I plan to watch: Ep 16 - ComfyUI, AnimateDiff, ControlNet & Blender Animated Masks - https://www.youtube.com/watch?v=QIhxY21inzk
This should cover using controlNets, animation diffusion and masks from blender. I'm not a blender guy, and thats ok, im mostly interested in seeing him use an image sequence as a mask to make cool shit. What are control nets? They lets you do cool things like automatically identify poses or depth mask out people in a photo. What is animation diffusion, its a sampler that lets you make image sequences that are coherent, so an animation. What is blender? Its a 3d modeling application that i know very little about, thats not needed. What should i have downloaded/what hardware do i need?
Hardware:

• Computer.
  Preferably an NVIDIA gpu, but if you have an AMD then it should still work for you but you will be rendering on the CPU level.

Downloads:

• ComfyUI https://github.com/comfyanonymous/ComfyUI/releases/download/latest/ComfyUIwindowsportablenvidiacu121orcpu.7z <- direct download of the portable version. Unzip and place where you want. If you are a windows user like me, place it in the root of your C folder cuz reasons.

• A model : https://civitai.com/models/25694/epicrealism <- put in models/Checkpoint

• a VAE (just in case) : https://huggingface.co/stabilityai/sd-vae-ft-mse-original/blob/main/vae-ft-mse-840000-ema-pruned.ckpt <- put in your VAE fold in models/VAE/

• ComfyManager : https://github.com/ltdrdata/ComfyUI-Manager <- put this in your custom nodes directory in the compfyUI folder. All custom nodes comes from here, makes shit super easy to add to and modify.

• Once comfyManager is installed, restart Comfy and then you should see the manager button. Clicking that will open an options window that will let you install custom nodes and missing nodes. My suggestion is to click the custom nodes and install animation diffusion, AnimateDiff-Evolved, and Crystools.

• You will need models for your animation diffuse to work, so to install those, there will be another button in the manager for downloading models. Then search for mmsdv15. That should automatically put it in the correct directory for you.

• You will want control nets installed (ignore the warning that the manager gives). then you will want the control net models installed, https://huggingface.co/lllyasviel/ControlNet/tree/main/models. its about 40 gigs for all of them, but you can get away with just depth and pose.

• Finally, if you want some extra stuff to have fun with, try some loras like https://civitai.com/models/82098/add-more-details-detail-enhancer-tweaker-lora <- put it in your lora folder.

• even EXTRA stuff, you can add an embedding like https://civitai.com/models/82839?modelVersionId=88012 <- put that in your embedding and when enters negatives for what you DONT want to show up in the generated work, you can add this embedding and that will help refine the image.

It may seem like a lot, but really its not that bad once you get into it. Everything is visual, all node based. So you just link up boxes to other boxes. Its fun and I'm just trying to have a good time making shit. Join me while we explore a new horizon of art akin to the camera or the moving picture. Thanks!