Video Frame Extractor: Local PNG Captures

Extract decoded PNG frames from a local video

Video Frame Extractor loads a file from your device into the browser’s native video decoder. You can play, pause, seek and capture a displayed frame as a PNG, or extract a bounded gallery at regular time intervals. The existing settings panel, timeline and gallery remain the workspace. This tool does not upload your selected video to a conversion server, download a remote video or transcode an unsupported codec.

Use it to inspect a clip you are allowed to process, choose a presentation still or create a small visual sequence for a review. A captured frame preserves the decoded pixel dimensions within the stated limits, but it cannot recover detail that the source codec discarded. PNG encoding avoids adding another lossy image-compression step; it does not turn a compressed source into an original camera image.

Load, seek and capture

Choose a nonempty video no larger than 250 MiB. MP4 and WebM are common containers, but container names do not guarantee decoder support: the video codec inside the file must also be supported by your browser. A file with a recognized extension and no MIME type may be accepted for decoding; a non-video file or unreadable clip is rejected. Wait for readable duration and dimensions before capturing.

Use the play control for an overview, then pause and move the timeline to the relevant moment. The small step buttons move approximately 33 milliseconds, or one second for the larger steps. A 33-millisecond step is close to one frame at 30 frames per second, but it is not a frame-index operation. Variable frame rates and other frame rates can select different decoded images than an exact frame-number workflow would.

Capture Current Frame adds the current decoded image to the gallery. Seeking must finish before capture; the tool does not substitute an arbitrary fallback image when decoding is unfinished. Each gallery item displays the video timestamp used for capture and offers Download PNG. A browser can resolve nearby requested timestamps to the same encoded frame, especially where frame spacing or keyframe seeking limits precision.

Original interval exercise and expected results

Make a short clip that you own with a changing colored background and a visible time label. For example, use a twelve-second clip and the five-second extraction interval. Auto-Extract Interval should request timestamps zero, five and ten seconds, producing three gallery items when all seeks decode successfully. The endpoint at twelve seconds is excluded because extraction visits timestamps strictly before the duration.

Compare the displayed images with playback at those moments. The image at five seconds should contain the frame the browser decodes near that time, not the initial frame repeated because a timer fired too soon. A synthetic fixture can verify this behavior without uploading somebody else’s media. If the encoded duration is slightly different from twelve seconds, use the actual metadata duration when predicting the number of samples.

Change the interval to one second for a shorter owned clip and compare the gallery count. The proposed count is the duration divided by the interval, rounded upward, plus existing gallery frames. The tool rejects a request that would exceed sixty total frames. Choose a larger interval, use a shorter clip or download and clear the current gallery. It does not silently truncate a promised complete interval extraction.

Local processing, cancellation and memory

The file is referenced through a browser object URL and decoded locally. Canvas captures and PNG data are held in application memory until you clear the video, choose another file or reload. The selected video bytes are not submitted to a SimplyUtils media-processing endpoint. Site-wide network requests are separate from this local pipeline, so local processing does not imply that the entire website is offline.

Automatic extraction pauses playback and temporarily mutes the video. It restores the previous mute setting afterward. Clearing or replacing a video cancels pending extraction so frames from an old file cannot appear in the new gallery. Closing the page also cancels the work and releases the file reference. Existing downloads on your device remain yours to manage separately.

Memory usage grows with image dimensions and the gallery count. The tool accepts dimensions up to 8192 pixels on either side and sixteen million pixels in total, and keeps at most sixty frames. Those are application budgets, not a promise that every device can comfortably hold that much data. On a low-memory phone, start with a small clip and a few captures before attempting a larger batch.

Errors and practical limitations

A video must have a finite positive duration and readable dimensions. Live-stream-style or malformed duration metadata is unsuitable for interval extraction. Each seek waits for a decoded frame and expires after ten seconds if it cannot complete. A timeout is an error, not a successful capture of an earlier frame. The gallery is updated only after the requested automatic sequence completes successfully.

If the player remains blank, test the file in your browser’s native player and check the codec. Try exporting an owned clip as a widely supported MP4 or WebM with normal duration metadata. Renaming a file extension does not convert its codec. If playback works but seeking is slow, try a shorter clip, reduce its resolution or re-encode with more useful seek points in an appropriate local editor.

This page does not extract audio, package a ZIP, perform scene detection or establish permission to reuse a still. Download only frames you are entitled to save or share. For an error report, provide the browser version, container and codec if known, clip duration, dimensions, selected interval and exact message. A small non-confidential reproducer is more useful than sending a private full-length video.