
Krita's documented animation video-export route does not preserve transparency; transparent regions render as black. Export a PNG image sequence when a compositor needs alpha, and make a separate review movie with a deliberate background. A transparent canvas does not guarantee a transparent video file.
Key Takeaways
- Check whether the receiver needs an overlay asset or a finished opaque movie.
- Preserve transparency in an appropriate image-sequence handoff.
- Review semitransparent edges over contrasting backgrounds in the receiving tool.
Confirm that the source has real transparency
Hide any painted background layer in a copy of the project and inspect the artwork. A white painted layer is still opaque, and a checkerboard drawn into the picture is still image content. Use the checkerboard diagnostic guide if the source came from an image generator or a downloaded preview.
Once actual transparency is confirmed, separate the source state from the export route. Krita's Render Animation documentation explicitly notes the video transparency limitation. Changing the apparent canvas background is not a substitute for choosing a delivery that carries the required alpha information.
Ask the receiving editor what formats and frame timing they support before rendering the full animation. A PNG sequence can preserve per-frame transparency, but the receiver still needs to know how to assemble and time the files.
Create different files for different review tasks
For the compositing asset, export the required image sequence and include every necessary frame for the agreed timing. Keep a note of FPS, start and end frame, dimensions and whether the sequence contains every output frame or only unique drawings. Those choices determine how another editor reconstructs the motion.
For a review movie, place the artwork over a deliberate background appropriate to the review. This lets the client assess action and timing without mistaking black transparency fill for part of the intended design. Keep that backing separate in the native project so it does not contaminate the transparent deliverable.
Do not remove black from the encoded movie as the first repair. A keyed reconstruction can damage dark artwork and soft edges, and it cannot reliably recover the original partial transparency. Return to the source and export the alpha-bearing images instead.
Inspect the compositing result
Imagine a hypothetical animated smoke flourish intended to sit over several wedding-film shots. Open representative exported PNG frames in the receiving compositor and place them over both a bright image and a dark image. Examine wisps, soft edges and dark painted details, then play the assembled sequence through its first and final exposure.
The proposed acceptance criteria are that intended background areas remain transparent, partial edges blend as intended and the motion duration matches the approved review. A single clean thumbnail is not enough to establish those properties across the sequence.
Use the render range guide if the delivered interval needs trimming. Preserve the native project, transparent frames and backed review movie as distinct deliverables. This gives collaborators both a usable compositing source and a clear visual reference, without relying on a video route that the application documents as opaque.
Related tasks
- Why Does the GIMP Eraser Paint Black or White Instead of Transparency?
- Why Does My Transparent AI Image Still Show a Checkerboard?
- How to Render and Verify a Repaired Frame Range in Krita
Sources and scope
- Render Animation: Video export is made from a prerendered PNG sequence; exporting the image sequence as well preserves it for a failed encode.; Only Unique Frames saves keyframes rather than every timeline frame.; Video transparency is not supported by this rendering route; transparent areas render black.; On Android the operating-system encoder is used, not FFmpeg; do not install FFmpeg for this workflow.
Official-documentation explanation with a proposed diagnostic example.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: AI-generated conceptual editorial illustration; not a real application screenshot, observed output or before-and-after test.