Rumor of a new Hasselblad 100 megapixels back

[eronald]

What causes you to think that the matrices you quote were made to ISO 17321 specs. Whoever made the profiles could derive their matrices.

Bill

I would really like to read the spec ...

my email is edmundronald at gmail

[ErikKaffehr]

Hi,

OK, so you used data from DxO-mark.

Now, let's look at DxO-mark data comparing two same generation Sony Cameras. The Sony A900 and the Sony A850. So, you would expect them to have identical sensors, but the DxO-measurements vary.

If we look at color conversion matrix in the EXIF file for each camera, they are identical. So, Sony obviously regards the sensors the same. So, I would guess that this is just sample variation.

The DxO-mark ratings are almost identical between the A-900 and the A850, so it doesn't seems that the A850 would have a more permissive CFA either. 

Best regards
Erik

The text in front of them in the links I gave is "Color matrix as defined in ISO standard 17321".

[landscapephoto]

I would really like to read the spec ...

my email is edmundronald at gmail

Each downloaded copy of the ISO standards is marked with the reference of the downloader. If you had a copy, would you infringe the ISO copyright and send your traceable copy to an anonymous gmail mailbox?

[landscapephoto]

If we look at color conversion matrix in the EXIF file for each camera, they are identical. So, Sony obviously regards the sensors the same.

You obviously have no idea of the difference between RGB filter values and camera native color space. Exif tags 0xc621 to 0xc626 do not correspond to RGB filter values.


Further, let me cite the two matrixes again for reference:

P 45+
R raw   3.19   -1.76   -0.43
G raw   -0.15   1.5   -0.35
B raw   0.18   -0.98   1.8

H3DII-39
R raw   3.19   -1.83   -0.37
G raw   -0.21   1.64   -0.43
B raw   0.18   -0.98   1.8

Some of the coefficients are more than 20% apart, particularly the green line. Dismissing these differences as "sample variations" is outrageous, particularly for a person who will normally find significance in differences which are less than a feather apart.

After a good night sleep, I reread the complete discussion. I am not interested in pursuing the discussion any further, enough has been said.