Archival image reproduction on Kodak Creo IQ3 scanner
Hi I am sure there are many on this list who have had to deal with reproduction of historical documents for state or national archives. I have a client who is using a Kodak Creo iq3 scanner with OxyGen software to scan documents ranging from 19th century glass plates, albumen and silver gelatine photographs, water colour and ink illustrations, hand coloured B&W images right up to modern day emulsions. We have built two profiles in ProfileMaker. One based on a reflective photographic target for standard emulsions and one based on a Colourchecker SG target for other reflective media. With a small amount of tweaking in Profile Editor we can get the white, grey and black patches on a X-Rite greyscale checker to match the target RGB values in Adobe RGB within one level. My question is, is there a subject based spectral response to the scanner light source that will render each of these types of original material differently with respect to the colour checker patches. In other words, are we likely to have to build a material specific correction for different types of originals? We have found, purely visually on a small range of samples, that there seems to be one type of cast for one material and a different cast for something else. Without going into specifics, is this an accepted issue and would a workable solution be scan all reflective originals with the same scanner profile and run a different photoshop action as a correction layer for each different type of material. This would seem simpler than having 10 different edited versions of the scanner profile, and would be easy to adjust over time as more samples are taken into account. Thanks in advance. Simon Olding ICC Imagetec mail@iccimagetec.com.au www.iccimagetec.com.au photo galleries: www.simonolding.com P: (03) 6223 7882 M: 0424 022 831
On Jul 21, 2011, at 2:28 AM, mail@iccimagetec.com.au wrote:
My question is, is there a subject based spectral response to the scanner light source that will render each of these types of original material differently with respect to the colour checker patches.
Yes.
In other words, are we likely to have to build a material specific correction for different types of originals?
Depends....
We have found, purely visually on a small range of samples, that there seems to be one type of cast for one material and a different cast for something else.
That's a good start.
Without going into specifics, is this an accepted issue and would a workable solution be scan all reflective originals with the same scanner profile and run a different photoshop action as a correction layer for each different type of material. This would seem simpler than having 10 different edited versions of the scanner profile, and would be easy to adjust over time as more samples are taken into account.
Image capture devices will inherently provide different spectral response information dependent on the characteristics of the capture device itself, the light source used to illuminate the target and the target's physical composition itself. Standardize the first two, the third remains to vex critical reproduction. Your approach is similar to what others have settled on. When it's possible to create a target - say watercolor or pastel artwork - that can be used in place of a color checker-type device, profiling for the system then becomes a more worthwhile endeavor but only as long as the custom target and subsequent artwork are consistent with one another. Working with legacy materials as you've mentioned renders this solution irrelevant, so a visually-calibrated approach is typically the path of least resistance towards reaching an acceptable reproduction. (My background isn't as extensive as others on this list I admit. I've done a fair amount of reproduction of artwork created by my wife and some of her friends as well as others in a variety of media. Invariably I end up getting as close to a visual match as I feel I can using the method you've outlined, whether capture is scanner- or camera CCD-based, or the medium the artwork was created in. Of course you must also consider the overall gamut and consistency of the output device your reproductions will be rendered upon, but that's kind of a no-brainer when it comes to stuff of this nature.)
participants (2)
-
mail＠iccimagetec.com.au
-
Stephen Clark