Відео

I like it. Just wish I could do this on my 1060.

perfect

Great tutorial, thanks for this! Question tho: Is there a way to feed it an image to be animated like the sourced video? Like say I want to animate a specific, original character singing. Can I provide an image of said character and a video of someone singing and have comfy replace that person with the character? Or those Animatediff works through prompts only at the moment?

Nice

Does anyone know how to save the videos on phone?

"Not Joke" My Galaxy A54 got the Galaxy AI too but (unofficial)

What? Have you been following sora, kling ai, etc.?

✈️

Prompt for this , plz😊

Am I supposed to be impressed? Honestly, we really seem to be hitting a wall with the transformer architecture but no one will admit it.

Ok. (There’s not much to say)

Wtf why you look like a AI, your eyes and mouth look wierd when moving

Great! Heading there!

I would dump water on their servers

Sad, they are creating AI to kill creativity. Now youtubers like you who use ai to write videos all sound the same and boring, and art by ai is genereic as fuck too. Sad to see

This is dope. Check out my generation!

谢谢你给我发了邮件 我今天才看见 非常感谢

Thanks! Been looking for a tut on AnimateDiff!!!

Your explanation is also created with AI. 🧠💡🌐

hey i have rtx 3080 10gb and im stuck on : Requested to load AnimateDiffModel Loading 2 new models 0%| | 0/24 [00:00<?, ?it/s]

how can i control clip vision on this workflow my friend ?

Use the prompt below: "I'm working on this cool short film project. My plan is to use MidJourney to generate a bunch of images. Then, I'll run those images through another AI called Gen2 to bring them to life. After that, I'll piece all those animated scenes together to tell a really captivating story. Here's the story idea: A 30 year old man who's stuck out in the middle of the sea during a sea storm, they finally spot an island in the distance. Here’s where I need your help: 1. Provide 10 simple comma-separated keywords describing the main character; what they look like and what they're wearing 2. Develop the story further and break it down into different shots, for each shot, provide 10 simple comma-separated keywords in this this structure: - Camera angle and position - Action, environment, time of day, lighting. - Don't include: Camera movement, sound."

Is it possible that I put the real person's image into the CHATGPT or Midjourney, so I make some shots based on that person? I am currently working on a new short film, and I wanna make some footages in the film by AI imaging technology.

mine does not come with the seed node. i have been looking for it in the nodes list to see if i need to install it but cant seem to find it. any suggestions?

I want to try this, thanks man, love your channel

Would it be easier to use rotobrush for the closeup shot mid air?

Thank you so much for your lessons, I made a short AI film using your lessons, I set up IPADAPTER and AnimateDiff and got an amazing result, I like it very much ua-cam.com/video/MFmTsgkLXRA/v-deo.html

ua-cam.com/video/PPjATWJqHAI/v-deo.htmlsi=we3yqh3vkhwSVirW

How do I do this?

Is there a new and better workflow for VideoToVideo with ComfyUI? Maybe time for a next videooo?

Hi, need your help please, I just dragged and drop a workflow into comfyUI and got this error : When loading the graph, the following node types were not found: MiDaS-DepthMapPreprocessor VHS_VideoCombine VHS_LoadVideoPath. And i am trying to install missing but not working, please help me. Thank you bro.

Poor guy😂 the s21 just got galaxy ai

does this works on a 1070 gtx 8gb vram ?

Ai name

....ok, I was totally on board until it came out looking cgi. Did you just use the clouds? 😅

How many vram and ram should i have for that? I have 32 ram- 8vram

Any suggestions on how to do a longer video? I want to use more than 4 images, how do i add nodes?

I'm having some issue in IPAdapter model loader, it keeps showing undefined/null. Any solution?

Fixes for current version - June 2024: IPAdapter won't work in folder 'ComfyUI_IPAdapter_plus' => Go to 'ComfyUI\models' folder and add a folder named 'IPAdapter' and place the IPAdapter plus there. (Now you can find the IPAdapter) Next you'll still have two red nodes in video reference and keyframe. Change those with 'IPAdapter Advanced' nodes (double click to search for nodes) and link the lines to these new nodes then remove the old ones. Make sure all connections are made like on the broken nodes.

i cant load the IpAdapter Nodes, it just dont show up in even if i refresh, can you help me? i did exacly that you did

Thank you for the excellent explanatory video. Due to my low PC specs, I got an error about memory limitation and could not proceed any further. Is there any instructions or notes on how to use the Google Colab version of comfyui?

What editing app nam

Hey, i am using Comfy UI in Think Diffusion for the first time. When i am adding your json file inside it, then they ask me to install the missing nodes. Once i install those missing nodes and restart the UI, then still it shows that nodes are missing. I a m unable to solve this problem. Can you please help me sir?

😢😢😢😢😢😢

why does the final output video turn super dark when i use super bright images??

What settings can I use for 1.5 I have 3090 RTX with 8gb ram but keeps getting the cuda memory error with the settings in the video.

Hi, when i'm at the "cmd" step and i paste "git clone [link]", it tells me that the word "git" is not recognized as a command. What yall guys suggest me ? (my pc is set in french if it helps)

🔥🔥🔥🔥

Tutorial 🥺🥺

When loading the graph, the following node types were not found: MiDaS-DepthMapPreprocessor VHS_VideoCombine IPAdapterApply VHS_LoadVideoPath Nodes that have failed to load will show as red on the graph