Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: QTR-Create-ICC tool. I'm about to give up.  (Read 1530 times)
texshooter
Full Member
***
Offline Offline

Posts: 218


« on: July 21, 2013, 01:04:16 AM »
ReplyReply

I've linearized my printer/paper combo for Epson ABW printing because I find the ABW mode results to be too dark in the shadows (I use the DARK tone setting, which looks better than LIGHT, NORMAL, DARKER, or DARKEST). My expectation from the linearization process was that this 21-step gray chart would print out such that no two steps appear the same. In my case, the last three darkest steps are blocked up on my Ilford Smooth Luster Duo paper, which I'm trying to remedy. Below are tables that show that the QTR-Create-ICC tool did little to change the densities of the 21-step chart print.  I can't see any difference between the printout of this chart before or after linearizing with QTR. Am I doing or expecting something wrong?

Logged
Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1674


WWW
« Reply #1 on: July 21, 2013, 11:04:09 AM »
ReplyReply

It's clear from the density readings that you have made that the numbers are in fact different but you say that visually you cannot tell the difference..  This could be because the paper doesn't resolve these black patches very well which isn't surprising.  I wouldn't expect QTR to have much if any impact on the heel of the ABW curve.  In my experience it's more to do with the midrange tones.
Logged

texshooter
Full Member
***
Offline Offline

Posts: 218


« Reply #2 on: July 21, 2013, 03:04:56 PM »
ReplyReply

There are three QTR profiling scripts to choose from:

QTR-CREATE-ICC
QTR-CREATE-ICC-RGB
QTR-LINEARIZE-DATA

I don't remember which one I ran. Is there a difference between these?
Logged
RHPS
Jr. Member
**
Offline Offline

Posts: 76


« Reply #3 on: July 22, 2013, 02:08:48 AM »
ReplyReply

Just looking at the original data I would say that your printer is giving you a pretty linear response to start with, so I wouldn't expect to see any great improvement from "linearization".

What isn't clear to me is what you actually did between the second and third tables. You clearly created an ICC profile, but did you then use it when you re-printed the 21-step chart though ABW? You can only do this on the Windows platform - the Mac doesn't allow it any more. But, as I said at the start, I wouldn't expect to see any significant change to something that is already very linear.

As far as the visual side goes, I suspect that you just need a brighter viewing light to see the difference in the very dark tones. The Lab measurements clearly show that these tones are differentiated, but you need a lot of light to actually see the difference. You might like to try holding the print in front of a bright light and looking through it, like a transparency - it's easier to see the steps that way.
Logged
texshooter
Full Member
***
Offline Offline

Posts: 218


« Reply #4 on: July 22, 2013, 02:43:50 AM »
ReplyReply

Just looking at the original data I would say that your printer is giving you a pretty linear response to start with, so I wouldn't expect to see any great improvement from "linearization".

What isn't clear to me is what you actually did between the second and third tables. You clearly created an ICC profile, but did you then use it when you re-printed the 21-step chart though ABW?

Yes, I used the ICC profile when printing with Windows. Because, the lab values after linearization were so close to the before values, my first rection was to assume the variation was due to spectrophotometer reading margin of error rather than ICC adjustments. But now I'm starting to believe my Epson printer's ABW mode is as good as it gets. In which case, QTR is obsolete. My guess is that QTR was worth its weight in gold before Epson released their ABW mode--that was when, 2007? Long story short, I'm sticking with ABW. I can always tweek it with luminosity curves.
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3758


« Reply #5 on: July 22, 2013, 02:46:04 AM »
ReplyReply

Given that it is so hard to see the difference in the darkest 3 steps, while there is a measurable difference, maybe full linear isn't the best way for that paper?

Just a thought ...

Cheers,
Bart
Logged
Ernst Dinkla
Sr. Member
****
Offline Offline

Posts: 2867


« Reply #6 on: July 22, 2013, 06:17:23 AM »
ReplyReply

There are three QTR profiling scripts to choose from:

QTR-CREATE-ICC
QTR-CREATE-ICC-RGB
QTR-LINEARIZE-DATA

I don't remember which one I ran. Is there a difference between these?

In QTR you start with the linearize tool after the partitioning of the inks and use that in the QTR driver. Then you can build the ICC profile and depending on the application you print from you can select the top one for greyscale or the RGB one if the application is not reliable with a greyscale ICC profile.

If it is about ICC profiling an ABW mode of an OEM driver, Epson, HP, etc, you use the RGB one and you have to rely on ABW being linear already or the linear character is made in the calibration step of the printer (HP). The last is what I use with my HP Z3100. Qimage to print from.
There is another possibility with the normal drivers where you create curves in Photoshop to linearize the print more by trial and error (OEM inks) or partition/linearize custom quad inks and incorporate the curves in an ICC printer profile with the ICC profile creator. Use the RGB one is my advice. Paul Roark asked Roy to implant that feature. That is what I use to drive an HP Officejet K5400 with 4 diluted HP Vivera monochrome inks. Qimage to print from.

You can still shift things that they become more to your taste if for example the viewing light is too low to show the density steps in the shadows. The (spectro)meter measured density differences so the paper can cope and a bad meter would probably have difficulty to distinguish differences in the shadows instead of exaggerate them.

Some time after the QTR tools were created for QTR itself QTR (linearisation and) profiling for OEM drivers started with the appearance of ABW in the Epson 2400 (Steve Kale) and not before ABW appeared. It is more likely that ABW got better in time.

--
Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
July 2013, 500+ inkjet media white spectral plots.
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad