QTR-Quadtone RIP

Joost Horsten wrote:
> --- In QuadtoneRIP@yahoogroups.com, Howard Shaw <glassman@...> wrote:
>> Several greyscale inksets are designed to be compatible with each 
> other 
>> in this sense. I believe the different positions in the various 
>> piezography sets were all of matching density as are the current K7 
>> neutral & sepia sets. That is how inks in matching positions can be 
>> interchanged and used with the same curves.
>>
>> Similarly MIS FS & FSN, the UT7 carbon & cool toners and UT-R2-Warm 
> & 
>> Neutral all include similar density inks in the equivalent 
> positions of 
>> the warm & neutral/cool sets. In that sense the luminance should be 
> the
>> similar whichever way they are mixed & matched.
>>
>> Perhaps in this sense, one generic profile would work for each 
> family of 
>> inks.

Paul Roark always tried to match at least Dmax in the 
different hue curves for one inkset, the rest was at that time 
matched to the eye and no spectrometer profiling was done. Ink 
limiting and linearisation in QTR could/should be done in the 
same way. That could mean some Dmax for carbon or neutral 
black curves (depends on the black ink color) have to be 
limited if the other hue choices limit Dmax by nature. But 
toning black at Dmax is hardly possible so even that gives a 
limited difference. The perceptual curve fits the Dmax and 
paper white and by that it should be very identical.

I'm actually more interested in a single profile for ABW 
drivers and the Z3100 one in particular than making it work on 
QTR..The Z3100 ABW driver is based on internal spectrometer 
calibration that is also done on the 3-4 grey inks. I have 
checked that with HP.  Calibration doesn't have to mean 
linearisation but at least it creates a stable base. I expect 
similarities in tone range between prints made with the 
different hue selections in the ABW driver based on 
information from the same HP man. What seems to lack is the 
Lab curve as I asked whether they used the color ICC paper 
profile as a base for ABW and that was denied. No alternative 
profile was used either. He was interested in the concept. 
This is one of the color experts for the Z3100 at HP in 
Barcelona. Based on this information I think that a custom 
made greyscale profile created by averaging 3-4 measurements 
(3 toned + 1 the basic grey ink color) will bring the 
perceptual curve right to the center Dmax of that small gamut 
and average the perceptual curve as well. I'm not advocating a 
generic profile but a custom one that fits that particular 
printer, paper, inkset, small gamut choice. Given the 
integrated calibration of the Z3100 it will without doubt be 
generic for all Z3100s.

> You have a point here. I'm not familiar with the K7 sets. The UT7 
> carbon and cool toners are supposed to be the same as those of the 
> UT3D set. I did not study them in all detail, but if you look to the 
> canned QTR curves for these, there are actual slight differences. I 
> confess I did not try to take the same settings and compare the 
> results. In the ideal case, you and Ernst are right. I just don't 
> know how ideal reality is.

> 
>> I am not familiar with UT3d which you have been using and it sounds 
> like 
>> this is a more complex set of inks for which the above scenario 
> does not 
>> apply.
> 
> With the lab a /selenium toner in the UT3D set the situation is 
> different indeed.
> Unlike for the carbon and cool toners, that each have a LK and a LLK 
> version, the selenium toner has just a single LLK version. The LK 
> part of the curve has to be built by adding a mix of the cool and 
> warm LK toners. Perhaps one could rely on just taking 50% warm and 
> 50% cool, but I don't see why there is a garantuee that that actually 
> results in a linear curve. 
> 
> My point is that I sympathize with Ernsts approach, but that building 
> the framework on a fundament of three pillars in stead of just one 
> would be more robust/accurate AND flexible. One just does not have to 
> assume anything, but ane can actually compensate for any 
> deviations/peculiarities.

That's what I wrote in my private mail on the 3 
blending/slider concept. There are maybe 10 different quad + 
CMYKkkk inksets around and I expect that we get a very complex 
mailing list here when the users of all the inksets start to 
experiment with the new features.

I expect some tone range compensations in the ABW drivers for 
the hue shifts. Not that I didn't think of your method. Before 
going back to a single profile I proposed a solution where the 
ink mix is based on blending between 3 ICC profiles. That 
would ask for a color engine anyway but it would provide a 
good base for both soft proofing and printing. Probably too 
far fetched as Roy replied that there is no mixing tool for 
ICC profiles. Another idea I had was adding a B&W rendering to 
normal ICC color paper profiles, say Tobie's approach but 
incorporated, neutral only (and accepted by the ICE standards 
committee). That's even more futuristic if you see the rate of 
developments there. I'm back on the ground again.

Right now I think that splitting the perceptual tone curve 
profile from color softproofing and the use of ABW like 
drivers for adding color to prints + soft proofing is the best 
possibility right now as it allows the flexibility I seek + 
less profiles + keeping greyscale files instead of RGB. The 
last may suffer partly if this has to be done from Qimage 
which I prefer to do. But I can escape the B&W profiled color 
printing route Tobie proposes. So the next step will be: 
talking with Mike Chaney to see whether the ABW driver can be 
fed with greyscale Tiffs in Qimage, have color management do 
the profile conversion and see whether the ABW softproof is 
then showing the profile conversion already before the hue is 
chosen. If that flow has to be RGB as Qimage doesn't know 
anything else doesn't matter to me, the data is gone after 
printing. The Epson K3 ABW driver accepts both RGB and 
greyscale input so with that one it isn't a problem.

There are a number of reasons why I am aiming at this workflow 
instead of using QTR etc. The Windows version of QTR with the 
GUI doesn't fit Qimage well, the hot folder I proposed at that 
time to overcome part of the problems does work but it is not 
really making it a good tandem. Paper sizes do not translate 
back to Qimage, it is a slow chain and not always reliable in 
larger prints. Soft proofing isn't available in QTR. More 
(RGB) data is spooled along the way. The chance QTR becomes a 
normal Windows driver is nil at short term. I intend to buy a 
Z3100 and will sell the 9000 quad. That step has many 
advantages for B&W as well,  both in maintenance and quality. 
QTR can't drive the HPs (yet:-) and I do not see the Windows 
version getting the softproof ability of the ABW drivers nor 
get the many features of Qimage.

To be honest I expect that is the way many of us will go on 
the longer term despite the wide choice of Quad inksets and 
RIPs right now. I do not underestimate Roy's efforts but I do 
not see QTR supporting both HP and Canon printers and add so 
many features while new printer models appear one after the 
other right now.


Met vriendelijke groeten,Ernst


|  Dinkla Grafische Techniek  |
|     www.pigment-print.com    |
|             ( unvollendet )            |