Re: scanner profiling (again)
Re: scanner profiling (again)
- Subject: Re: scanner profiling (again)
- From: Armand Rosenberg <email@hidden>
- Date: Tue, 16 Dec 2003 15:12:50 -0500
Jim,
At 1:01 PM -0500 12/16/03, Jim Rich wrote:
So one question to consider is why does each image I photograph have a
different white and black point?
That's THE question.
As I pointed out in my last post factors like exposure, the scenes content,
and film processing cause the wp and bp to vary. It is just that simple.
Let me repeat my starting assumptions, in case they were not clear: I
have a correct exposure and processing is reasonably consistent (with
positive film). That is the case I am interested in. Then wp and bp
should be reasonably consistent between target and image. I can
understand small deviations, but not major changes.
I don't understand why scene content would matter in this case. As I
said, the target contains patches corresponding to highlight and
shadow. So any scene's content should be covered by the target
patches.
It is the nature of light striking objects and then reflecting back on to
film or a sensor.
It is not perfect.
What is "not perfect" about this phenomenon? How is it going to
result in a properly exposed image that has areas with OD higher than
the GS23 patch or areas with OD less than the GS0 patch?
If this can happen and is a frequent occurrence, then the concept of
using that target to profile a scanner is flawed.
So a savvy operator then intervenes
to set the correct wp and bp for that situation.
OK. But that's outside a true ColorSync (ie, ICC) workflow, isn't it?
As for when to adjust the wp and bp. I adjust wp and bp after I have used
one pair of profiles. That is to Assgin and then Convert to my WS.
OK. Makes sense.
I am not sure how else it can be a Photoshop ICC based workflow that uses
non ICC savvy scanner software.
The scanner sw need not be "ICC savvy" itself -- all it has to do is
capture a raw scan with locked scanner settings. The profile can be
assigned later (eg, in PS.) Isn't that all that's required of an ICC
workflow?
It's the same idea as printing with a printer driver that isn't "ICC
savvy" itself: you can still use custom profiles just by letting PS
handle all profile conversions.
If you adjust the raw image before you
Assign the profile you are altering the raw state in which you created a
scanner profile. This makes the scanner profile invalid and is illogical in
an ICC workflow.
I agree.
(clip)
If the scanner and software allows you to set up all of the image
processing parameters before the scan then you are ahead of the curve when
it comes to being paranoid about quantization errors that Photoshop image
processing might or might not introduce.
It's true that you can set up the parameters in the scanner sw based
on preview, before clicking "scan," but really the adjustments are
done by the scanner driver sw AFTER the data is captured... The
capture itself seems unaffected by these parameters.
So I repeat once more: even if you do these settings in the scanner
software, they are still applied only after the data is captured, as
far as I can tell -- then it just does not matter whether you do this
in the scanner software or in PS, as long as you have a true 16-bit
data path.
(clip)
So back to THE question: with a correct exposure and consistent
processing, and using the same film as the target (nominally), why
should each image have drastically different wp and bp? To me, it
seems like the differences, if any, should be small.
Armand
_______________________________________________
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.