Initializing... drag & drop files here
Supports: MJPEG
Motion JPEG is the one video format where pulling a still costs you nothing structurally: every frame is already a complete, independently coded JPEG, so there is no predicted frame to reconstruct and no reference picture to guess at. This tool decodes the frame you ask for, hands it over as a lossless intermediate, and writes it out as a TIFF — the format archives, print shops and scientific workflows ask for. The one setting that decides whether the result is worth having is Compression Type, and its default is not the lossless one.
.mjpeg extension, so rename a .mjpg before uploading; the real format is then detected from the file's own bytes, which means a renamed AVI or QuickTime file carrying Motion JPEG is read correctly too..mjpeg, use Multiple Screenshots and pick a Capture Rate instead — see the frame table below for why..tiff or .tif. Convert and download — no sign-up, no watermark.TIFF is a container that can hold several different compression schemes, and this page exposes ten of them. They are not interchangeable. The verdicts below come from running the same extracted 640×480 Motion JPEG frame through each one and comparing the result pixel-for-pixel against the source frame.
| Compression Type | Lossless? | Relative size | Verdict |
|---|---|---|---|
| JPEG (the default) | No | Small | Re-compresses an already-compressed frame. Avoid unless size beats fidelity |
| LOSSY | No | Same as JPEG — it is an alias for it | No reason to pick it over JPEG |
| LZW | Yes — bit-exact | Moderate | The safe lossless choice; readable by essentially every TIFF application |
| DEFLATE | Yes — bit-exact | Clearly smaller than LZW | Best lossless size-to-compatibility ratio; the same algorithm PNG uses |
| ZSTD | Yes — bit-exact | Smaller still than Deflate | Excellent, but a modern extension — older readers will reject it |
| PACKBITS | Yes — bit-exact | Very large on photographic content | Simple run-length coding; only sensible for flat graphics |
| NONE | Yes — bit-exact | Largest by a wide margin | Raw pixels. Choose only when a downstream tool demands no compression |
| WebP | No | Smallest overall | Lossy, and WebP-in-TIFF is poorly supported outside libtiff-based readers |
| CCITT Fax 4 | n/a | n/a | Fails on a colour frame. It is a 1-bit bilevel scheme only |
| JP2K (JPEG 2000) | n/a | n/a | Fails. The encoder does not implement scanline JPEG 2000 output |
A useful historical note explains the default: the TIFF 6.0 specification (Adobe, final June 3 1992) requires Baseline readers to handle only three schemes — uncompressed, CCITT Group 3 1-D, and PackBits — and it explicitly records that "due to licensing issues, LZW compression support was moved out" of Baseline. Those patents expired long ago and LZW is universally supported today, so the historical caution no longer applies to you. Pick LZW or Deflate without hesitation.
A raw Motion JPEG stream stores no timestamps, no duration and no declared frame rate — it is just concatenated JPEGs. Decoders therefore assume a nominal rate, which has a direct consequence for the Time (seconds) box.
| Frame Selection setting | On a raw .mjpeg stream |
On a renamed AVI / MOV / MKV |
|---|---|---|
| Specific Frame, Time = 0 (default) | Works — returns the first frame | Works — returns the first frame |
| Specific Frame, Time > 0 | Returns nothing; the stream has no timestamps to seek against | Works — returns the frame at that timestamp |
| Multiple Screenshots, any Capture Rate | Works — walks the decoded stream and writes one file per interval | Works |
| Output when multiple frames are selected | Separate numbered image files, one per frame | Separate numbered image files, one per frame |
So on a genuine elementary stream there are exactly two reliable routes: take frame one, or take a series with Multiple Screenshots and keep the one you want. Capture Rate runs from 0.1 seconds up to 10 seconds per frame, so a fine rate on a short clip gives you close to every frame. If your source is really a container, convert it from that file instead and Time will behave normally.
No, and nothing can. The frame was compressed as a JPEG when the camera recorded it, and that loss is permanent. What a lossless TIFF does is stop further loss: it preserves the decoded frame exactly as it came out of the JPEG decoder, so every later crop, level adjustment or print pass starts from the same pixels. Saving as TIFF with the default JPEG compression throws that advantage away by re-compressing the picture a second time.
Structurally, yes. H.264, HEVC and AV1 code most frames as predictions from neighbouring frames, so an arbitrary still is a reconstruction assembled from a keyframe plus motion data and residuals, and it can carry blocking or smearing that the keyframe does not. Every Motion JPEG frame is a self-contained image, so the frame you extract is the frame that was recorded — which is exactly why surveillance, microscopy and machine-vision systems still use MJPEG despite its size.
The default favours small files, which suits the majority of TIFF conversions on the site, and it is applied consistently across image targets rather than being tuned per source format. On this page it is usually the wrong pick: the frame arrived as a decoded JPEG, and re-encoding it as JPEG-inside-TIFF adds a second generation of loss for no structural benefit. Change it to LZW or Deflate and the TIFF is a faithful copy.
Both are bit-exact, so choose on compatibility versus size. LZW has been in TIFF since Revision 5.0 in 1988 and every TIFF reader written in the last thirty years handles it, which makes it the right answer when the file is going to a client, a print shop or an archive. Deflate is the same algorithm PNG and ZIP use, is supported by all current software, and produces distinctly smaller files on photographic content. Use LZW when you do not control the recipient, Deflate when you do.
All ten, on one 640×480 frame extracted from a raw Motion JPEG stream, each output compared pixel-by-pixel with the source frame. LZW, Deflate, ZSTD, PackBits and None came back with zero differing pixels — genuinely bit-exact. JPEG and WebP compression both altered a large fraction of the pixels, confirming they are lossy. CCITT Fax 4 failed outright with "Bits/sample must be 1 for Group 3/4 encoding", and JP2K failed with "Compression scheme 33004 scanline encoding is not implemented". On that frame, Deflate produced a file under half the size of LZW while remaining identical to it in content.
Not from this page. TIFF does support multiple images in one file, but the conversion runs frame-by-frame: each selected frame becomes its own .tiff. Choosing Multiple Screenshots gives you a numbered set of separate files, which you can assemble into a multi-page TIFF afterwards in ImageMagick, Photoshop or a dedicated TIFF tool. If what you want is a moving result rather than stills, use MJPEG to GIF or MJPEG to APNG.
Because DPI is meaningless for a video frame. Dots-per-inch only describes how many pixels a device should place per physical inch when printing; it changes no pixel data. A camera frame has a pixel count, not a physical size, so the useful control here is Image resolution — Preset Resolutions, a percentage, or explicit Width x Height. Set the print size in your layout application, where the paper dimensions are actually known.
.tiff or .tif in the File extension list?They are the same format; only the filename differs. .tif is the three-letter form that survived from MS-DOS filename limits, and a few legacy Windows applications and some scientific instruments still look specifically for it. .tiff is the more common modern spelling. Nothing inside the file changes, so pick whichever your destination software expects and rename freely if you guess wrong.
It travels over an encrypted connection, is converted on our servers, and is deleted automatically after a few hours. Your files are never shared, never published, and never used for anything except producing your download. No account, no watermark, and no limit on how many frames you pull.