Fors' ACR Calibrator Measurement File

[mcmorrison]

Hello,

I have begun looking at using Thomas Fors' ACR Calibrator for my 5D. As I was sorting out the procedure, I found no mention of measuring the Colorchecker being photographed. As these also vary—and perhaps by as much as intended calibration correction?—it seems it should be measured.

I have found a section in the script that gives 8-bit Prophoto values for each patch, and wonder if it makes sense to measure my Colorchecker, convert the measurements to Prophoto, and replace the  values in the script with my own measurements. If so, I presume that an "absolute" conversion is the correct rendering to use. . .

Any thoughts?

Thanks!

Michael Morrison

[eronald]

Indeed.
In fact there are significant variations between colorcheckers.

Edmund

[Dinarius]

I have begun looking at using Thomas Fors' ACR Calibrator for my 5D. As I was sorting out the procedure, I found no mention of measuring the Colorchecker being photographed.
[{POST_SNAPBACK}][/a]

You are correct that the issue of the actual values of the chart being used is relevant.

Undoubtedly, there will be variations. However, for most people, taking measurements is both impossible and impractical.

Much better is to use BabelColor's excellent tables of average measurements, taken from 20 charts.

You can download the results and print them off from here>

[a href=\"http://www.babelcolor.com/main_level/ColorChecker.htm#ColorChecker_data]http://www.babelcolor.com/main_level/Color...lorChecker_data[/url]

Scroll down to the section headed "ColorChecker" data and click on "ColorChecker data and spectral.xls."

I convert everything to Adobe 1998 16bit TIFF and find the values excellent.

Two final points.......

1. I shoot fine art for a living and, in terms of achieving superb colour accuracy ( by this I mean making a proof print and viewing it and the original artwork in the same light.) in RGB is the least of your problems. The real problems arise at the CMYK stage.

2. Your question really belongs in Digital Image Processing and you might have received a lot more replies if you had posted it there. ;-)

D.

[Jack Flesher]

Actually the MacBeth color checkers are far more accurate than you might believe.  Robin Masters of Betterlight checked several, found flaws in some earlier editions and reported them to MacBeth.  The result is the current versions are very consistent and accurate.  Note that they do age and will discolor over time, so if you are requiring one for critical work, it should to be replaced with a new one every few years.

[mcmorrison]

Hello,

Thanks for your thoughts!

Regarding where to post this thread, I thought it would be a camera topic rather than an image processing topic, but I see your point. Alas.

I too photograph paintings for reproduction, although on a small scale. For these I shoot a target under the same lighting as the painting and make a profile with Eye One Match. This works pretty well for me, with most of my color correction work resulting from lighting issues (I use cross-polarized Solux lighting, rather than pony up for some Northlights. . .  :-)) ) I print all my work to an iPF5000, so have been spared CMYK travails. Have you seen Bill Atkinson's discussion on the truly profiled press he used for "Within the Stone"?

My interest in the ACR Calibrator is for improving the color of "regular" photos in Lightroom, not the reproduction work.

Best Regards,

Michael Morrison