Extract Video Frame
Extract Video Frame captures the decoded picture at a selected playback time and downloads it as a PNG. It is useful for thumbnails, visual references, and frame-level inspection without uploading the source. Choose the time by previewing nearby moments: seeking may land on the closest decodable frame, and motion blur or transitions can make one instant less useful than the next.
Runs locally — files stay on your deviceHow do I save a still image from a video?
Choose the video, enter the desired time in seconds, run the tool, and download the captured frame as a PNG.
A video frame is one decoded picture
Video presents a sequence of frames over time, but compressed files do not necessarily store every frame as a complete independent image. A player often seeks from a nearby keyframe and decodes forward to produce the picture at the requested time.
How it works
Enter a time within the video's duration. The browser seeks to that point, draws the decoded frame at its native pixel dimensions, and creates a PNG. Inspect the saved image for transition frames, blur, subtitles, or overlays before using it.
Key concepts
Timestamp and frame boundaries
A time in seconds may fall between frame boundaries. The captured picture comes from the frame the browser decodes for that playback position.
PNG preserves the decoded pixels
The PNG avoids an additional lossy JPEG step, but it cannot restore detail already removed by video compression.
When to use it
- Create a thumbnail candidate from a local recording.
- Capture a visual reference for feedback or documentation.
- Inspect what appears near a specific timestamp.
When not to use it
- Do not expect a frame beyond the video's duration.
- Do not use a single still to represent image quality throughout the video.
- Do not assume the capture removes burned-in subtitles, logos, or overlays.
Worked examples
Thumbnail from an interview
- Input
Time = 42.5 seconds- Expected output
PNG of the decoded interview frame
Previewing nearby times helps avoid closed eyes or motion blur.
Capture a title card
- Input
Time just after the title transition completes- Expected output
Clean still of the full title
A slightly later timestamp avoids a partially blended transition frame.
Frame from portrait video
- Input
Vertical 1080 × 1920 source- Expected output
Vertical PNG at the decoded frame dimensions
The capture keeps the source frame shape rather than forcing landscape output.
Common mistakes
Entering a time outside the video
The player cannot seek beyond the available duration. Correction: Inspect the duration first and choose a non-negative time below it.
Expecting the exact mathematical instant
Compressed seeking and frame cadence determine which discrete frame is available. Correction: Try nearby timestamps and keep the visually correct frame.
Assuming PNG improves the source
PNG preserves the decoded capture but does not recreate detail lost during video encoding. Correction: Use the highest-quality source available.
Best practices
- Preview several nearby timestamps before choosing a thumbnail.
- Check that the captured frame contains no unwanted private information or overlays.
- Keep the video source if another frame or higher-resolution export may be needed.
Related operations and differences
Video Metadata Viewer
Choose Video Metadata Viewer when you need to inspect video duration, dimensions, ratio and file size.; use Extract Video Frame specifically for frame.
Crop Image
Choose Crop Image when you need to crop an image to an exact local pixel rectangle.; use Extract Video Frame specifically for frame.
Image Resizer
Choose Image Resizer when you need to resize images to exact dimensions locally.; use Extract Video Frame specifically for frame.
PNG to JPG
Choose PNG to JPG when you need to convert PNG images to compact JPG files.; use Extract Video Frame specifically for frame.
Limitations and privacy
- Frame extraction depends on browser support for the source format.
- Seeking precision is limited by frame cadence, keyframes, and decoder behavior.
- HDR or color-managed video may not look identical after canvas capture.
- Only one still frame is exported at a time.
The selected video is decoded locally in this browser and is not uploaded to Toolbox.
Frequently asked questions
What image format is the captured video frame?
The tool exports the decoded frame as PNG.
Why is the captured moment slightly different from the entered time?
Video consists of discrete frames, and compressed seeking depends on the browser decoder and nearby keyframes.
Can frame extraction recover higher resolution?
No. The PNG is limited to the decoded source-frame dimensions and detail.
Is the selected video uploaded?
No. Decoding and capture happen locally in this browser.