Digital BW, The Print

Tried some combinations today:

QTR's latest set of profiles,16-2-2005

Vuescan, PWP, PS-CS, Qimage.

ACE, ICM, CMM, LCMS color engines.

With and without Black Point Compensation.

Zuber RGB Black target converted to 48 bit and the QTR RGB Lab
assigned to it -plus
of that file a conversion to 16 bit Greyscale target + QTR
Grey Lab
assigned to it.

Checked the RGB numbers of the P2P conversions or the "files
printed"in Photoshop.

Vuescan doesn't allow a choice of a color space to assign to
the scanned image other than from an internal list of color
space profiles. The last is a nice list for color images so
there's no reason to complain: AdobeRGB, Colormatch, Bruce,
Wide Gamut, Ektaspace etc. But adding another one from Windows
resources isn't possible. We have to ask Ed Hamrick for the
possibility to add either QTR RGB-Lab or QTR Gray-Lab to that
list. The printer profiles show but I do not see a need for
adding printer profiles to scanner software. Not having the
color space makes it impossible to check the workings of the
LCMS color engine with the profiles. The choices of B&W film
calibrations are quite good ( with developer times etc in the
manual for some films) so the film input side is reasonably
well covered, it is possible to make custom film profiles but
that's usually only done with color slide film. It doesn't
mean that there couldn't be a new B&W custom calibration added
but that's Ed territory.

Picture Window Pro.
To put it simple, avoid Picture Window Pro when you intend to
use QTR's profiles, color space and printer profiles
altogether, choice of LCMS or ICM color engine doesn't
matter. It is totally unpredictable, softproof shows
reversing of the image, errors, program goes flat in an
instance etc.  A neutral RGB file with the QTR RGB Lab space
profile assigned can be used but avoid profile conversions
etc. I think the program needs more than bugs taken out on
color management.

Qimage.
I don't think that the new QTR profiles are working nice
either in Qimage with its LCMS color engine. Best is the RGB
input with the QTR RGB Lab as its space and QTR Photo Paper as
the printer profile, even then there are differences to the PS
CS ACE conversions. Ferdinand has written about the deep
shadows but I see differences up to the 56 RGB number, the
Qimage numbers stay lower there.

Photoshop. CS
I have three color engine choices in PS now: ACE, ICM, CMM.
The last is the Kodak engine that came with the Kodak Color
Custom Tools. What can be expected ACE does it right, CMM
does it almost exactly like ACE in all choices. ICM
with Black Point Compensation on is not correct. Roy should
add a comment to the profile Read Me that it is better to use
ACE in PS CS for correct conversions.

Based on this I will keep my workflow as follows:
Scanning B&W film in Vuescan, filmtype selected that suits the 
real film best, AdobeRGB as the output color space, 16 bit 
grey Tiff, the Tiff will not have AdobeRGB assigned but gets 
the 2.2 Gamma when AdobeRGB is selected as the general output 
space in Vuescan,  import in PS on ACE color engine, P2P 
convert to QRT Gray-Lab with Black Point Compensation, save, 
edit, save, convert in PS to one of the QTR Gray paper 
profiles with BPC on. Then either print through Qimage with CM 
off or directly through QTR.

This way it can be a 16 bit greyscale Tiff for most of the 
time, have all the advantages of the right previews, and a 
consistent color management. I find it now more interesting to 
get the QTR Gray-Lab profile assigned to a greyscale Tiff in 
Vuescan than getting it right in Qimage. The water at that 
side has to calm down first I think.

Ernst