MIRawViewer is a converter?
Yes. You can convert to CDNG or ProRes.
MLVFS is for MLV files only and also it appears to involve downloading OSXFuse which hasn't got great reviews.
Huh? Hasn't got great reviews? Please explain. It's just an open source framework. People don't usually write "reviews" of code frameworks, so I'm very curious as to where you found any reviews, much less "not great" ones.
As far as I'm concerned, FUSE is actually quite amazing and extremely powerful. Using it was my idea, and I've been developing on this framework without a single hitch. Also haven't seen any stability issues. So,
what the heck are you talking about?
mlv_dump (the one for OSX) won't open as it's a Microsoft thing.
It works fine here and that has nothing to do with Microsoft.
According to Adobe's DNG validator, the image is actually 1920x1080:
./dng_validate -v M14-1724_C0000_00142.dng
Validating "/Users/david/Downloads/M14-1724_C0000_00142.dng"...
Uses little-endian byte order
Magic number = 42
IFD 0: Offset = 8, Entries = 34
NewSubFileType: Main Image
ImageWidth: 1920
ImageLength: 1080
BitsPerSample: 16
Compression: Uncompressed
PhotometricInterpretation: CFA
FillOrder: 1
StripOffsets: Offset = 16384
Orientation: 1 - 0th row is top, 0th column is left
SamplesPerPixel: 1
RowsPerStrip: 1080
StripByteCounts: Count = 4147200
PlanarConfiguration: 1
*** Warning: IFD 0 Software is not NULL terminated ***
Software: "RAWMagic 1.2.2"
CFARepeatPatternDim: Rows = 2, Cols = 2
CFAPattern:
Red Green
Green Blue
DNGVersion: 1.1.0.0
*** Warning: IFD 0 UniqueCameraModel is not NULL terminated ***
UniqueCameraModel: "Canon EOS 5D Mark III"
BlackLevelRepeatDim: Rows = 1, Cols = 1
BlackLevel: 1024.00
WhiteLevel: 15000
DefaultCropOrigin: H = 0.00 V = 0.00
DefaultCropSize: H = 1920.00 V = 1080.00
ColorMatrix1:
0.6722 -0.0635 -0.0963
-0.4287 1.2460 0.2028
-0.0908 0.2162 0.5668
ColorMatrix2:
0.6722 -0.0635 -0.0963
-0.4287 1.2460 0.2028
-0.0908 0.2162 0.5668
CameraCalibration1:
1.0000 0.0000 0.0000
0.0000 1.0000 0.0000
0.0000 0.0000 1.0000
CameraCalibration2:
1.0000 0.0000 0.0000
0.0000 1.0000 0.0000
0.0000 0.0000 1.0000
AsShotNeutral: 0.4761 1.0000 0.7000
BaselineExposure: +1.00
CalibrationIlluminant1: Standard light A
CalibrationIlluminant2: D65
*Tag51043: Byte, size = 8
00 40 24 17 00 00 00 00 .@$.....
*Tag51044: SRational = 24000/1001
*** Warning: IFD 0 Tag51081 is not NULL terminated ***
*Tag51081: "M14-1724_C0000"
NextIFD = 0
*** Warning: Too little padding on left edge of CFA image (possible interpolation artifacts) ***
*** Warning: Too little padding on top edge of CFA image (possible interpolation artifacts) ***
*** Warning: Too little padding on right edge of CFA image (possible interpolation artifacts) ***
*** Warning: Too little padding on bottom edge of CFA image (possible interpolation artifacts) ***
Raw image read time: 0.005 sec
Linearization time: 0.006 sec
Interpolate time: 0.050 sec
Validation complete
As a1ex stated, the black level is incorrect (1024) and the image shows up incorrect in ACR. However it looks fine in the Finder.
Also, in the Finder it shows up as being 1916x1076. This appears to be (somewhat) of a bug on Apple's part. The DNG spec recommends providing padding around the true "Active Area" of the DNG to avoid interpolation artifacts from demosaicing (see the warnings above). It appears that the Finder is ignoring the active area and forcing a padding to avoid interpolation artifacts. Of course, we want every pixel we can get so all raw video DNGs out of ML aren't going to sacrifice pixels for padding. Most DNG readers can handle there being no padding.