Re: Color spaces/ Digital Cameras
Re: Color spaces/ Digital Cameras
- Subject: Re: Color spaces/ Digital Cameras
- From: "tlianza" <email@hidden>
- Date: Wed, 30 Jan 2002 22:19:33 -0500
Hello Richard
I read your question with interest. I would first suggest that you visit
this site:
http://www.brucelindbloom.com/
and down load the Delta E 16 bit gamma 2.2 tif target. Bruce has provided a
wonderful test image that you should first use to check your basic
calibration
assumptions.
You didn't mention how you calibrated your system but as a first pass try
running this
target through the system. Follow the direction on the download page for
use with PS.
The image that you print should match the one on the screen, if it doesn't
you are going to
continue to have problems that have nothing to do with the camera.
Each camera manufacturer builds their own non-linearity into the front end
of the process and
then applies mapping to a space such as sRGB or AdobeRGB. The front end
processing often is (and should be)
very dependent upon absolute exposure so it is basically impossible to
gurantee that a camera profile measured at one
exposure will be accurate at another exposure or color temperature. The
cameras are designed to produce good looking
images, not good colorimetry. If they were colorimetrically accurate, most
users would be quite disappointed.
For instance, in low light levels, the scene contrast appears visually lower
and the scene appears to be less colorful even if
the absolute colorimetry indicates otherwise. If we reproduce this scene
perfectly from a chromaticity and relative lightness
basis, but view it at higher levels of illumination than the orignal scene,
it will appear to have more contrast and greater
colorfulness than the origninal scene. This may be very pleasing, but it
can lead to some problems. The same is true with
a brightly lit scene: it may appear to be very colorful and contrasty, but
if we reproduce those same chromaticities and the
exact relative (to white) lightnesses, on paper and view in normal indoor
lighting, the image looks flat. That's why the high
contrast papers are used in typical mini lab environments. The camera
manufacturers are trying to produce images that
look like transparencies when viewed under an sRGB assumption. This means
that the camera characteristic curve
is "S" shaped and it has a gamma that is higher than 1.0.
The key to the reproduction process is to make sure that image on the screen
represents what is printed. The cameras
are designed to make pretty pictures, not colorimetricly correct images.
Tom Lianza
Technical Director
Sequel Imaging Inc.
25 Nashua Rd.
Londonderry, NH 03053
email@hidden