In my recent wanderings, I downloaded the Adobe DNG Converter in order to snarf the DCP profiles. In writing my recent post in No graduate saturation (example pic included) - #29 by ggbutcher, I used dcamprof to convert some of the Nikon D7000 DCPs to ICCs. Being curious, I plotted the primaries from a few to see what they did to the gamut, and was surprised by this:
The green-dot gamut is the Standard profile, looks like what a calibrated camera profile should look like. The other (blue-dot) is the Neutral profile from the “look” profiles, e.g., Vivid, Landscape, Portrait, etc. It’s gamut is a fair bit smaller than the Standard, but when I use it as a camera profile the output looks pretty good. In fact, it seems to do a better job at the blue correction than the blue Y thing I did in the other thread.
And so now I’m wondering about my camera calibration. It would seem Adobe has a different idea about it (Standard) than Argyll, and the look profiles (Neutral, Vivid, etc) are more about working profile, but the best result with them seems to be as the original camera profile. I’m considering using the Adobe profiles in my everyday processing, but I need to understand them first. I can post pictures from the different combinations, but not today, my wife is home now and has poured wine…
So the primaries are outside the CIE 1931 color space because there is no real lighting that exists that can excite only the green sensels on your sensor.
Standard really pumps up the yellow-greens and the violets, at least in relation to Neutral. It’s hard to say which is actually more accurate without using a test chart…
If you have the time, could you pick a camera make and model and plot Adobe Neutral vs Adobe Standard vs one of the RT profiles?
Please pick one which was updated in the last 2 months (~1.1MB in size, so not Nikon D7000). I would suggest the SONY ILCE-7M2. Files here for convenience: Filebin | s6cgi3by01fxmsxu
Copy primary lines to clipboard
Paste primary lines into the “exiftool Paste” text box at http://glenn.pulpitrock.net/xyY.html
Click “Parse to XYZ”
Click “Compute xyY”
Click “Plot xy”
Ta-Da.
Rinse and repeat. The xyY page will plot up to 5 gamuts.
For convenience, here are the exiftool lines, with a FileName line that will populate the Profile box in the xyY plot page:
FileName : RawTherapee
Red Matrix Column : 0.69107 0.25027 0.0006
Green Matrix Column : 0.13867 0.71164 0.0528
Blue Matrix Column : 0.13446 0.03809 0.7715
FileName : Adobe Standard
Red Matrix Column : 0.59061 0.3866 0.23964
Green Matrix Column : 0.24512 0.5739 0.0005
Blue Matrix Column : 0.12846 0.03949 0.58475
FileName : Adobe Camera Neutral
Red Matrix Column : 0.79776 0.28798 -2e-05
Green Matrix Column : 0.13516 0.71193 0
Blue Matrix Column : 0.03127 9e-05 0.82489
FileName : Adobe Camera Standard
Red Matrix Column : 0.79776 0.28798 -2e-05
Green Matrix Column : 0.13516 0.71193 0
Blue Matrix Column : 0.03127 9e-05 0.82489
Observations:
Adobe Camera Neutral and Camera Standard use the same primaries.
The Adobe Standard profile would indicate they think this is a crappy camera with respect to red response (?)
RawTherapee (obtained from the github repository) and the Adobe Camera profiles are relatively close.
The RawTherapee DCP uses AdobeRGB gamut compression (contrast + saturation compression), to avoid issues with color clipping, usually prevalent in reds and blues.
