Nikon original colors for Capture One

"Someone is looking for fish, someone is looking for wildfowl, someone is looking for mushrooms. Another is looking for money - and finds the wildfowl, mushrooms and fish." Mikhail Zhvanetsky

Long and hard I had searched a way to get Nikon original colors for excellent professional raw converter Capture One, and finally I found them some months ago. Now I can make icc-profiles for Nikon cameras, released until 2014 (Nikon Df is last of them), and get colors in Capture One exactly the same as Nikon Capture NX-D.

But there is an obvious question: How these colors match the Nikon original colors?

Let’s try to find the answer.

We need to take one image of color target (reference), developed in Capture NX-D, and the second image of color target (sample), developed in Capture One and test icc-profile. And we need to compare color patches.

(My NikonD5100-EO-NXD-Standard.icc and others files, ColorLab Utility Freeware, and Nikon D5100 source nef)

And now step by step. (But you can repeat these comparisons again, starting from D5100hVFAI00200.NEF, not only from my finished tiff files.)

Reference: NikonD5100_CC_NXD-Standard.tif developed in Capture NX-D, picture_control: Standard, noise_reduction: off, all_another: default, tiff: 16-bit AdobeRGB
Sample: NikonD5100_CC_С1-Standard.tif developed in Capture One, base_icc_profile: NikonD5100-EO-NXD-Standard.icc, base_curve: Film Standard, sharpening: No, noise_reduction: disable, all_another: default, tiff: 16-bit AdobeRGB

Then, these files had converted from AdobeRGB to LAB color space in Photoshop:
NikonD5100_CC_NXD-Standard_LAB.tif
NikonD5100_CC_C1-Standard_LAB.tif

Then, in X-Rite ColorLab Utility:
File->Open(NikonD5100_CC_NXD-Standard_LAB.tif) ; Filter->Layout and Format->Spot Colors…->Width:6, Height:4->OK

This window is left opened.

Then, in X-Rite ColorLab Utility:
File->Open(NikonD5100_CC_C1-Standard_LAB.tif) ; Filter->Layout and Format->Spot Colors…->Width:6, Height:4->OK

Then, in X-Rite ColorLab Utility:
Special->Comparing…->Delta E 2000->Save Report(NikonD5100_CC_NXD-Standard_DeltaE2000.txt)

NikonD5100_CC_NXD-Standard_DeltaE2000.txt describe detailed results.
Worst Delta E 2000 = 3.37

All these steps can be repeated for ColorChecker SG in a similar way.
NikonD5100_CCSG_NXD-Standard_DeltaE2000.txt describe detailed results.
Worst Delta E 2000 = 4.35

After all we can read on X-Rite ColorLab Utility user manual page 32: “ColorLab can read 16 bit images, but it converts the data internally into 8 bit when it loads an image.”
In addition, X-Rite ColorLab Utility last release date: 1/6/2006.
And most likely the measurement results are not accurate enough.

So I took advantage of other instruments, and I repeated measurements ColorChecker SG again.
I measured the RGB-values of color patches using Argyll scanin.exe: NikonD5100_CCSG_NXD-Standard.val и NikonD5100_CCSG_C1-Standard.val.
Then I converted from RGB-values to LAB-values for D50 illuminant: NikonD5100_CCSG_NXD-Standard_LAB16.txt и NikonD5100_CCSG_C1-Standard_LAB16.txt.
And finally I compared these files using X-Rite MeasureTool Utility:
NikonD5100_CCSG_NXD-Standard_DeltaE2000-16.txt describe detailed results.
Worst Delta E 2000 = 1.17

Wow!
End of story.

Afterword:
It is very likely that I can make D2XMODE1 (my favourite), D2XMODE2 and D2XMODE3 icc-profiles for Nikon cameras that originally do not have them in the cameras picture control kit: Neutral, Portrait, etc. Actually I already made D2XMODE1 icc-profile for Nikon D5100. Pictures turn out quite well, but for test it is necessary to make the target images in the same conditions using D5100(I do not have it, and have to borrow) and using my D300, that will be a benchmark for comparison.

And, as a gift to fans of CCD-sensors colors, I posted all my icc-profiles for the Nikon D80: EO_ICC_NikonD80.zip.