DNG Profile Editor - RGB readings

[Dinarius]

I have two questions regarding the DNG Profile editor.

Having created a colour table of a Gretag CC (under the Chart tab), I now want to tweak the image further.

1. In the Color Tables tab, what color space are the RGB readings in? e.g. The blue patch on the chart reads 15, 12, 49. If I was making a Fraser-type profile in ACR, I would be aiming for the blue patch to be 53, 64, 143. Where can I find equivalent values for editing using the DNG profile editor?

2. When making a profile in ACR using the Fraser method, it is first necessary to adjust the grey patches so that the overall luminosity and contrast of the Gretag CC fall into a certain range. How is it possible to create a profile in the DNG Editor without first doing this? Or maybe I should be using the Tone Curve tab before the Color Tables tab?

3. This one has me really confused.........In the Color Tables tab, I click on a patch in the image of the Gretag CC. e.g. the Blue patch. It appears highlighted in the adjustment window. The readings are 15, 12, 49. I begin to make adjustments to the patch by either dragging the circle on the colour wheel or by moving the sliders. The blue patch in the image of the CC visibly alters. Great. But, the RGB readout doesn't change! Why not?  

Thanks.

D.

Ps. I should add that I could make edits using Lab values, if necessary, rather than RGB values, but those values don't change either when I make adjustments! Very, very odd.

[Tim Lookingbill]

If you're on a Mac, you can use the Digital Colormeter set to LAB to get the gamma corrected preview readouts of the CC chart demonstrated below.

The readouts in DNG Profile Editor is in linear space. Can't do anything about it.

[Tim Lookingbill]

Just some pointers I can recall from the last time I attempted to do what you're doing...

Let the Profile Editor wizard do all of the work and don't try to edit each patch with the color sampler editor or you'll introduce noisy artifacts applying this custom edited profile on other images who have colors close to the edited patches in question.

You can be 3-5 points off in any of the Lab numbers compared to the X-rite provided Lab data for each patch before you'll notice a visual difference and even then it will appear so subtle that HSL tweaks in ACR/LR will get you the rest of the way.

Profiling your camera this way is more about getting all the color blocks in a row before you decide to edit the image for "Pleasing" results. IOW profiling isn't about getting you "Pleasing" colors automatically. It's about getting the camera to render what's actually in the scene according to Lab readouts which may not give you the results you expect ("Pleasing") for every image you shoot.

Again ACR/LR's HSL, Color Temp, Split Tone panels are better and faster for correcting what the profile doesn't.

Believe me, you're in for a heap of work tweaking the profile using the DNG Profile Editor and creating a profile for every lens and lighting arrangement. The differences between all those profiles to every possible lighting and lens combo will be so subtle it won't be worth it.

[Dinarius]

Many thanks for the comprehensive reply.

I'm on Vista x64 and my problem outlined in 3. above still applies and renders the DNG Profile Editor useless for me. Namely, that the colour picker readings don't change when I move the sliders (or drag the circle in the colour wheel).

So, if I don't know what the new reading is, I can't generate a profile.

Would love to know if anyone else with Vista x64 has the same problem.

D.

Edited: It now appears that the DNG Profile Editor's inability to not give updated RGB readings (i.e. the reading you get when you open an image file is the only one) is uniform, regardless of the OS you're using.

In other words, you have to make the adjustments visually - unlike in ACR where you can check the changes with the eyedropper tool.

This renders the DNG Profile Editor useless from my point of view.

I'm surprised that Jeff Schewe devoted time to it in "Camera RAW with Adobe Photoshop CS4" to the exclusion of the old (and far more accurate IMHO) Fraser method. And is the older method still available as a download on www.realworldcameraraw.com ?