How to Generate Depth Maps from Video
Create a relative-depth video in HEVC MP4, understand its limits and review the depth maps before using them in a compatible editing or AI video tool.
Artplayer AILettura di 4 min

In questa pagina
Prima di iniziare
A video depth map represents relative depth as a scene changes over time. Artplayer AI turns an ordinary video into a depth clip you can download as HEVC MP4 for use in a compatible editor or AI video tool. This is a visual depth representation: it does not supply measured distances, a 3D skeleton or a finished character animation.
Prepare a short, representative source
Choose MP4, FLV, TS, AVI, MOV, WMV or MKV video up to 10 minutes and 10 GB. The application accepts a long edge up to 2560 pixels and a short edge up to 1440 pixels, including portrait orientation. Only SDR video is supported. The file is checked before a task can start.
Begin with a short section that includes the subjects, camera motion and overlaps you need to evaluate. Thin objects, motion blur and changing occlusion deserve particular attention when reviewing the result. Keep your source so you can compare the original scene with the depth representation.
Understand the depth file you receive
The tool uses a fixed processing mode and returns an HEVC-encoded MP4. It offers no output-resolution, frame-rate or model selector. The downloaded video represents relative depth; it is not calibrated in meters and does not include raw depth matrices or motion-capture joint coordinates.
Playback depends on the browser, operating system and codec support. If the browser cannot preview HEVC, download the file and open it in a compatible player or editor. A working download and a working browser preview are separate checks. Audio retention is not guaranteed; use the original soundtrack separately when needed.
Review the checked quote before processing
The rate is 5 credits per input minute, with fractional minutes included in the calculation. Round the total up to whole credits, with a minimum of 3 credits per task. A 30-second clip costs 3 credits; a 61-second clip costs 6 credits.
The local estimate is provisional. After upload, the server checks the source and produces the quote accepted when starting the task. That accepted charge stays fixed during processing. The task continues in the cloud after you leave the page, and you can return through My files.
Review the result before editing with it
Watch the depth clip alongside the source, including transitions and moments when one object passes in front of another. Review the complete clip for unstable boundaries or abrupt changes. A single clear frame cannot establish that the representation is consistent over time.
Download the result before its retention window ends. Uploaded originals remain available for 24 hours after upload completes, and results remain available for 7 days after completion. The cover illustration explains depth layers; it is not a supplier output or a quality comparison.
- Check the actual duration, frame rate and dimensions in your player or editor.
- Verify HEVC support and inspect the entire clip, especially overlaps and fast motion.
- Treat the video as relative depth information, not measured geometry or skeletal animation.
Vuoi provare lo strumento?
Scegli il file, controlla le impostazioni disponibili e il preventivo in crediti, poi avvia quando sei pronto.
Extract video depth