Initializing... drag & drop files here
Supports: MRW
MRW is Minolta's raw format — the undeveloped sensor readout from DiMAGE bridge cameras and the Dynax/Maxxum digital SLRs. Konica Minolta announced on 19 January 2006 that it was leaving the camera business and began transferring its SLR operations to Sony on 31 March 2006, so no camera has written a new MRW file since, and support for reading them has been quietly draining out of consumer photo software ever since. This guide is for people holding an orphaned MRW archive who want a flat, viewable, genuinely lossless TIFF preservation copy, and most of its length goes to a single dropdown — because the setting this page opens with is the wrong one for archival work.
.mrw files onto the page or click "Add Files" to browse. You can queue a batch of frames from a DiMAGE or Dynax/Maxxum body in one go..tiff and .tif.Everything about whether this conversion is archival or not lives in one control. TIFF is a container that can hold pixels compressed several different ways, and our TIFF writer passes your choice straight through to the encoder — including the fallback, which is JPEG. That is a reasonable default for someone making a quick preview and a bad one for someone rescuing a raw archive, so it is worth two seconds of attention.
| Compression Type | Lossless? | What it actually is | Use it when |
|---|---|---|---|
| LZW | Yes | The classic lossless TIFF scheme; the most widely readable of the lossless options | Default choice for a preservation master |
| DEFLATE | Yes | zlib/ZIP compression, the same family PNG uses; usually tighter than LZW on photographic detail | You want the smallest lossless file and your tools are modern |
| PACKBITS | Yes | Simple run-length encoding (Macintosh RLE); compresses photographs only modestly | Very old software has to open it |
| ZSTD | Yes | Zstandard; fast and efficient, but a recent addition that older TIFF readers do not know | Your whole toolchain is current |
| NONE | Yes | No compression at all — raw pixel bytes | You want the most defensively simple file on disk |
| JPEG | No | Lossy DCT compression inside the TIFF wrapper — the setting this page opens with | A small preview, never a master |
| LOSSY | No | A literal alias for JPEG — the same encoder value, so it produces the same lossy result | Never; pick JPEG or a lossless option deliberately |
| WebP | No (as used here) | WebP-compressed tiles inside the TIFF; the quality control applies | A small file where TIFF is a hard requirement |
| CCITT Fax 4 | Yes, but | Bilevel Group 4 fax coding — it only handles 1-bit black-and-white | Never on a photograph; it fails on colour |
| JP2K | Depends | JPEG 2000 codestreams inside TIFF; also fails on ordinary colour input here | Never on a raw photo |
Two interactions are worth knowing. First, the "Quality Preset" dropdown ("Very High (Recommended)" and six other steps) becomes a JPEG quality factor — our writer converts "Very High" to Q95 — and a quality factor only means something when the compression is lossy. Pick LZW and the preset has nothing to act on. Second, "Specific file size" works by searching for a quality value that lands on your target; with a lossless compression selected there is no quality to search, so the only lever left is the Auto Scale checkbox shrinking the pixel dimensions. If you are producing an archival master, leave both alone.
A TIFF is not a repackaged MRW, and this is the part people underestimate. The MRW holds one brightness value per photosite sitting behind a colour filter array — no full-colour pixels exist in the file yet. Producing a TIFF means demosaicing that mosaic into three channels per pixel, applying a white balance, and running a tone curve to get from linear sensor readings to something that looks right on a screen. Those decisions get baked into the output permanently.
That is why a lossless TIFF is still not a raw file. The compression is lossless in the sense that the TIFF encoder discards nothing you hand it; the interpretation that produced those pixels is what you can no longer freely revisit. It is also why the TIFF is dramatically larger than the MRW despite starting from the same photograph: one value per photosite becomes three full channels per pixel. Note that our server-side decoder makes its own development choices and will not reproduce Minolta's in-camera processing exactly. It is not doing any HDR tone-mapping — just a standard raw development.
| Property | MRW | TIFF |
|---|---|---|
| Type | Camera raw — undeveloped sensor data | Developed raster image |
| Owner / standard | Minolta, then Konica Minolta (proprietary, undocumented publicly) | Tagged Image File Format, an open published spec |
| Colour data | One value per photosite behind a colour filter array | Full RGB channels per pixel |
| Still being produced? | No — no camera has written one since 2006 | Yes, actively used for print and archival |
| Editing latitude | Full: white balance, exposure and highlight recovery are all still open | None: the render is baked in |
| Compression | Camera-side raw encoding | Your choice — LZW/Deflate/PackBits/ZSTD/None lossless, JPEG/WebP lossy |
| Browser display | None | Safari only; other browsers need an add-on, per MDN |
| Best for | The irreplaceable original you keep forever | The viewable, editable, print-ready copy you actually use |
This path expects a genuine Minolta .mrw. A truncated file from a failed card copy, or some other raw format renamed to .mrw, will either fail to decode or come out flat — re-copy the original from the camera or card if you still can. This page also exposes no bit-depth control and no DPI control, so if your workflow specifically requires a 16-bit TIFF or an embedded ICC profile and print resolution, do that development in a desktop raw editor that reads Minolta files and export the TIFF from there. Whatever you do, keep the .mrw. It is the only copy that still holds the recoverable highlight, shadow and white-balance latitude, and nothing will ever produce another one.
Only if you change one setting. The "Compression Type" control on this page starts on JPEG, and JPEG inside a TIFF wrapper is still lossy DCT compression — the container being TIFF does not make it archival. Switch it to LZW, Deflate, PackBits, ZSTD or None and the TIFF encoder writes back exactly the pixels it was handed. That is the encode side. The development side — turning the sensor mosaic into RGB — is a one-way interpretation regardless of which compression you pick, which is the reason you keep the MRW as well.
It is the fallback in our TIFF writer, and it optimises for file size rather than fidelity, which suits the common case of someone making a shareable copy. It is genuinely the wrong default for a raw archival user, so we would rather say so plainly than let you discover it later. There is also a second entry in that list labelled LOSSY that maps to the identical JPEG encoder value, so choosing it changes nothing — treat LZW as your starting point on this page and ignore both.
LZW if you care about compatibility, Deflate if you care about size. LZW is the lossless TIFF compression that essentially every imaging tool and print RIP has read for decades. Deflate is the zlib/ZIP algorithm and usually squeezes photographic detail a bit harder, at the cost of being less universally understood by very old software. PackBits is a simple run-length scheme that ancient applications can open but that barely compresses photographs; ZSTD is fast and efficient but recent enough that older readers will not recognise it; None writes uncompressed pixel bytes and is the most defensively simple option of all.
No, and that is expected rather than a fault. That dropdown is a quality factor — our writer turns "Very High (Recommended)" into Q95 — and a quality factor is only meaningful to a lossy encoder. With LZW, Deflate, PackBits, ZSTD or None selected there is nothing for it to trade away, so the output is the same whichever preset is showing. It does matter if you deliberately choose JPEG or WebP compression.
Not on this path. This page renders a standard TIFF and exposes no bit-depth selector and no DPI field — the visible controls are Compression Type, Quality Preset, Specific file size, Image resolution and the .tiff/.tif extension. If your requirement is specifically 16 bits per channel or a particular print resolution tagged into the header, develop the MRW in a desktop raw application and export from there. For a viewable, lossless, correctly-sized copy of an orphaned archive, what this page produces is sufficient.
Treat it as unreliable and do not build a workflow on it. TIFF can carry EXIF, but raw metadata written by a discontinued camera line passes through a decoder that was reverse-engineered rather than documented, and maker-note blocks in particular are the first thing to get dropped. If capture dates and camera settings matter to your archive, read them off the original MRW with an EXIF tool and record them alongside — and keep the MRW, which is the authoritative copy of that metadata.
None in the bytes. .tif is a three-letter holdover from the old eight-character filename limit and .tiff is the same Tagged Image File Format; the "File extension" group simply names the output to match whatever your software expects. Pick whichever your cataloguing or print workflow wants and the file itself is identical either way.
Your MRW is uploaded to our servers over an encrypted connection, decoded and written out as a TIFF there, and deleted automatically after a few hours. There is no sign-up, no watermark, and your photographs are never shared or made public. The practical constraint on a large batch is upload size and time rather than anything on the conversion side, so if you are digitising a whole archive it is usually faster to work in modest batches. And keep the originals: the TIFF is a copy, not a replacement.