Printed on the back of the packaging for the GT Color Checker, is a chart that describes the 24 squares in terms of their color stated in terms of CIE (1931), as values of x. y and Y. The bottom row of squares numbered 19 through 24 are shades of gray; named "white" through "black".

I am unfamiliar with CIE (1931), but the values for "Y" look quite to be very close to what "L" would be in Lab scale which is a 0-100 scale of luminance. Can anyone shed any light on this connection please?

I would like to use the "white" square which has a "Y" value of 90 and estimate what that would be in RGB. Ninety percent of 256 would be 230 mathematically. Can I correctly infer that Y = .9 = RGB = 230?

Jerry Reed

Nope. You're forgetting about the gamma curve built in to editing spaces. The RGB values will vary depending on which space you convert to.

LAB and XYZ are well-defined color spaces; however rgb is just a generic notation, you need to attach a colorspace eg. sRGB or AdobeRGB for rgb values to acquire meaning. If you are a mathochist go to www.brucelindbloom.com and look at the various color calculators, else get a book.

Edmund

When lighting an art piece, with a GT MacBeth Color checker substituted for the art object to build the lighting, using the Preview Mode in Capture Shop with the eyedropper placed on the white square, for images to be converted to ADOBE 1998 from RAW after a 16-shot , what value read by the eyedropper do you use to ensure adequate distributions of the tones?

Jerry Reed

I adjust so that the lightest neutral square (white) is about 250 with exposure set to 0 in ACR, converting to ProPhoto. ProPhoto is your best bet for artwork because it can contain more saturated colors than Adobe RGB or sRGB.

Bruce Lindbloom's site has an extensive set of tables with the mathematically correct RGB values for the colochecker squares in all the popular color spaces ... check it out

http://brucelindbloom.com/index.html?ColorCheckerRGB.html

I aim for 200,200,200 in adobe RGB 1998 on the second square from the left. (200,200,199 is the exact number, but once I've used that square for my white balance ...)

Jeremy:

Thank you. This is just what i was looking for, and answers a question that I had but had not asked, which is, "I am I the only one who sees the left most "White" square as having a lower value for the "BLUE" channel?" In every color space that square is shown with a lower blue channel value, so it must have printed that way and it is not a problem that has crept into my color management.

Thanks again,

Jerry Reed