[argyllcms] Re: Camera profiling?

On 2008 May 27, at 8:43 AM, Graeme Gill wrote:


None of this is to say that a profile won't help, just that it
cannot be perfect, and in general photographs are always going
to require some level of subjective adjustment, unless they are
all shot under identical conditions.
That was the idea -- one profile per lighting setup. For example, there're a half-dozen wooden discs about 26" in diameter. Once I get what I want for one, I'll just swap out the art works and keep everything else identical.
Some of the larger, more abstract works will need individualized lighting...but I'm willing to do one profile per shot if it's worth it. It's not /that/ much of a hassle, compared to everything else involved. 


Note that a scanner is a much more controlled situation in that
the illuminant is fixed, and the characteristics of the test chart
can be adjusted to be similar to the type of media that will be scanned. 


I have both  a ColorChecker and the scanner chart  that comes with
the i1 Pro. Will either (or both) of those suffice?


Sure, but you may have to take particular care about the illumination
evenness, and may also have to manually get rid of any perspective
distortion (at least until the next release of scanin, which should
then take care of this aspect automatically).
I can take extra care in making the test chart perpendicular...and there's always Photoshop if I'm off too far.
Will scanin give a recognizable warning, or should it just be the first thing I assume is worng if it complains? 


Don't be surprised
if they give different results though, since I think that the
ColorChecker (DC or SG ?)


It's the regular 24-patch kind


are composed of individual colorants,
while (I presume) the i1 pro scanner chart is some sort of print
process.
That was my first guess, but I can't see any hints of a halftone screen or dithering, even with an 8x loupe. It's printed on glossy photographic paper. (I'll have to be extra careful to avoid glare, I guess.) Dye sub, perhaps? 

I'll try an experiment in a bit....

Cheers,

b&

Other related posts: