Reset to defaults
Need help?MKV to TIF — Pulling a Frame Out as a Tagged Image File
This is a video-to-still conversion, not a video conversion. It decodes your Matroska file, takes either one frame at a timestamp you choose or a whole series of frames across the clip, and writes each one as a TIF image. TIF is the format print, archival, forensic, and scientific workflows standardise on because it can store a frame with no compression loss at all — but that is a property of the compression setting you pick, not of the extension, and the default on this page is a lossy one. The reference below covers what the format actually is, what each control does, and which options quietly fail.
TIFF at a Glance
| Property | Value |
|---|---|
| Specification | TIFF Revision 6.0, "Final — June 3, 1992", authored at Aldus Corporation; stewardship passed to Adobe |
| Extensions | .tif and .tiff are the same format; the File extension control on this page lets you write either |
| Structure | A tag-directory container (IFDs) rather than a single fixed layout, which is why one TIFF can hold wildly different data than another |
| Compression | Selectable per file — lossless, lossy, or none, set by the Compression Type control |
| What this page writes | A single-frame TIF per captured frame, produced by libvips, with the compression scheme you choose |
| Metadata retained | EXIF, XMP, IPTC and ICC data the pipeline carries through |
| Colour source | Whatever the decoded video frame is, converted to 8-bit RGB — there is no bit-depth control on this path |
| Practical role | Editing masters, print, evidence stills, anything that will be re-saved repeatedly |
Compression Type: What Each Option Actually Does
This is the single most important control on the page, and its default is not the one most people assume. JPEG is preselected, and JPEG is lossy — which defeats the reason most people reach for TIF in the first place. Note also that LOSSY and JPEG are the same scheme in this list; both resolve to JPEG compression. The measurements below come from our own runs writing a single 1920×1080 frame through each setting.
| Compression Type | Lossless? | Behaviour on a 1080p colour frame |
|---|---|---|
| JPEG (default) / LOSSY | No | Smallest of the usable options, but re-compresses the frame — pick something else if pixel fidelity is the point |
| LZW | Yes | The safe lossless default; opens in essentially every image editor ever written |
| DEFLATE | Yes | Lossless and typically smaller than LZW; needs a reasonably modern reader |
| ZSTD | Yes | Lossless, smallest of the three lossless schemes in our runs, newest and least universally supported |
| PACKBITS | Yes | Lossless but a simple run-length scheme — on photographic frames it barely beats no compression at all |
| NONE | Yes | Raw raster. A 1920×1080 RGB frame lands at 6,220,800 bytes by definition — three bytes per pixel, no header savings |
| WebP | No, as configured here | WebP's lossless mode is off in this pipeline, so this is a lossy scheme despite living inside a TIFF |
| CCITT Fax 4 | — | Fails on colour frames. It is a bilevel (1-bit) fax scheme and rejects anything with more than one bit per sample |
| JP2K | — | Fails here. The JPEG 2000 scanline encoding path is not implemented in the TIFF writer |
If you want the thing TIF is famous for — a bit-exact copy of the decoded frame — change Compression Type to LZW, or to DEFLATE/ZSTD if your downstream tool reads them.
Frame Selection Reference
| Setting | Default | What it produces |
|---|---|---|
| Specific Frame | Selected, with Time (seconds) set to 0 | One TIF, taken at the timestamp you type. Leaving it at 0 grabs the very first frame — which on a film is often black |
| Multiple Screenshots | Capture Rate of "1 second per frame" | A separate TIF for every captured frame across the whole clip. Rates run from 0.1 s per frame up to 10 s per frame |
| Image resolution | Keep original | Presets are named by height (4320p down to 16P) and the width follows the source aspect; Width, Height and Width × Height are also available |
| Image Compression | Quality Preset at "Very High (Recommended)" | Specific file size is the alternative when you need a byte budget rather than a quality target |
How to Convert MKV to TIF
- Upload Your MKV File: Drag your
.mkvonto the page or click "Add Files" to browse. Queue several videos to run them through identical settings. - Choose the Frame in Frame Selection: Pick Specific Frame and enter the Time (seconds) for a single still, or Multiple Screenshots and a Capture Rate to export a TIF for every captured frame across the clip.
- Set the Compression Type: Change Compression Type from its JPEG default to LZW for a genuinely lossless still, or DEFLATE/ZSTD for smaller lossless files. File extension switches the output between
.tifand.tiff, and Image resolution downscales if you do not need full frame size. - Convert and Download: Click Convert and download your TIF. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours — no sign-up, no watermark.
Frequently Asked Questions
Is the TIF this produces actually lossless?
Only if you change the Compression Type. The control ships preselected on JPEG, which is a lossy scheme, so a straight run with default settings gives you a re-compressed frame in a TIF wrapper. Switch it to LZW, Deflate, ZSTD, PackBits, or None and the frame is written exactly as the decoder produced it. This is the single thing worth checking before you extract stills you intend to grade, analyse, or archive.
What is the difference between .tif and .tiff?
Nothing at the file level — they are the same format, and the byte contents this page produces are identical either way. The three-letter form is a holdover from DOS-era 8.3 filenames. Use the File extension control to write whichever one your downstream software expects; some older scientific and prepress tools match strictly on one or the other, which is the only reason the choice exists.
Why do CCITT Fax 4 and JP2K fail on my video frame?
Because neither can represent what you are giving it. CCITT Group 4 is a bilevel fax compression scheme and requires exactly one bit per sample, so any colour or greyscale frame is rejected outright with a bits-per-sample error. JP2K fails for a different reason: the JPEG 2000 scanline encoding path is not implemented in the TIFF writer here. Both entries appear in the dropdown but neither will complete on a normal video frame — use LZW instead.
Why is a single TIF bigger than several seconds of the MKV?
Because they store fundamentally different things. Video compression only records what changes between frames, spreading a few kilobytes across each second of motion. A TIF stores one complete raster: an uncompressed 1920×1080 RGB frame is 6,220,800 bytes by arithmetic alone, and even lossless compression only shaves that down so far. This is expected and is why lossless stills are used sparingly rather than by the thousand.
Which compression should I choose if I just want it to open everywhere?
LZW. It has been in the TIFF specification since Revision 6.0 in 1992 and is read by essentially every image editor, RIP, and viewer in existence. Deflate and ZSTD produced smaller lossless files in our runs but assume a more modern reader — ZSTD in particular is recent enough that older prepress and scientific tools will not know what to do with it. PackBits is universally supported but compresses photographic frames so little that you may as well pick None.
How do I capture a frame from a specific moment rather than the start?
Set Frame Selection to Specific Frame and type the timestamp into Time (seconds). The field defaults to 0, which grabs the very first frame — often a black or logo frame on anything produced professionally. If you are not sure of the exact moment, use Multiple Screenshots with a fine Capture Rate to sample the clip, then keep the one you want and discard the rest.
Will HDR or interlaced footage come through correctly?
Not automatically, and this is worth knowing before you extract stills for colour work. There is no tone-mapping stage in the pipeline, so an HDR source gets a plain colour conversion and the resulting TIF can look flat or washed out compared with what you saw on a HDR display. There is likewise no deinterlacing filter, so a frame grabbed from interlaced footage keeps its combing artefacts baked into the still. Grade or deinterlace before uploading if either applies.
What is the difference between extracting to TIF, PNG, and JPG?
TIF with a lossless scheme and PNG are both pixel-exact copies of the decoded frame; TIF is the one that print, prepress, and archival workflows expect, while PNG is far smaller and web-friendly — convert MKV to PNG if that is the destination. JPG re-compresses the frame and softens fine detail, but is the right pick for a quick thumbnail or preview: convert MKV to JPG. If your source is an MP4, MOV, AVI, or WebM rather than Matroska, the general video to TIFF converter covers those.
How are my files handled, and how long are they kept?
Your MKV is uploaded over an encrypted connection, decoded on our servers, and both the upload and the extracted stills are deleted automatically after a few hours. There is no sign-up, no watermark, and files are never shared or made public. Matroska itself is an open container — formally specified in RFC 9559 in October 2024 — and the H.264, HEVC, VP9, and AV1 streams it usually holds all decode here before the frame is written out.