Depth Map Extractor
The Depth Map Extractor node turns a video into its depth map: a black and white video in which things close to the camera are bright and things far away are dark.
What it does
Connect one video and run. The depth map comes out as a video with the same length, shape and frame rate as the one you put in, without sound. It goes on to the next node like any other video, for example into an Output node.
Inputs and outputs
- Video: the one video to read. Pictures, sound and text won't connect, and an Upscaler only connects when it is set to Video.
- One video at a time: a Media Library connects only when it holds exactly one video. With two files in it, the wire is refused with "Video takes one video, but Media Library holds 2."
- The input takes one connection. Once one is connected it won't take a second, and hovering it tells you to disconnect it first. A workflow that already has two on it won't start until you disconnect one.
- Depth video (output): the depth map. It connects wherever a video does.
There is no Flow input. The node runs as soon as the video it reads is ready.
If no video is connected, the workflow check says "connect a video" for this node before anything runs.
Limits
- Videos up to 30 seconds. A longer one is refused with a message that says how long your video is. Nothing is cut off.
- A file that is not a video is refused with a message.
Settings
None. There is no model to pick and no prompt to write, and the node needs no provider key.
Tips
- While the node works, the card shows the same moving fog as the other generation nodes. When it is done, the depth video plays in the card.
- Once a result exists, the Download button under the card becomes active.
- To keep the depth video in your Media library, connect the node to an Output node.