Video frame extractor for full-resolution stills

How the video frame extractor differs from a player screenshot

A screen capture records whatever the display happened to show: the player at its current CSS size, perhaps scaled for a high-density screen, sometimes with controls or a cursor in the way. This video frame extractor takes a different route. It asks the browser to decode the video at the timestamp you chose, then draws that picture into a canvas with exactly the decoded track’s width and height. A 3840 × 2160 source therefore returns a 3840 × 2160 still even when the player above is only six hundred pixels wide.

That distinction matters for thumbnails, slide decks, product documentation and reference images. It does not manufacture detail that compression already removed, and it cannot turn motion blur into a sharp photograph. It does prevent the page layout from throwing away a second round of resolution before the image is saved.

Choose the moment by sight or by timestamp

Use the player to find the scene, pause, and press Use playhead. For a timestamp from an edit list, subtitle or review note, type it directly as seconds, mm:ss, or hh:mm:ss. Decimal milliseconds are accepted, so 1:02.500 means one minute, two seconds and five hundred milliseconds. Each capture is added to the receipt without closing the video; move to another moment and repeat. One image downloads directly, while several can be returned together in a ZIP.

The result is tied to a time, not to a made-up frame number. Web browsers expose the current media time, but they do not expose one dependable source-frame index across MP4, MOV, WebM, constant-rate and variable-rate video. Calling 12.500 seconds “frame 375” assumes 30 frames per second and becomes wrong the moment a phone varies its frame rate. The receipt therefore records the timestamp the browser actually presented.

Video in, still image out

MP4

Full name	MPEG-4 Part 14
Extension	.mp4, .m4v
Format type	Video container — usually H.264 or H.265 with AAC
MIME type	video/mp4

PNG

Full name	Portable Network Graphics
Extension	.png
Format type	Lossless raster image, with alpha
MIME type	image/png

JPG

Full name	Joint Photographic Experts Group
Extension	.jpg, .jpeg
Format type	Lossy raster image
MIME type	image/jpeg

PNG keeps the decoded pixels; JPG spends some of them

Choose PNG for interface recordings, diagrams, text, line art and any image that will be edited again. PNG is lossless, so there is no useful quality slider to show. Choose JPG for camera footage when a smaller file matters more than preserving every decoded edge. Its quality control changes the new JPEG compression only; it cannot undo the video codec’s earlier compression. A high-quality JPG can still be dramatically smaller than a 4K PNG, which is why both outputs belong in the same extraction job.

If the still is destined for a format that needs transparency, remember that a normal video frame is opaque even when it is saved as PNG. The container and codec decide whether an alpha track exists, and ordinary MP4 camera files do not have one. Converting the filename cannot invent a cutout. For an existing animation rather than a video, the separate GIF frame extractor lets you choose frames from the GIF itself.

Why format support follows the browser

The page uses the same decoder your browser uses for ordinary playback. That keeps the clip local and avoids downloading a second video engine, but it also makes codec support honest: MP4 with H.264 and WebM work broadly, while HEVC inside an iPhone MOV and some tracks inside MKV depend on the browser and operating system. A file extension names a container, not the codec inside it. If this browser cannot play the picture, the page cannot truthfully extract it. Safari on a recent Apple device is often the best route for HEVC; exporting an H.264 MP4 is the portable fallback.

Nothing is uploaded while any of this happens. The selected file becomes a temporary local object URL, the decoded picture is copied to a canvas in this tab, and the PNG, JPG and optional ZIP are written here too. You can verify that behavior yourself with the steps in the local-processing network guide. When you are done, leaving or reloading the page releases the temporary video and image URLs.

When a moving result is the better answer

A still works for a cover, evidence of one state, a presentation or a reference board. It is the wrong answer when the motion explains the thing: an interaction, reaction, transition or short demonstration. In that case, trim the video into a GIF only when the destination demands an image that plays by itself. Keep the original video whenever the destination accepts it; inter-frame video compression preserves motion far more efficiently than a stack of independent pictures.