Video Control Map Creator (Depth, Pose, Lines)
Upload a video and the workflow saves a depth map video plus a second control map such as line art, pose or canny. A utility for preparing guidance clips.
- A10G
- Video editing
- ControlNet
- Free template
How it works
This is a utility, not a generator. It reads a video, makes control maps for each frame, and writes them out as new videos.
VHS_LoadVideo from comfyui-videohelpersuite loads the input video with the frame count and forced FPS you set (defaults are 81 frames and 15 fps). ImageResizeKJv2 resizes the frames to a size chosen by VideoResolutionSelector from the aspect ratio (Horizontal, Vertical or Squarish), the quality tier (HQ, MQ, LQ) and a VideoModel setting with options like I2V480p, I2V720p, T2V1.3B and T2V14B. These names are resolution presets for Wan-style video models.
DownloadAndLoadDepthAnythingV2Model and DepthAnything_V2 produce the depth video using depth_anything_v2_vitg_fp32.safetensors. A second pass uses AIO_Preprocessor from comfyui_controlnet_aux with the map type in EXTRA_MAP_TYPE. Two VHS_VideoCombine nodes save the two results as videos.
Tips
- Match the resolution preset to the model you plan to use the maps with.
- Lower the frame count to test a preprocessor choice before running a full clip.
- Set EXTRA_MAP_TYPE to none if you only need depth.
- Use the same FPS for the control video and the generation workflow so the frames line up.
What you can use it for
- Preparing depth videos for video ControlNet or VACE-style workflows
- Extracting line art, pose or canny maps from clips
- Batch converting clips to guidance videos
- Checking how a preprocessor sees your footage
Run it in three steps
Open the template
Click Run on ComfyDeploy. The workflow, its custom nodes and models are set up for you on a cloud machine.
Add your inputs
Provide input video and extra map type in the playground or the ComfyUI editor.
Run, share or deploy
Run it on an A10G, share it with your team as a playground, or deploy it as an API endpoint.
Frequently asked questions
What GPU does the video control map workflow need?
The recommended GPU is an A10G. It runs the Depth Anything V2 vitg fp32 model on every frame, plus one more preprocessor.
Which control maps can I create?
Depth is always made with Depth Anything V2. The EXTRA_MAP_TYPE dropdown picks an additional AIO Preprocessor map, with options such as AnyLineArt, Canny, DensePose and Depth Anything, or none.
How do I run the video control map creator as an API?
Deploy it on ComfyDeploy and call the endpoint with the input video and options. The result is returned from the API response.
