Canon CR2 RAW format conversion to typical cine file formats and image sequences
like DPX, Cineon, EXR and QuickTime movies. It features 32bit floating point image path and fast multi-CPU
parallel processing. For a complete list of features read on.
For a quick test-run with crw2dpx just download the demo!
If you have questions, there might be an answer already in the FAQ section.
- works with Canon CR2 RAW files *
- floating point RAW conversion, keeping all highlight and shadow detail
- crop / scale image using high quality algorithms
- deblur to compensate the sensor's OLP filter
- selectable debayer algorithm (bilinear and LMMSE)
- fix blown highlights (i.e. remove magenta cast)
- specify the number of CPU cores to be used for parallel processing
- automatic dead / hot pixel detection and fixing
- apply camera ICC profile
- timecode burn-in
The following image-sequence file formats are supported:
- DPX: 10bit log
- Cineon: 10bit log
- TIFF: 8bit sRGB, 8bit Rec709, 16bit linear with embedded ICC profile
- PNG: 8bit sRGB, 8bit Rec709, 16bit linear with embedded ICC profile
- EXR: 16bit half, 32bit float, uncompressed or ZIPS, ZIP, PXR24
- Targa: 8bit sRGB, 8bit Rec709
- SGI: 8bit sRGB, 8bit Rec709, 16bit linear
- Radiance HDR: uncompressed or with RLE compression
The following movie file formats and codecs are supported:
- QuickTime: 2vuy, v210, v410, Animation (RLE), PNG, Photo-JPEG (4:4:4 and 4:2:2), Aja Kona 10bit (R10k and R10g), Blackmagic r210, ProRes 422 and 4444**
- AVI: YVYU, v210
- Piranha ARC: 8bit Rec709, 10bit log, 16bit half
If supported by the selected output file format,
all files written will have metadata (e.g. timecode, frame number, camera make, model and serial number) embedded.
Once the RAW sensor data has been extracted from the source files all image processing
is performed using a 32bit floating point pipeline. Thus, instead of clipping all values
below the camera's black level and above its white-point, files converted with crw2dpx
retain all image data provided by the camera sensor. ***
"I spent most the day yesterday trying to get a good conversion and everything was junk and or incredibly slow until I found yours!"
Marc D. Rienzo, Director / VFX Supervisor & Artist
"I shall be an enthusiastic promotor of this software to all I work with!"
Joseph Williams, Little Fish Films, London
You can purchase a license from within the application using PayPal. The license costs 49 EUR and will only work on the machine
it was ordered from. Free updates for all upcoming 1.x versions are included.
If you want to purchase a license for a machine that has no internet access, please get in touch using the contact form.
Please specify the operating system you want to run crw2dpx on, so that my reply can be a bit more specific.
A public beta of the upcoming crw2dpx v1.5. It adds GPU processing and also writes 12bit and 16bit DPX files as well as MayaIFF files up to 32bit floating point.
The extended-dynamic-range processing (EDR) now comes with a second option to generate monochrome highlights.
No release date for the final v1.5 yet, but if you have an existing license this version will work with the same license code.
For use on render-farms or similar environments I can provide a command-line version of crw2dpx. In order to negotiate pricing and the required features please get in
touch with me using the contact form. It is possible to support other output formats or add special features
like rendering on the GPU, automated generation of proxy files, 3D support or error logging to a logfile.
The cr2Reader plugin for Nuke uses the same imaging pipeline as crw2dpx. This is an add-on for crw2dpx users and will only work on a machine with a valid crw2dpx license on it.
On other machines output will be watermarked. Please note that this plugin is largely untested because I don't have access to Nuke.
If you need a version for some other operating system than Linux or for a different Nuke version, please let me know.
* for a complete list of supported cameras, check the manual that comes included with the downloads
** OS X 10.7 and later only, makes use of an Apple API and requires the AVFoundation framework
*** requires output file format that does not clip to 0-100% range