Re: ICC Profile location for PS RAW
Re: ICC Profile location for PS RAW
- Subject: Re: ICC Profile location for PS RAW
- From: Graeme Gill <email@hidden>
- Date: Wed, 12 Nov 2003 15:21:24 +1100
Derrick Brown wrote:
>
On 11/4/03 6:19 PM, "Jeff Schewe" <email@hidden> wrote:
>
> Just to be very clear, when you "profile a camera" you are NOT profiling the
>
> camera-you are profiling the camera, the lighting and the exact raw
>
> conversion process involved. You can't profile just the camera separate from
>
> the lighting and the raw processing.
>
>
I would suggest being a little more specific here.
>
Yes Jeff, you are "profiling the Camera, the "raw converter", and the "COLOR
>
TEMPERATURE" of the lighting. Not the "Lighting".
There is more to an illuminant than it color temperature. Illuminants
with identical color temperature but differing spectrums can lead to
quite different appearances when used to illuminate things.
When you profile a camera, you are profiling the effects of the
illuminant spectrum, the test target reflectance spectrums,
the camera sensor color matching functions (CMFs, or sensor spectral
sensitivity spectrums), and any electronic processing that
follows. Note that if you make a change to the illuminant
spectrum (even if the color temperature remains the same), or
illuminate objects that do not have the same set of reflectance
spectrums for each color that the test chart has, that the resulting
color from the camera can change. No colorimetric profile can compensate
perfectly for these sorts of effects. Having a camera with CMFs that
are similar to the ones in our eyes will improve the usefulness of
colorimetric profiles, but is often at odds with making a practical input device.
The ACR idea of profiling at two different color temperatures
(and therefore spectrums), and then interpolating between these
profiles to adjust the white point, is an interesting approach to
accounting for the metameric affects of a change in the illuminant
spectrum with white point. I'd imagine it is less effective if the actual
illuminant spectrum is quite different to the two used to create the ACR
profile, but overall it sounds like it should be more effective than
just changing the white point in a colorimetric space.
Graeme Gill.
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.