Re: digital camera profiling
Re: digital camera profiling
- Subject: Re: digital camera profiling
- From: "Fleisher, Ken" <email@hidden>
- Date: Wed, 20 May 2009 09:22:37 -0400
- Thread-topic: digital camera profiling
For the work you are doing, I'm afraid that you may not be able to avoid
local and selective corrections when using the equipment you have listed if
you want better results. The problem is that the camera's spectral
sensitivity functions are different from the standard observer (the way
people see color). While some pigments will reproduce correctly, others will
not due to the spectral nature of the particular pigments (cobalt blue is a
typical offender). What you are asking for is the holy grail for those of us
doing art reproduction, but no three-channel cameras will ever be perfect
for this use.
Spectral imaging, or at least multi-channel imaging with 5 or more channels
(wide-band), will give you an advantage. There are many examples of these
systems at universities and research departments, but there are no systems
that I know of that are commercially available or ready for normal
production.
Polarizing could also be complicating the profiling process which might
reduce the accuracy of the resulting profile. I understand why you need to
do that (we also have to do it often), but this can slightly change the XYZ
values of your chart that you are capturing. Perhaps try measuring the chart
using a polarizing filter on your spectro and then using that measurement
file for your reference data. It's at least worth a try. I have not
personally felt this really improved the results though, but some suggest it
does.
Another thing you can try to improve your results is to either replace or
supplement the ColorChecker SG with a custom target made from the same
pigments that the art is made from (or at least a sampling of pigments that
represent the spectral space of the artwork). Knowing which pigments to use
is not always straight forward, and can be time consuming and expensive to
prodcuce. By doing this, you will calibrate the camera to the correct
spectral space and will likely improve your results. You can find some
articles about this at <http://art-si.org/>. For example
<http://art-si.org/PDFs/Processing/TechnicalReportPigments_04.pdf> will give
you one set of 11 pigments that may work for you.
Most likely, however, if you want more accurate results, you will have to
resort to some local corrections.
Ken
On 5/20/09 8:42 AM, "Derek Lambert" <email@hidden> wrote:
> I have few cameras that we have trouble getting accurate color. We used them
> for reproduction of very old printed pieces mainly but also old paintings.
> Some of them are under glass and cannot be removed because of their price
> and insurance requirements. We use polarized filters on lenses and polarized
> gels on strobes not to get reflections. This is studio setup with controlled
> strobes.
>
> Any suggestion what can I do to make custom color profiles. I tried
> ProfileMaker, Monaco Profiler and BasicColor Input with ColorChecker SG and
> none of the custom profiles I made are close enough. Should I even bother
> with profiling?
>
> I need the best possible color reproduction when there is no need for
> mousing around in Photoshop with making masks and selective color
> corrections. I don't mind one glogal correction but don't have time for
> selections and local corrections. All of this art is going to be printed in
> a catalog on a sheetfed offset press and sold on auction.
>
> My cameras: few Hasselblad H2 with Leaf Aptus 22 Backs using Leaf Capture
> software and one Sinar P2 using Sinar software.
>
> Derek Lambert
--
Ken Fleisher
Photographer
Imaging & Visual Services
National Gallery of Art
Washington, D.C.
Phone: (202) 712-7471
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