Re: GaMapICC v0.5
Re: GaMapICC v0.5
- Subject: Re: GaMapICC v0.5
- From: Graeme Gill <email@hidden>
- Date: Thu, 28 Jan 2010 16:12:20 +1100
Marco Ugolini wrote:
But what is "real life" here? the source RGB image, once adjusted for the
desired tonalities and look, becomes my "reality", the ideal I aspire to in
the final conversion in the face of the necessary compromises.
"Real life" is the appearance of the image on the display and the
appearance of the image when printed and viewed.
Besides, when I have two images each converted to the destination output
space, of which one looks darker than the other (the darker one matching the
source image more closely in tonalities on the same calibrated and profiled
display), what kind of strange miracle is going to make the LIGHTER one a
better match to the source image?
The "strange miracle" is called the human visual system. The appearance
of things is affected by the viewing conditions, and two colors that,
even though they may have the same XYZ values, may not appear to
be identical if the viewing conditions differ.
And a display and a print generally have different viewing conditions.
So doing a colorimetric conversion from CMYK back into a display
base soft proof almost certainly doesn't reflect the actual appearance
of the print made from the CMYK file.
I'm aware of how the substrate affects the printed results, but I'm also
aware of how adaptation compensates for the paper's white point.
This is more than just white point, viewing conditions affects the perceived
tonal range amongst other things.
That is definitely not so. The tonalities match very closely. Even keeping
in mind the adaptations necessitated by the substrate, the tonalities are
still a better match in the Photoshop conversion.
Well, as I said, I don't see that using the particular soft proof
workflow I've used, although I have no doubt that you see it in yours.
Again, let's not forget chromatic/white point adaptation. Whatever may seem
dull before adaptation will look better after adaptation.
I don't see how that can be. The relation of the colors to the white
point remains relatively constant - that's the point of allowing for
white point adaptation. Whether an image looks duller or lighter really
comes down to maximum black density and viewing condition effects -
absolute light levels + adaptation/contrast with the surrounding
illumination.
Well, it's what it is. That's how the ball rolls in my practical world of
real-world production. You work with what you're given. And your judgment of
what looks "natural" is bound to differ from mine.
Sure, although my observation was also based on things like the
histogram, and other logic, such as the fact that something
that is obviously meant to represent blood should probably
appear to be red, that there is contrast and detail in
the axe handle and head that has disappeared in the mud, and
that there is color and contrast in the woodgrain that has also
been lost, and that the white background is obviously
not natural and has been added as a mask, and hence cannot
be used as a reference. But on the other hand, you have
more to go on, within the context of how the image was
supplied and what its intended use is.
The point of the exercise is which procedure makes the best of the given
source image as it is, without questioning it.
Gamut mapping is part science and part art. So some judgment about
what "looks good" guided by the intent and purpose of the reproduction
is always going to be needed.
Graeme Gill.
_______________________________________________
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