Skip to content

Video frame and thumbnail contact sheet extractor

Turn any MP4 or WebM clip into crisp full-resolution still frames, ZIP sequences, and labeled contact sheets.

Your files are open only in this tab. Nothing is uploaded.

Video input

Extract individual PNG, JPEG, or WebP frames and compose contact sheets locally in this tab.

Drop a video file here

MP4, WebM, MOV, or MKV decoded in this browser.

How to use Video frame extractor

  1. Load a local clip. Drop an MP4, WebM, MOV, or MKV file onto Drop a video file here, or click Load sample video to inspect six synthetic frames offline.
  2. Choose your capture mode and format. Pick Extract every N seconds with an interval between 0.2 and 120 seconds, or switch to Capture current frame and select PNG, JPEG, or WebP.
  3. Run extraction and review the grid. Click Extract frames or pause the player and click Capture current frame to append still shots with MM:SS.t badges.
  4. Build a contact sheet or ZIP archive. Set Columns, Tile width, Header title, and Show timestamps on tiles, then click Build contact sheet (.png) or Download all frames (.zip).

Your camera, microphone and recordings stay on this device. Nothing is uploaded or streamed anywhere.

Extracting video frames by interval or playhead

Capturing sharp thumbnails from MP4 or WebM footage lets you grab full resolution PNG, JPEG, or WebP frames offline without uploading large video files to third-party servers. A local video frame extractor decodes frames directly inside your browser tab, keeping unreleased product demos and private interviews strictly on your device.

Sweeping recordings with Extract every N seconds

Dropping an MP4, WebM, MOV, or MKV file reads native dimensions and duration from browser metadata, while clicking Load sample video generates 6 deterministic 320 × 180 frames spaced 1.5 seconds apart. In Extract every N seconds mode, the Interval (seconds, max 60 frames) input accepts values from 0.2 seconds up to 120 seconds, defaulting to 2 seconds.

The interface builds a timestamp schedule starting at 0 seconds up to 60 frames maximum and stops 0.04 seconds before the end so trailing seeks remain valid. You can halt long sweeps anytime using the Cancel button beside the progress indicator.

Pausing on exact moments with Capture current frame

Switching to Capture current frame lets you scrub the muted HTML5 <video> player to any reaction shot, slide transition, or subtitle cue before clicking Capture current frame. Each capture appends a numbered frame card (FRAME 1, FRAME 2) showing its MM:SS.t badge, pixel dimensions, and byte size.

Video timelines structured under the Library of Congress MPEG-4 File Format guide index sample timestamps inside container tables so browser decoders seek cleanly between keyframes. If you still need raw footage before pulling thumbnails, record a fresh webcam clip first in the same suite.

How does canvas capture keep full frame resolution?

Every captured frame draws onto an offscreen canvas sized to the video track’s native videoWidth and videoHeight before encoding, preserving every pixel of source detail. Only storyboard thumbnails scale down to 640 px max width before entering the worker for contact sheet tiling.

Encoding PNG, JPEG, and WebP via canvas toBlob

Individual frame downloads pass the full-size <canvas> surface into the MDN HTMLCanvasElement toBlob API using your chosen Frame image format. Lossless PNG (.png) preserves pixel-exact graphics and text overlays, while JPEG (.jpg) and WebP (.webp) encode at 0.92 quality to save disk space.

Each decoded frame simultaneously creates an in-memory object URL for instant grid previews and one-click Download buttons. Unwanted shots drop cleanly with the remove (×) button, automatically revoking object URLs and clearing stale contact sheets.

Handling codec support and variable frame timing

Browser playback engines natively decode H.264 MP4 and VP8/VP9 WebM streams, whereas specialized HEVC or ProRes MOV files depend on host OS codec availability. Seeking waits up to 1,500 ms for the seeked event and 800 ms for loadeddata if decoding buffers stall momentarily.

Infinite-duration streams recorded from live pipes trigger a temporary seek probe so valid timeline bounds resolve before extraction begins. When working with looping animated graphics instead of video containers, switch to our animated GIF frame splitter for disposal-aware frame layers.

Contact sheet tile layout, formats, and export options

Selecting the right format and grid geometry balances output byte size against storyboard readability across multi-minute recordings. Lossless storyboard grids follow the W3C Portable Network Graphics Specification so timestamp badges remain sharp at 240 px, 320 px, or 480 px tile widths.

Control Options or range Default setting Behavior in tool Best use case
Extract every N seconds 0.2 to 120 seconds (0.5 step) 2 seconds (60 max) Seeks sequentially and appends numbered frames Automated timeline previews and action sweeps
Capture current frame Any paused timestamp Manual scrubber Appends current playhead frame to existing list Picking exact YouTube thumbnails or slide stills
Frame image format PNG, JPEG, or WebP PNG (0.92 lossy alt) Sets canvas.toBlob encoder for individual downloads Lossless UI grabs vs compact photo shares
Columns & Tile width 2–6 cols; 240 / 320 / 480 px 3 cols at 320 px Controls grid density and PNG pixel dimensions Compact PDF review sheets vs detailed frames
Show timestamps on tiles Checked or unchecked Enabled Draws FRAME N and MM:SS.t pills on tiles Editorial timecode logging and client notes

Deleting sensitive microphone audio before distributing preview clips works cleanly alongside still extraction when you keep confidential clips silent before sharing in a separate tab. Every generated asset stays isolated inside browser memory until you save it manually.

Tip: Set Frame image format to PNG (.png) when capturing software UI screencasts or slide text, and keep Columns at 3 with Tile width at 320 px for crisp storyboard contact sheets that remain readable on laptops without exceeding reasonable file sizes.

Building storyboard PNGs and ZIP archives locally

Creating a labeled video contact sheet generator output or packaging dozens of extracted stills happens asynchronously inside a background worker so browser tabs never freeze. This architecture lets editors assemble storyboard sheets and multi-frame archives on flights or air-gapped workstations.

Bilinear RGBA tiling inside the background worker

Clicking Build contact sheet (.png) transfers downscaled RGBA pixel buffers (640 px max source width) to everyday.worker.ts over the MDN Web Workers API documentation channel. The pure TypeScript composer calculates aspect-correct row heights, paints a #111827 backdrop with 16 px gutters, and runs bilinear resampling per tile.

When Header title is populated, a 5×7 bitmap font stamps your custom heading across the top margin, while rgba(15, 23, 42, 0.8) pill overlays display FRAME labels and MM:SS.t timecodes along tile bottom corners. Clicking Download contact sheet saves <stem>-contact-sheet.png directly from worker bytes.

Bundling frame sequences for client review offline

Saving 20 or 60 individual stills one by one slows down review cycles, so Download all frames (.zip) packages every extracted PNG, JPEG, or WebP file into <stem>-frames.zip inside worker memory. Extracted filenames increment cleanly (-frame-01, -frame-02) to preserve chronological playback order inside desktop viewers.

Linux build pipelines or archival teams needing .tar.gz bundles instead of standard .zip archives can repackage the extracted ZIP immediately with our TAR & .tar.gz Extractor. Closing or reloading the page revokes every temporary blob URL and clears source footage from RAM.

Questions

Are extracted video frames saved at full resolution?

Yes. Individual frame downloads paint the decoded video onto a canvas sized to its native coded width and height without resizing down. Only storyboard contact sheets downsample tiles to your selected 240 px, 320 px, or 480 px column width.

What appears on the generated contact sheet PNG?

The Web Worker arranges your captured frames into a 2 to 6 column grid with 16 px padding over a dark slate background. When Show timestamps on tiles is enabled, each thumbnail displays a FRAME label and MM:SS.t badge.

Can I capture frames offline from confidential client footage?

Yes. Your video opens through a local object URL inside the HTML5 video element, and canvas serialization happens strictly inside tab memory. Contact sheet composition and ZIP bundling execute inside a dedicated Web Worker with zero network uploads.

Which video formats work for browser frame extraction?

Any MP4, WebM, MOV, or MKV file whose video codec your browser supports natively can be decoded and captured. Standard H.264 MP4 and VP8 or VP9 WebM clips open reliably across Chromium, Firefox, and Safari.

How many frames can interval mode extract in one pass?

Extract every N seconds steps across the video timeline from 0.2 seconds to 120 seconds per step and caps each automatic sweep at 60 frames. For longer movies, raise the interval or capture specific keyframes manually with Capture current frame.

Why does JPEG or WebP export produce smaller files than PNG?

PNG stores lossless pixels, which preserves crisp interface text at larger byte sizes. Choosing JPEG or WebP encodes frames with 0.92 lossy quality, shrinking photographic video captures substantially for web thumbnails and quick client previews.

Updated