Re: Archival image reproduction on Kodak Creo IQ3 scanner
Re: Archival image reproduction on Kodak Creo IQ3 scanner
- Subject: Re: Archival image reproduction on Kodak Creo IQ3 scanner
- From: Stanley Smith <email@hidden>
- Date: Mon, 25 Jul 2011 15:04:03 -0700
Chiming in late here, as I only occasionally see this listserv, but this is indeed an interesting dialog. The goal to achieve "accurate" color when photographing artwork is an elusive one. It's a goal that has spawned a lot of research, study, and dialog in the museum community. It is clearly difficult to do, and from my experience using an SG target to calibrate capture will do a great job if your goal is to reproduce an SG target, but not always so great with the paints. ColorSage seems like a great idea, but most conservators (including those at the Getty) are reluctant to allow the physical touching of a painting with an spectrophotometer. An accurate spectrophotometer that can measure at a distance is very expensive.
Fransiska Frey and Susan Farnand's recent Mellon-funded research project Current Practices in Fine Art Reproduction, is a good summary of these issues, and provides some really interesting quantitative analysis of how museums deal with this. The final report "Benchmarking Image Interchange Cycles" is available here: http://artimaging.rit.edu/
In my opinion any RGB sensor based capture system is never going to cut it if your goal is to sneak under the average one deltaE threshold. It seems to be unreasonable to expect to characterize millions of colors based on their response to three. Multi-spectral systems show a lot of promise here, and I worked with Dr. Roy Berns earlier this year at the Getty to test a practical system, with some really good results that will be available in the future. Ken Boydston's MegaVision multi-spectral system is another approach that yields some very accurate captures-- with the bonus of also capturing great UV and IR. The problem seems to be that those of us who really care about this are a very small segment of the market, and unless a tweak of an existing RGB system can be implemented (this is what Dr. Berns is doing-- adding two color filters to create a 6-color multi-spectral capture), we may be stuck with slightly deficient color.
Stanley Smith
Head of Collection Information and Access
J. Paul Getty Museum
1200 Getty Center Drive, Suite 1000
Los Angeles, CA 90049-1687
(310) 440-7286
>>> On 7/24/2011 at 11:54 PM, in message <email@hidden>, Ernst Dinkla <email@hidden> wrote:
On 07/23/2011 05:05 PM, Robin Myers wrote
> When I made the measurements for ColorSage (Better Light's program using HP Artist), I selected as wide a range of colors as possible, with as wide a range of lightness as possible. As noted, having a wider sampling improves most color management profiling results and this one is no different. But it is not necessary to measure every possible blend of two pigments to get good sampling. For instance, in one painting it appeared that the artist had mixed a yellow pigment with a black pigment in a gradient between the two, so I sampled the yellow, the black and one blended color area between the two. It was not necessary to get every intervening mixture between the two blended colors.
>
> In most artworks it actually becomes difficult to meet the 50 sample minimum requirement without repeating. With one painting of a cathedral at night, consisting of blends between only two basic colors, black and a yellow-orange, to get the 50 minimum measurements required sampling just about every possible mix between the two primary colors, with many measurements being repetitions.
>
> Standard issues of measurement also apply. If the artwork has tiny areas and your spectrophotometer has a large measuring area, then the measurement will include the desired patch plus the surrounding colors, thus diluting the spectral result. Also, when using an i1Pro, or similar instrument, they must touch the artwork surface, so using these instruments is only possible on artworks such as dry oil or acrylic paintings where they will not be damaged by the contact. So if you want to measure pastels, pencil drawings, or other artworks with easily damaged surfaces, then you have to use a non-contact measuring device such as a Photo Research instrument.
>
> If you want more information please feel free to contact me.
>
> Robin Myers
>
> Robin Myers Imaging
> email@hidden
Robin,
Thank you, I have a better understanding of the system now. Next
question is of course whether a Canon sensor differs much on its RGB
filters that it could substitute a Nikon for the Ergosoft version of
this technology. Top Sony models are there identical to the Nikons
mentioned.
I see some analogy in the correction matrix (ccmxmake) for Argyll's
Dispcal screen calibration utility where spectrometer measurements are
used to correct/adapt a colorimeter for different monitor panel
colorants. The colorimeter still used for the calibration task as it
measures lower light levels better.
--
Met vriendelijke groeten, Ernst
Dinkla Gallery Canvas Wrap Actions
| Dinkla Grafische Techniek |
| www.pigment-print.com |
| ( unvollendet ) |
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden