Thank you for this tutorial, yours is by far the easies tutorial of animatediff to follow along with this workflow and I got it to work finally:)
Glad it helped!
You're a legend! This is one of the best ai tutorials I've seen! Thank you! 💯
glad you think so :)
Great tutorial! Thanks a bunch for making this!
You're very welcome!
THANK YOU BROTHER! IT WORKED BEAUTIFULLY!!!!!!!!!!!!!!!!!!
Glad it helped
our savior has posted again!
🙏😁
thanks @MDMZ this was by far the easiest ComfyUI tutorial personally for me. I tried to get into ComfyUI more but this video really helped me. Now I would love to do all the work in ComfyUI.
glad it was helpful
Woooow! I really love this tutorial! Thank u very muuuuch
Glad it was helpful!
Awesome video ❤
Thank u so much
Thank you!
You're welcome!
thanks for the tutorial!
Happy to help!
that is sooooo cooool, thanks
glad you liked it
Wow, that's a good video.
Glad you think so!
Very cool guide! How did you add the AnimateDiff combine node?
Edit: nevermind had to restart comfy for it to appear.
Awesome video Mate 🙂 I am definitely gonna try this. Comfyui still looks overwhelming to me, but i just have to dive into it i guess 🙂 Why is it so much better then A1111? Because of the output quality? Or the workflow and speed of rendering?
Thanks Lenny! glad you like the video, ComfyUI is certainly much easier to setup, it still has its own limitations but I like how easy it is to load/import workflows, customize them, and having visibility over the generation queue, I did hear from other users that it runs faster on lower end GPUs as well, but not sure as I didn't do any speed comparison tests
@@MDMZ He Mohammed, Thanx for the great explanation, i am definitely going to dive into it. I like that the loading and importing of workflows is easy 🙂 But i heard it can't run Deforum yet.
Wish you a nice weekend 🙂
@@digital_magic yeah I would still use A1111 for things like deforum, have fun!
@@MDMZ Thanx and i am glad you share the same opinion about deforum 🙂
Hey great vid - I want to use my own images to create an animation - do I have to add an image to the animated nodes?
not yet possible
i saved it to watch later i hope this will help me alot
have fun!
Girl is doing her hair with her hand while wearing a helmet! Great Art! Lol 😂
Right?!
Do you have a guide on how to do this with SDXL models?1.5 feels like such a downgrade now (and theres the whole forgetting how to prompt for 1.5 lol)
This is an awesome video. Is there any way we can make a bit longer duration vid???
I think it's possible now
ещё бы сиды каждого кадра были идентичными как Runway Gen-1...
Fantastic tutorial! Thanks a lot, right now Im watching your second video on AnimateDiff. Then I will watch what you have on Automatic 1111. Can you please help us to use the LoRAs?
Have fun!
thanks
You're welcome!
@@MDMZ i need workflow: img2img of animatediff-comfyui. Can u help me? Thanks.
Is it possible to start with a ready (single) image instead of a text2img prompt? I have an image I wish to animate - can I start with it and then do img2img with a prompt what I want to achieve?
not yet
Hi, Topaz Video AI is error 404'ing. So I'm trying to work out the upscaling via comfyUI and wondering if I have to re prompt everything again after dropping the upscaling workflow in, or can i just add the node into my current workflow (and how/where to connect it) and then running the prompt again to upscale? I from A1111 where upscaling is so much more simpler lol.
I'm looking into that as well, it would be much more efficient, currently you have to do it manually if you move from the normal workflow to upscaling workflow.
just fixed the topaz link, thanks for pointing it out
Please tell me system requirements to install this software? Please 🙏
You really deserve millions of subscribers and views.
Thanks 🙏
o meu deu esse erro [WinError 126] Não foi possível encontrar o módulo especificado. Error loading "C:\Users\Freitas\Desktop\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\lib\c10.dll" or one of its dependencies.
Hi, friend! Can this be run from Google Colab? or only from local with GPU?
Yes! I believe it's possible, Olivio Sarikas has a tutorial on it
hiya, my AnimateDiff_loader folder (orange one in top ) is in red colour saying its missing somethig so what should i do?
Hii. Thank you for this tutorial. 🙏🧡 How can I add zoom in camera?
I believe there are new models for that which I havent covered on the channel yet, but you can try specifying the movement in the prompt, sometimes it works
How many resources are needed? How much minimum Vram do I need?
not working ComfyUI manager when i see the interface the input showing Default not manager
hi, check the pinned commend
Thank you so much for this!
ComfyUI is so much more powerful.
You're a Top G!
You earned new subscriber as well, thanks again!
Please do more video's like this on ComfyUI.
Thanks for the sub!
Hey, i have a problem. Generation proces is too slow for me (much slower than in automatic1111). Time for generating one animation was too long, look at it: "[2:28:46
hey, most likely it's because you have 6GB of VRAM, this works best with 12GB or higher
Is there a way to use controlnet or images to get animatediff to generate a video?
❤❤
why windows detects virus when I trying to download comfyUI ?
it's normal, it should be safe
FYI, going through this (thanks by the way) I had to also install ComfyUI-VideoHelperSuite custom node set due to the AnimateDiff pages workflow examples now requiring it.
Thanks for sharing!
can you tell me how to get a lot of results at once? For some reason it doesn't work to get more than one, thanks for the video!
not sure what you mean, are you referring to the prompt travel method? I'm working on a separate video for that
For some reason all of my outputs have the same semi-transparent watermark across the video in the bottom third of the frame. Added watermark to the negative prompt but it persists. Any idea? Is it something embedded in one of the animatediff models?
watermark? that's weird, try using a different model
@@MDMZ Read somewhere that the animateDiff models were trained with shutterstock images and are prone to watermarking
how can i use comfyui (or Automatic1111's Web UI) to implement a transcription lipsync feature so the model reads your text?
Awesome video. Can you try to make it work on gg collab for some one with weak computer like myself
hi, that's possible, you can check the tutorial by olivio sarikas
How to control the movement so we can make real short movies? Instead of random weird movements?
This is a super new tool, cant ask much from it yet, but i like yoir ambition 😉
🥳👏👏
Thank you
You're welcome!
@@MDMZis this free or we need to pay for this ai
It tells me that I'm missing ADE_AnimateDiffLoaderGen1, but neither in the manager nor in Google does it say what it is and where to install it, so it doesn't let me do anything, not even the most basic thing.
that's weird, try to re-install everything from scratch, and if you haven't yet, open the manager and install missing custom nodes
"When loading the graph, the following node types were not found:
VHS_VideoCombine
Nodes that have failed to load will show as red on the graph. ComfyUI"
Wanted to manually install it, moved to the right folder but the same error comes out, can anyone advise how to solve it? (EDITED) everything worked when I redid everything as shown in the video, try again, it worked for me!
Same issue.I tried the animateDiff Combone (depreciated) node and deleted that other one...now the error message is huge I dunno how to add tbe ''Combine'' node. Help... 😭
everything worked when I redid everything as shown in the video, try again, it worked for me!@@kanohane
I will be adding fixes in the pinned comment
The install manager doesn't appear to load - any ideas as to why this isn't working? Great video by the way!
hi, can u try downloading the latest version of "Git for windows", and reinstall comfyui+manager
My ComfyUI dont have the "manager" button, why ?
hi, check the pinned comment
how do you getting the moviments?
hi, the entire process is explained in the video
Hey I get the following error code:
"When loading the graph, the following node types were not found:
ADE_AnimateDiffLoaderWithContext
Nodes that have failed to load will show as red on the graph."
Any ideas?
ComfyUImanager and Animate diff are properly installed but no luck. I cant get the error to go away
update ComfyUI
updated. but the same error showing
@@MDMZ
I'm running into an issue that says "Error occurred when executing KSampler:
cannot access local variable 'orig_memory_required' where it is not associated with a value"
Any help?
hi, check the pinned comment
Hi thanks for the video! I'm getting this error whenI drag in the setup image into the UI:
When loading the graph, the following node types were not found:
ADE_AnimateDiffLoaderWithContext
Nodes that have failed to load will show as red on the graph.
HELP! I've tried reinstalling and restarting.
hi, can you try to follow the steps from @lihuelalejandro7786's comment ?
Yes thanks - that got me the VHS node but the actual AnimateDiff Loader is installed but showing up as missing. I'm thinking it has something to do with being on a Mac unfortunately. Thanks @@MDMZ
To anyone else having this error : "ADE_AnimatedDiffLoaderwithContext" and trying EVERYTHING with no luck - in manager do an Update All - Finally worked. Thought I was going crazy doing several clean installs and restarts.. The only thing that worked was doing this as the last step.
@@imrsvhk are you on a Mac and using ComfyUI through Pinokio?
@@tonytitani410 Sorry No, windows.
Can you use stable warp fusion with ComfyUI?
warpfusion works on google colab
Another question my man, how can I import settings on stablewarp easily like deforum? I Couldnt find in the colab@@MDMZ
Does img2img works in comfyui?
yes
Hi! I can't see the cutsom node called "AnimateDiff (Kosinkadink version)". I only have "Animate Diff Evolved" which is published by Kosinkadink. I've isntalled it, but cant find any nodes called "Animate Diff Loader". Anyone had the same issue?
does it work after installing that ?
you can try to import the Wokflow first, and use the manager to install missing nodes, it should be able to detect the nodes
Getting error- node types not found when load workflow- I see it in the Custom note folder animate diff loader w context and vhs video combine... CUI is not seeing custom node ...did restarts
hi, check the pinned comment
What is the specs required for installation of COMFY UI ?
I haven't found an official list of requirements from the developers, but it's important to have a decent GPU, with atleast 12GB of VRAM to get it run properly
I am getting the VHS_VideoCombine errors, where do I put ComfyUI-VideoHelperSuite ?
what does the error say ?
hii , can you please provide the workflow it not there anymore and also not working .
Hi, just checked the animatediff page and the workflows are still there, maybe u can try a different browser
Could I for example create say Naruto running and then fighting nothing. Then make a video of me in animatediff of me fighting back nothing l, then edit it together. Or do I need t go down the 2D animation path?
you will definitely need to do some compositing in software like After Effects, have fun!
@@MDMZ that what I was afraid of 😂 #herewego
Hi sweetie the image workflow we drop n now has a video module instead of animatediff combine I think it is depricated.
can you be more specific, there are a few in the video
do you know if thsi works on google colab?
yes it does
Could not find ComfyUI manager button in the Comfy UI manager website
Hey there, the only button you need on that page is the download button, it's a big blue button 😁
if you dont see the manager on your ComfyUI, make sure you place the manager file in the comfyUI directory before running it, and restart comfyUI
I have the same issue! After I extract Comfy ui, I download the Manager, extract it like instructions, and still no Manager in the GUI. Any advice please ? Thank you in advance@@MDMZ
Hi, nothing happens to me at all when I insert the photo on the ComfyUi page. Help
which photo are you using ? try loading the PNG file instead of the GIF file
comfyui is the way of the future
I agree
Hello Is it working on mac Intel ?
I believe some people managed to make it work, check the ComfyUI installation guide
Hi, I was following a tutorial, but comfyUi manager was not installing...
check the pinned comment
I still have an error after I installed the custom node ComfyUI-VideoHelperSuite, could you please advise how I can fix this issue?
hi, what's the error?
Hi @@MDMZ, the error says: "When loading the graph, the followingnode types were not found: VHS_VideoCombine" The output node for the image shows red in the graph. Thanks!
@@AM7AM012 open manager, and install missing custom nodes
cant figure out why but this crashes on the ksampler every time. cant figure out why.
Do you see an error ?
how to install ffmpeg? thanks
in the config manager i dont have "manage" cant install nothing. tried many versions of it but it's the same
hi, check the pinned comment, there's a fix for that
@@MDMZ installing git i fixed
@@allago11 nice
hello sir.. why my config manager missing
Hi,check the pinned comment
make a video for the loras please :D
👌
For some reason I can't install ComfyUI Manager
hi, check the pinned comment
Hi and thanks for the tutorial. I get this error:
Error occurred when executing ADE_AnimateDiffLoaderWithContext:
invalid load key, '
hi, it looks like you have a typo somewhere, double check your prompts, I might be wrong tho, so check the pinned comment for more support
@@MDMZ Hi there! I'm getting the same error as Eran, any more suggestions on how to fix thisd? Many tnx!
you lost me after saving the file to the vae folder! i am a chef and artists so i managed to get that far then lost the plot please help....... brain fog now!
hi, are u unable to find the VAE folder ?
Make A Video On how to make wonderful anime series/short movie using
AI tools, editing Software also unreal engine 5 or blender.
Step by step process
Sound effects
Characters
Storyline
Amazing looking scene or environment
Intro song
Theme
❤❤
😊😊
It is possible
Plz make a Video
Thanks for the idea!
something is wrong, i do all the steps, use the workflows provided, and they do not produce the same result, it is superbad colors, low quality, weird rgb artefacts
strange, feel free to drop your workflow on discord, I'll have a look
@@MDMZ thank you, doing now :)
I want to learn from u
well, u came to the right channel
Is it free
yes
VHS_Video Combine errror
hi, can you try to follow the steps from @lihuelalejandro7786's comment ?
thank you Bro@@MDMZ
same workflow, not woking 😐
how so ? getting an error ?
Comfy just isn't as popular unfortunately. The intimidating workspace just scares too many people off.
True, but I think the interest will grow over time
For example me, I’m leaving automatic1111, it’s too damn buggy, everywhereeeee bugs bugs bugs…….. errors errors errors……..
@@deadbushlover3511 I haven't had that yet with A1111. I guess I get overwhelmed by comfy with everything in my face at once. That and I could never get it to work like I wanted it to.
Hold up....... you sound familar
I wonder if you just heard one of my videos before 😂
No doubt, great tutorial by the way@@MDMZ , the other probably was too, I've watched a few. :) And now on to ComfyUI going to have cross the node threshold sometime.
It’s too bad none of you tutorial hounds ever learn anything all the way before you attempt to teach it online.
I like that though. Keeps everyone doing everything wrong lol
@mdmz Hi i would love to learn in vid2vid to achieve consistency result. make one tutorial especially in this topic
Prompt outputs failed validation
VAELoader:
- Value not in list: vae_name: 'vae-ft-mse-840000-ema-pruned.safetensors' not in ['taesd', 'taesdxl']
make sure you place the model files in the right location