Digital BW, The Print

> From: Paul Roark <paul.roark@...>

> 
>> ... The outbound portion of what
>> QTR Create ICC does is simply automating this measurement, calculation and
>> adjustment transformation curve.
> 
> Yes, and it requires the i1 or similar, I assume.

It requires a device which can measure luminance.  If you are not worried
about soft proofing for hue then you could simply plug the a* and b* data
with zeros.  If you want to proof hue then the device needs to be a
spectrophotometer. You just need to get the data in the form that QTR Create
ICC can read it.  It then does the scaling etc and generates the specific
type of file that the ICC spec demands - not insignificant work.


> 
> Are those ICCs useable in the Print with Preview procedure with the Epson
> driver?

Some clarification.  When you are in Print with Preview you have not left PS
yet.  Here you are telling PS to either do or not do a conversion of the
file data for a selected ICC profile BEFORE hand off to the Epson driver.

Answer:  Yes I use them all the time.  I have profiled my Epson Adv B&W
settings, one for each hue - neutral, warm, sepia and cool.  I do not mess
with the darker and other settings (other than hue).  I prefer to edit the
image in PS rather than in the driver which has limited editing ability.

> 
> I use an X-Rite for my curves and simply input the points manually in
> Photoshop curves.  I don't find it a problem, but I understand that
> automation is needed for many.  I hope my tutorial helps those who want to
> do it via eyeball.
> 
> What I'm shooting for in a universal EZ B&W hextone (+ maybe septone)
> approach.  (I can make inks from K3 as well as MIS if needed -- no one gets
> a monopoly here.)  I'll help make the initial curves for the different
> printers, but I want a very low entry barrier, student and starving
> artist-compatibel, system that uses the Epson driver.  I'd really welcome
> the QTR ICC approach alongside the manual PS curves approach to linearizing.
> One advantage of the ICC approach is that PS Elements can use them and not
> Transfer Functions -- at least in the Elements version I have.

One area worth a lot of thought is the thing I mentioned above.  QTR Create
ICC simply needs the data set up in the right way.  If you don't want to
soft proof hue then you don't need to measure a* and b* and can simply plug
the data with zeros.  So now all you need is a densitometer and a method for
getting data organised in the right format.  Now you're down to a cheap
densitometer and a $50 shareware fee.  That's really not that bad.

I can attest to the enormous amount of work Roy put into QTR Create ICC.  I
guess I understand it a reasonable amount - I never tried to get into the
actual math of the scaling - and I've described in summary form in this
thread a lot of ground that took Roy (and I) weeks/months to understand.  He
then had to go on and programme it which simply left me in the dust.  To do
the transfer curve approach you at least need to programme all the scaling
maths (even if just in a spreadsheet).  I have no idea how someone would go
about eyeballing that.  I would have thought the cheap densitometer, data
formatting guide (or Roy's expansion of the data flexibility of QTR Create
ICC) and the $50 shareware fee would be an admirable cheap entry point.