Yahoo Groups archive

QTR-Quadtone RIP

Index last updated: 2026-04-28 23:12 UTC

Thread

can't properly linearize ink profile in QTR 2.4.3.10

can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Philip Schwartz

My workflow:
- create ink profile in QTR
- select curve and print 21 step tablet
- measure steps after dry-down using X-Rite 810
- enter measured values into QTR linearization array and regenerate curve

Problem:
Steps never achieve the ideal density for the measured dMax.
When I linearize using Photoshop transfer curves, I can get within
.01 for all steps, but I am often off by at least .05 - .10 in
multiple steps when I linearize using QTR. Multiple iterations through
the process don't improve matters; things often get worse. 

Can anyone shed some light on this? I assume I should be entering
*measured* values into the linearization array, and not *derired*
values. Too bad QTR doesn't present the interpolated curve and let
users edit it directly ...

Re: can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Jeff Randall

--- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@...> 
wrote:
>
> My workflow:
> - create ink profile in QTR
> - select curve and print 21 step tablet
> - measure steps after dry-down using X-Rite 810
> - enter measured values into QTR linearization array and 
regenerate curve
> 
> Problem:
> Steps never achieve the ideal density for the measured dMax.
> When I linearize using Photoshop transfer curves, I can get within
> .01 for all steps, but I am often off by at least .05 - .10 in
> multiple steps when I linearize using QTR. Multiple iterations 
through
> the process don't improve matters; things often get worse. 
> 
> Can anyone shed some light on this? I assume I should be entering
> *measured* values into the linearization array, and not *derired*
> values. Too bad QTR doesn't present the interpolated curve and let
> users edit it directly ...

.
.
When you create the initial non-linear ink profile, the linerization 
array should be empty.

When you do "multiple iterations through the process" do you clear 
the linerization array each time?

Re: can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Philip Schwartz

Sorry, I should have specified this in my workflow. Yes, I first
generate the curve with the array empty. For each iteration, I clear
out the array, regenerate the curve, then populate the array and regen
the curve again. 
--- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@...> wrote:
Show quoted textHide quoted text
>
> --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@> 
> wrote:
> >
> > My workflow:
> > - create ink profile in QTR
> > - select curve and print 21 step tablet
> > - measure steps after dry-down using X-Rite 810
> > - enter measured values into QTR linearization array and 
> regenerate curve
> > 
> > Problem:
> > Steps never achieve the ideal density for the measured dMax.
> > When I linearize using Photoshop transfer curves, I can get within
> > .01 for all steps, but I am often off by at least .05 - .10 in
> > multiple steps when I linearize using QTR. Multiple iterations 
> through
> > the process don't improve matters; things often get worse. 
> > 
> > Can anyone shed some light on this? I assume I should be entering
> > *measured* values into the linearization array, and not *derired*
> > values. Too bad QTR doesn't present the interpolated curve and let
> > users edit it directly ...
> 
> .
> .
> When you create the initial non-linear ink profile, the linerization 
> array should be empty.
> 
> When you do "multiple iterations through the process" do you clear 
> the linerization array each time?
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Terence Lowe

I, too struggled with the QTR linearization process.  My approach was to 
try to get the best density measurements by modifying the ink setup and 
grey curve tabs, but with the linearization array clear.  This took many 
iterations.  Eventually, when I thought that the measured density curve 
was close to the ideal density curve (I plotted both in Excel for 
comparison purposes), I linearized, printed the step tablet again and 
checked the new, linearized density measurements to confirm (or 
otherwise) that my final curve was good enough.

I made the assumption that it was incorrect to re-linearize the 
linearized curve during the many iterations and so only linearized once 
as the final step in curve creation..  I hope this was the correct 
assumption.

Cheers,

Terence Lowe.

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Philip Schwartz

Thanks Terence. I am a new to QTR but have lots of experience printing
fine art B&W. QTR has lots of controls, but they are not well
documented so results are often not predictable, making the profiling
process incredibly tedious. I am spending far too much time profiling,
and not enough time creating and printing images :( Hoping this will
change ...

--- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@...> wrote:
>
> I, too struggled with the QTR linearization process.  My approach
was to 
> try to get the best density measurements by modifying the ink setup and 
> grey curve tabs, but with the linearization array clear.  This took
many 
Show quoted textHide quoted text
> iterations.  Eventually, when I thought that the measured density curve 
> was close to the ideal density curve (I plotted both in Excel for 
> comparison purposes), I linearized, printed the step tablet again and 
> checked the new, linearized density measurements to confirm (or 
> otherwise) that my final curve was good enough.
> 
> I made the assumption that it was incorrect to re-linearize the 
> linearized curve during the many iterations and so only linearized once 
> as the final step in curve creation..  I hope this was the correct 
> assumption.
> 
> Cheers,
> 
> Terence Lowe.
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-20 by Jeff Randall

Phil: Have you printed Keith Coopers test image
<http://www.northlight-
images.co.uk/article_pages/black_and_white_test.html>
and looked at the circular gradient and printed the offsetgrad8.jpg 
file located in this group's Files section to see if you can 
actually see the non-linearities you are measuring?

Another thing to try is a 51 step calibration approach using the 51-
step file in the OneEye folder of the QTRip folder.  I had problems 
getting a linear profile using the 21-step approach, but the 51-step 
approach worked very well for me.  YMMV.

BTW, I stretched the 51step image to print landscape -- bigger 
patches for my clumsy hands.


--- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@...> 
wrote:
>
> Thanks Terence. I am a new to QTR but have lots of experience 
printing
> fine art B&W. QTR has lots of controls, but they are not well
> documented so results are often not predictable, making the 
profiling
> process incredibly tedious. I am spending far too much time 
profiling,
> and not enough time creating and printing images :( Hoping this 
will
> change ...
> 
> --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@> wrote:
> >
> > I, too struggled with the QTR linearization process.  My approach
> was to 
> > try to get the best density measurements by modifying the ink 
setup and 
> > grey curve tabs, but with the linearization array clear.  This 
took
> many 
> > iterations.  Eventually, when I thought that the measured 
density curve 
> > was close to the ideal density curve (I plotted both in Excel 
for 
> > comparison purposes), I linearized, printed the step tablet 
again and 
> > checked the new, linearized density measurements to confirm (or 
> > otherwise) that my final curve was good enough.
> > 
> > I made the assumption that it was incorrect to re-linearize the 
> > linearized curve during the many iterations and so only 
linearized once 
> > as the final step in curve creation..  I hope this was the 
correct 
Show quoted textHide quoted text
> > assumption.
> > 
> > Cheers,
> > 
> > Terence Lowe.
> >
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-21 by Philip Schwartz

Hi Jeff. I am familiar with Keith's image, but I didn't actually try
it. My reasoning: I need to create curves that will produce
(reasonably) identical prints on different papers. If I can't reliably
linearize my profiles this will be an exercise in futility. I may get
good prints, but none will match. I did also try the 51-step wedge,
but I kept getting QTR errors complaining that the linearization array
sequence was invalid because it was not constantly increasing so curve
generation failed. To be honest, I got this message most of the time
using 21 values, but did manage to generate a couple of curves. It
seems like QTR doesn't like multiple steps with the same value, or
steps that are too close together, or, horrors, a value that actually
decreases. All of these things can happen in my world, and I don't
know how to fool QTR without invalidating my measurement.
I also tried to load a Photoshop CS acv I use all the time, but got a
QTR error about invalid values. Sorry, I'm not in front of that
computer at the moment but I have captured all these errors on my home
 PC. 
--- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@...> wrote:
Show quoted textHide quoted text
>
> Phil: Have you printed Keith Coopers test image
> <http://www.northlight-
> images.co.uk/article_pages/black_and_white_test.html>
> and looked at the circular gradient and printed the offsetgrad8.jpg 
> file located in this group's Files section to see if you can 
> actually see the non-linearities you are measuring?
> 
> Another thing to try is a 51 step calibration approach using the 51-
> step file in the OneEye folder of the QTRip folder.  I had problems 
> getting a linear profile using the 21-step approach, but the 51-step 
> approach worked very well for me.  YMMV.
> 
> BTW, I stretched the 51step image to print landscape -- bigger 
> patches for my clumsy hands.
> 
> 
> --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@> 
> wrote:
> >
> > Thanks Terence. I am a new to QTR but have lots of experience 
> printing
> > fine art B&W. QTR has lots of controls, but they are not well
> > documented so results are often not predictable, making the 
> profiling
> > process incredibly tedious. I am spending far too much time 
> profiling,
> > and not enough time creating and printing images :( Hoping this 
> will
> > change ...
> > 
> > --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@> wrote:
> > >
> > > I, too struggled with the QTR linearization process.  My approach
> > was to 
> > > try to get the best density measurements by modifying the ink 
> setup and 
> > > grey curve tabs, but with the linearization array clear.  This 
> took
> > many 
> > > iterations.  Eventually, when I thought that the measured 
> density curve 
> > > was close to the ideal density curve (I plotted both in Excel 
> for 
> > > comparison purposes), I linearized, printed the step tablet 
> again and 
> > > checked the new, linearized density measurements to confirm (or 
> > > otherwise) that my final curve was good enough.
> > > 
> > > I made the assumption that it was incorrect to re-linearize the 
> > > linearized curve during the many iterations and so only 
> linearized once 
> > > as the final step in curve creation..  I hope this was the 
> correct 
> > > assumption.
> > > 
> > > Cheers,
> > > 
> > > Terence Lowe.
> > >
> >
>

RE: [QuadtoneRIP] Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-21 by Tom Moore

Phil

Looking over your posts on this topic it's not clear to me what you are
changing on each iteration. I don't know of any mechanism in QTR for one
linearization to refine the results of a previous linearization. Iteration
is useful, but as Terrence Lowe points out it should occur before
linearization. To elaborate somewhat on Terence's suggestion, here's what I
do:

After the curve is setup on the Ink Setup tab and I am happy with the limits
etc on that tab, I print a step wedge, measure the values and plot them in
Excel. I compare this non linearized curve with the so called ideal curve
for the dmax I am achieving with the ink/paper I'm using. If there is a
"big" discrepancy between the measure curve and ideal, I'll adjust the
Highlight, Shadow and Gamma parameters on the Gray Curve tab and print
another step wedge, measure and plot it. You may also have to adjust the
same parameters on the Toner tabs if appropriate.

I don't have any suggestion as to how to judge "big". All I can suggest is
that if you're not happy with the results of linearizing your curve, you
could try making your "big" smaller. Also, the Gray Curve parameters are
sometimes useful in eliminating flat spots or small reversals that make
linearization impossible.

I may adjust these parameters, print, measure and plot the results several
times before I feel the curve is close enough to get QTR to linearize it
satisfactorily. After doing several curves the number of iterations required
drops off. Of course once you've done one curve for an ink/paper combo you
can usually use the same values for other curves for that paper - at least
as a starting point.

As an indication of how much adjustment can be required, the default and
final values for these parameters for a curve I created for Premier Fine Art
are:
Parameter       Default         Final		
---------       -------         -----
Highlight         10              6
Shadow            10              4
Gamma              1             1.6

The Gamma adjusts the mid-tones and the other parameters adjust the obvious
parts of the curve. In general, "L"arger values "L"ighten the curve. As
noted in the User Guide, the Gamma adjustments are dramatic and should be
small compared to the other 2 parameters. Once you reach acceptable values
for these parameters, you can use the measured values from the last
iteration as the linearization values to create your final curve. If you
need to adjust these parameters again, you must clear the linearization
array and print and measure a step wedge so you can enter the new
linearization values (and regenerate a new curve).

I recall that adjusting these parameters was suggested in a post by Roy
because the linearization algorithm is not able to do an effective job on
curves that require a large adjustment. I don't recall him stating how close
you need to be to get a good linearization.

I created the curveval2.xls spreadsheet (available in the files section) as
a tool to aid in visualizing the curve. I use Xkey, a key-stroke grabber
from X-Rite, to place the values from my 810 directly into the spreadsheet.
I find I can make a quite good (in my humble opinion) curve using the 21
step wedge. The 51 step wedge is more work and, because of the smaller step
is more prone to reversals.

Hope this helps. Good luck.

Tom Moore
Show quoted textHide quoted text
> -----Original Message-----
> From: QuadtoneRIP@yahoogroups.com [mailto:QuadtoneRIP@yahoogroups.com] On
> Behalf Of Philip Schwartz
> Sent: Wednesday, December 20, 2006 7:19 PM
> To: QuadtoneRIP@yahoogroups.com
> Subject: [QuadtoneRIP] Re:can't properly linearize ink profile in QTR
> 2.4.3.10
> 
> Hi Jeff. I am familiar with Keith's image, but I didn't actually try
> it. My reasoning: I need to create curves that will produce
> (reasonably) identical prints on different papers. If I can't reliably
> linearize my profiles this will be an exercise in futility. I may get
> good prints, but none will match. I did also try the 51-step wedge,
> but I kept getting QTR errors complaining that the linearization array
> sequence was invalid because it was not constantly increasing so curve
> generation failed. To be honest, I got this message most of the time
> using 21 values, but did manage to generate a couple of curves. It
> seems like QTR doesn't like multiple steps with the same value, or
> steps that are too close together, or, horrors, a value that actually
> decreases. All of these things can happen in my world, and I don't
> know how to fool QTR without invalidating my measurement.
> I also tried to load a Photoshop CS acv I use all the time, but got a
> QTR error about invalid values. Sorry, I'm not in front of that
> computer at the moment but I have captured all these errors on my home
>  PC.
> --- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@...> wrote:
> >
> > Phil: Have you printed Keith Coopers test image
> > <http://www.northlight-
> > images.co.uk/article_pages/black_and_white_test.html>
> > and looked at the circular gradient and printed the offsetgrad8.jpg
> > file located in this group's Files section to see if you can
> > actually see the non-linearities you are measuring?
> >
> > Another thing to try is a 51 step calibration approach using the 51-
> > step file in the OneEye folder of the QTRip folder.  I had problems
> > getting a linear profile using the 21-step approach, but the 51-step
> > approach worked very well for me.  YMMV.
> >
> > BTW, I stretched the 51step image to print landscape -- bigger
> > patches for my clumsy hands.
> >
> >
> > --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@>
> > wrote:
> > >
> > > Thanks Terence. I am a new to QTR but have lots of experience
> > printing
> > > fine art B&W. QTR has lots of controls, but they are not well
> > > documented so results are often not predictable, making the
> > profiling
> > > process incredibly tedious. I am spending far too much time
> > profiling,
> > > and not enough time creating and printing images :( Hoping this
> > will
> > > change ...
> > >
> > > --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@> wrote:
> > > >
> > > > I, too struggled with the QTR linearization process.  My approach
> > > was to
> > > > try to get the best density measurements by modifying the ink
> > setup and
> > > > grey curve tabs, but with the linearization array clear.  This
> > took
> > > many
> > > > iterations.  Eventually, when I thought that the measured
> > density curve
> > > > was close to the ideal density curve (I plotted both in Excel
> > for
> > > > comparison purposes), I linearized, printed the step tablet
> > again and
> > > > checked the new, linearized density measurements to confirm (or
> > > > otherwise) that my final curve was good enough.
> > > >
> > > > I made the assumption that it was incorrect to re-linearize the
> > > > linearized curve during the many iterations and so only
> > linearized once
> > > > as the final step in curve creation..  I hope this was the
> > correct
> > > > assumption.
> > > >
> > > > Cheers,
> > > >
> > > > Terence Lowe.
> > > >
> > >
> >
> 
> 
> 
> 
> 
> Yahoo! Groups Links
> 
> 
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-21 by Philip Schwartz

Thanks for the clarity of your explanation -- I now see the workflow
in my original post was way too sketchy :) I am actually following
exactly the process you describe. I am obviously not using the
controls on the gray curve tab effectively. I'm sure the problem is
that the corrections I am making are not based on any understanding of
how the Highlight and Shadow parameters work on the Gray Curve tab. I
am clear on gamma, but I have been holding this param constant. I am
also using an Excel spreadsheet that plots the curve against the ideal
for a specific dMax. I think I just need to dive back in and get a
better feel for the QTR controls. I am also still not clear on exactly
how to partition inks and set limits, and if these decisions are
flawed, linearization is moot, but this is another topic. Any hints
here would be much appreciated. 

--- In QuadtoneRIP@yahoogroups.com, "Tom Moore" <r.t.moore@...> wrote:
>
> Phil
> 
> Looking over your posts on this topic it's not clear to me what you are
> changing on each iteration. I don't know of any mechanism in QTR for one
> linearization to refine the results of a previous linearization.
Iteration
> is useful, but as Terrence Lowe points out it should occur before
> linearization. To elaborate somewhat on Terence's suggestion, here's
what I
> do:
> 
> After the curve is setup on the Ink Setup tab and I am happy with
the limits
> etc on that tab, I print a step wedge, measure the values and plot
them in
> Excel. I compare this non linearized curve with the so called ideal
curve
> for the dmax I am achieving with the ink/paper I'm using. If there is a
> "big" discrepancy between the measure curve and ideal, I'll adjust the
> Highlight, Shadow and Gamma parameters on the Gray Curve tab and print
> another step wedge, measure and plot it. You may also have to adjust the
> same parameters on the Toner tabs if appropriate.
> 
> I don't have any suggestion as to how to judge "big". All I can
suggest is
> that if you're not happy with the results of linearizing your curve, you
> could try making your "big" smaller. Also, the Gray Curve parameters are
> sometimes useful in eliminating flat spots or small reversals that make
> linearization impossible.
> 
> I may adjust these parameters, print, measure and plot the results
several
> times before I feel the curve is close enough to get QTR to linearize it
> satisfactorily. After doing several curves the number of iterations
required
> drops off. Of course once you've done one curve for an ink/paper
combo you
> can usually use the same values for other curves for that paper - at
least
> as a starting point.
> 
> As an indication of how much adjustment can be required, the default and
> final values for these parameters for a curve I created for Premier
Fine Art
> are:
> Parameter       Default         Final		
> ---------       -------         -----
> Highlight         10              6
> Shadow            10              4
> Gamma              1             1.6
> 
> The Gamma adjusts the mid-tones and the other parameters adjust the
obvious
> parts of the curve. In general, "L"arger values "L"ighten the curve. As
> noted in the User Guide, the Gamma adjustments are dramatic and
should be
> small compared to the other 2 parameters. Once you reach acceptable
values
> for these parameters, you can use the measured values from the last
> iteration as the linearization values to create your final curve. If you
> need to adjust these parameters again, you must clear the linearization
> array and print and measure a step wedge so you can enter the new
> linearization values (and regenerate a new curve).
> 
> I recall that adjusting these parameters was suggested in a post by Roy
> because the linearization algorithm is not able to do an effective
job on
> curves that require a large adjustment. I don't recall him stating
how close
> you need to be to get a good linearization.
> 
> I created the curveval2.xls spreadsheet (available in the files
section) as
> a tool to aid in visualizing the curve. I use Xkey, a key-stroke grabber
> from X-Rite, to place the values from my 810 directly into the
spreadsheet.
> I find I can make a quite good (in my humble opinion) curve using the 21
> step wedge. The 51 step wedge is more work and, because of the
smaller step
> is more prone to reversals.
> 
> Hope this helps. Good luck.
> 
> Tom Moore
> 
> > -----Original Message-----
> > From: QuadtoneRIP@yahoogroups.com
[mailto:QuadtoneRIP@yahoogroups.com] On
> > Behalf Of Philip Schwartz
> > Sent: Wednesday, December 20, 2006 7:19 PM
> > To: QuadtoneRIP@yahoogroups.com
> > Subject: [QuadtoneRIP] Re:can't properly linearize ink profile in QTR
> > 2.4.3.10
> > 
> > Hi Jeff. I am familiar with Keith's image, but I didn't actually try
> > it. My reasoning: I need to create curves that will produce
> > (reasonably) identical prints on different papers. If I can't reliably
> > linearize my profiles this will be an exercise in futility. I may get
> > good prints, but none will match. I did also try the 51-step wedge,
> > but I kept getting QTR errors complaining that the linearization array
> > sequence was invalid because it was not constantly increasing so curve
> > generation failed. To be honest, I got this message most of the time
> > using 21 values, but did manage to generate a couple of curves. It
> > seems like QTR doesn't like multiple steps with the same value, or
> > steps that are too close together, or, horrors, a value that actually
> > decreases. All of these things can happen in my world, and I don't
> > know how to fool QTR without invalidating my measurement.
> > I also tried to load a Photoshop CS acv I use all the time, but got a
> > QTR error about invalid values. Sorry, I'm not in front of that
> > computer at the moment but I have captured all these errors on my home
> >  PC.
> > --- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@> wrote:
> > >
> > > Phil: Have you printed Keith Coopers test image
> > > <http://www.northlight-
> > > images.co.uk/article_pages/black_and_white_test.html>
> > > and looked at the circular gradient and printed the offsetgrad8.jpg
> > > file located in this group's Files section to see if you can
> > > actually see the non-linearities you are measuring?
> > >
> > > Another thing to try is a 51 step calibration approach using the 51-
> > > step file in the OneEye folder of the QTRip folder.  I had problems
> > > getting a linear profile using the 21-step approach, but the 51-step
> > > approach worked very well for me.  YMMV.
> > >
> > > BTW, I stretched the 51step image to print landscape -- bigger
> > > patches for my clumsy hands.
> > >
> > >
> > > --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@>
> > > wrote:
> > > >
> > > > Thanks Terence. I am a new to QTR but have lots of experience
> > > printing
> > > > fine art B&W. QTR has lots of controls, but they are not well
> > > > documented so results are often not predictable, making the
> > > profiling
> > > > process incredibly tedious. I am spending far too much time
> > > profiling,
> > > > and not enough time creating and printing images :( Hoping this
> > > will
> > > > change ...
> > > >
> > > > --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@> wrote:
> > > > >
> > > > > I, too struggled with the QTR linearization process.  My
approach
Show quoted textHide quoted text
> > > > was to
> > > > > try to get the best density measurements by modifying the ink
> > > setup and
> > > > > grey curve tabs, but with the linearization array clear.  This
> > > took
> > > > many
> > > > > iterations.  Eventually, when I thought that the measured
> > > density curve
> > > > > was close to the ideal density curve (I plotted both in Excel
> > > for
> > > > > comparison purposes), I linearized, printed the step tablet
> > > again and
> > > > > checked the new, linearized density measurements to confirm (or
> > > > > otherwise) that my final curve was good enough.
> > > > >
> > > > > I made the assumption that it was incorrect to re-linearize the
> > > > > linearized curve during the many iterations and so only
> > > linearized once
> > > > > as the final step in curve creation..  I hope this was the
> > > correct
> > > > > assumption.
> > > > >
> > > > > Cheers,
> > > > >
> > > > > Terence Lowe.
> > > > >
> > > >
> > >
> > 
> > 
> > 
> > 
> > 
> > Yahoo! Groups Links
> > 
> > 
> >
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-21 by Tom Moore

--- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@...>
wrote:
>
> Thanks for the clarity of your explanation -- I now see the workflow
> in my original post was way too sketchy :) I am actually following
> exactly the process you describe. 

 Well, If you had responded to Terrence Lowe's post I wouldn't have
written my detailed description.

I am obviously not using the
> controls on the gray curve tab effectively. I'm sure the problem is
> that the corrections I am making are not based on any understanding of
> how the Highlight and Shadow parameters work on the Gray Curve tab.

Other than larger values give lighter tones, there is not much to
understand. The rest is empirical.

 I
> am clear on gamma, but I have been holding this param constant. I am
> also using an Excel spreadsheet that plots the curve against the ideal
> for a specific dMax. I think I just need to dive back in and get a
> better feel for the QTR controls. I am also still not clear on exactly
> how to partition inks and set limits, and if these decisions are
> flawed, linearization is moot, but this is another topic. Any hints
> here would be much appreciated.

Yes. This is another area where some experience helps. I think that
setting ink limits can be the most common source of problems. Too much
ink can cause the problems you are having. The partition values
(relative densities) are less critical.

Creating your own curve for a paper/ink combination for which a curve
is included in QTR, can be useful in understanding the methodology for
curve creation.

Tom Moore

> 
> --- In QuadtoneRIP@yahoogroups.com, "Tom Moore" <r.t.moore@> wrote:
> >
> > Phil
> > 
> > Looking over your posts on this topic it's not clear to me what
you are
> > changing on each iteration. I don't know of any mechanism in QTR
for one
> > linearization to refine the results of a previous linearization.
> Iteration
> > is useful, but as Terrence Lowe points out it should occur before
> > linearization. To elaborate somewhat on Terence's suggestion, here's
> what I
> > do:
> > 
> > After the curve is setup on the Ink Setup tab and I am happy with
> the limits
> > etc on that tab, I print a step wedge, measure the values and plot
> them in
> > Excel. I compare this non linearized curve with the so called ideal
> curve
> > for the dmax I am achieving with the ink/paper I'm using. If there
is a
> > "big" discrepancy between the measure curve and ideal, I'll adjust the
> > Highlight, Shadow and Gamma parameters on the Gray Curve tab and print
> > another step wedge, measure and plot it. You may also have to
adjust the
> > same parameters on the Toner tabs if appropriate.
> > 
> > I don't have any suggestion as to how to judge "big". All I can
> suggest is
> > that if you're not happy with the results of linearizing your
curve, you
> > could try making your "big" smaller. Also, the Gray Curve
parameters are
> > sometimes useful in eliminating flat spots or small reversals that
make
> > linearization impossible.
> > 
> > I may adjust these parameters, print, measure and plot the results
> several
> > times before I feel the curve is close enough to get QTR to
linearize it
> > satisfactorily. After doing several curves the number of iterations
> required
> > drops off. Of course once you've done one curve for an ink/paper
> combo you
> > can usually use the same values for other curves for that paper - at
> least
> > as a starting point.
> > 
> > As an indication of how much adjustment can be required, the
default and
> > final values for these parameters for a curve I created for Premier
> Fine Art
> > are:
> > Parameter       Default         Final		
> > ---------       -------         -----
> > Highlight         10              6
> > Shadow            10              4
> > Gamma              1             1.6
> > 
> > The Gamma adjusts the mid-tones and the other parameters adjust the
> obvious
> > parts of the curve. In general, "L"arger values "L"ighten the
curve. As
> > noted in the User Guide, the Gamma adjustments are dramatic and
> should be
> > small compared to the other 2 parameters. Once you reach acceptable
> values
> > for these parameters, you can use the measured values from the last
> > iteration as the linearization values to create your final curve.
If you
> > need to adjust these parameters again, you must clear the
linearization
> > array and print and measure a step wedge so you can enter the new
> > linearization values (and regenerate a new curve).
> > 
> > I recall that adjusting these parameters was suggested in a post
by Roy
> > because the linearization algorithm is not able to do an effective
> job on
> > curves that require a large adjustment. I don't recall him stating
> how close
> > you need to be to get a good linearization.
> > 
> > I created the curveval2.xls spreadsheet (available in the files
> section) as
> > a tool to aid in visualizing the curve. I use Xkey, a key-stroke
grabber
> > from X-Rite, to place the values from my 810 directly into the
> spreadsheet.
> > I find I can make a quite good (in my humble opinion) curve using
the 21
> > step wedge. The 51 step wedge is more work and, because of the
> smaller step
> > is more prone to reversals.
> > 
> > Hope this helps. Good luck.
> > 
> > Tom Moore
> > 
> > > -----Original Message-----
> > > From: QuadtoneRIP@yahoogroups.com
> [mailto:QuadtoneRIP@yahoogroups.com] On
> > > Behalf Of Philip Schwartz
> > > Sent: Wednesday, December 20, 2006 7:19 PM
> > > To: QuadtoneRIP@yahoogroups.com
> > > Subject: [QuadtoneRIP] Re:can't properly linearize ink profile
in QTR
> > > 2.4.3.10
> > > 
> > > Hi Jeff. I am familiar with Keith's image, but I didn't actually try
> > > it. My reasoning: I need to create curves that will produce
> > > (reasonably) identical prints on different papers. If I can't
reliably
> > > linearize my profiles this will be an exercise in futility. I
may get
> > > good prints, but none will match. I did also try the 51-step wedge,
> > > but I kept getting QTR errors complaining that the linearization
array
> > > sequence was invalid because it was not constantly increasing so
curve
> > > generation failed. To be honest, I got this message most of the time
> > > using 21 values, but did manage to generate a couple of curves. It
> > > seems like QTR doesn't like multiple steps with the same value, or
> > > steps that are too close together, or, horrors, a value that
actually
> > > decreases. All of these things can happen in my world, and I don't
> > > know how to fool QTR without invalidating my measurement.
> > > I also tried to load a Photoshop CS acv I use all the time, but
got a
> > > QTR error about invalid values. Sorry, I'm not in front of that
> > > computer at the moment but I have captured all these errors on
my home
> > >  PC.
> > > --- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@>
wrote:
> > > >
> > > > Phil: Have you printed Keith Coopers test image
> > > > <http://www.northlight-
> > > > images.co.uk/article_pages/black_and_white_test.html>
> > > > and looked at the circular gradient and printed the
offsetgrad8.jpg
> > > > file located in this group's Files section to see if you can
> > > > actually see the non-linearities you are measuring?
> > > >
> > > > Another thing to try is a 51 step calibration approach using
the 51-
> > > > step file in the OneEye folder of the QTRip folder.  I had
problems
> > > > getting a linear profile using the 21-step approach, but the
51-step
> > > > approach worked very well for me.  YMMV.
> > > >
> > > > BTW, I stretched the 51step image to print landscape -- bigger
> > > > patches for my clumsy hands.
> > > >
> > > >
> > > > --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@>
> > > > wrote:
> > > > >
> > > > > Thanks Terence. I am a new to QTR but have lots of experience
> > > > printing
> > > > > fine art B&W. QTR has lots of controls, but they are not well
> > > > > documented so results are often not predictable, making the
> > > > profiling
> > > > > process incredibly tedious. I am spending far too much time
> > > > profiling,
> > > > > and not enough time creating and printing images :( Hoping this
> > > > will
> > > > > change ...
> > > > >
> > > > > --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@>
wrote:
> > > > > >
> > > > > > I, too struggled with the QTR linearization process.  My
> approach
> > > > > was to
> > > > > > try to get the best density measurements by modifying the ink
> > > > setup and
> > > > > > grey curve tabs, but with the linearization array clear.  This
> > > > took
> > > > > many
> > > > > > iterations.  Eventually, when I thought that the measured
> > > > density curve
> > > > > > was close to the ideal density curve (I plotted both in Excel
> > > > for
> > > > > > comparison purposes), I linearized, printed the step tablet
> > > > again and
> > > > > > checked the new, linearized density measurements to
confirm (or
> > > > > > otherwise) that my final curve was good enough.
> > > > > >
> > > > > > I made the assumption that it was incorrect to
re-linearize the
Show quoted textHide quoted text
> > > > > > linearized curve during the many iterations and so only
> > > > linearized once
> > > > > > as the final step in curve creation..  I hope this was the
> > > > correct
> > > > > > assumption.
> > > > > >
> > > > > > Cheers,
> > > > > >
> > > > > > Terence Lowe.
> > > > > >
> > > > >
> > > >
> > > 
> > > 
> > > 
> > > 
> > > 
> > > Yahoo! Groups Links
> > > 
> > > 
> > >
> >
>

Re:can't properly linearize ink profile in QTR 2.4.3.10

2006-12-21 by Philip Schwartz

Thanks, Tom. I did respond to Terence, but included more details in my
response to Jeff. I do appreciate everyone taking the time to help
out. I encountered a number of error messages when trying to gen
curves, load acv's, load profiles for editing, etc. which is
stretching my learning curve :) If someone is interested I could send
a summary of these offline. I haven't been able to find an error log 
for QTR -- does it maintain one? Would be helpful, especially as one
can't cut/paste from the error dialogs generated by the GUI.


--- In QuadtoneRIP@yahoogroups.com, "Tom Moore" <r.t.moore@...> wrote:
>
> --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz" <pschwart@>
> wrote:
> >
> > Thanks for the clarity of your explanation -- I now see the workflow
> > in my original post was way too sketchy :) I am actually following
> > exactly the process you describe. 
> 
>  Well, If you had responded to Terrence Lowe's post I wouldn't have
> written my detailed description.
> 
> I am obviously not using the
> > controls on the gray curve tab effectively. I'm sure the problem is
> > that the corrections I am making are not based on any understanding of
> > how the Highlight and Shadow parameters work on the Gray Curve tab.
> 
> Other than larger values give lighter tones, there is not much to
> understand. The rest is empirical.
> 
>  I
> > am clear on gamma, but I have been holding this param constant. I am
> > also using an Excel spreadsheet that plots the curve against the ideal
> > for a specific dMax. I think I just need to dive back in and get a
> > better feel for the QTR controls. I am also still not clear on exactly
> > how to partition inks and set limits, and if these decisions are
> > flawed, linearization is moot, but this is another topic. Any hints
> > here would be much appreciated.
> 
> Yes. This is another area where some experience helps. I think that
> setting ink limits can be the most common source of problems. Too much
> ink can cause the problems you are having. The partition values
> (relative densities) are less critical.
> 
> Creating your own curve for a paper/ink combination for which a curve
> is included in QTR, can be useful in understanding the methodology for
> curve creation.
> 
> Tom Moore
> 
> > 
> > --- In QuadtoneRIP@yahoogroups.com, "Tom Moore" <r.t.moore@> wrote:
> > >
> > > Phil
> > > 
> > > Looking over your posts on this topic it's not clear to me what
> you are
> > > changing on each iteration. I don't know of any mechanism in QTR
> for one
> > > linearization to refine the results of a previous linearization.
> > Iteration
> > > is useful, but as Terrence Lowe points out it should occur before
> > > linearization. To elaborate somewhat on Terence's suggestion, here's
> > what I
> > > do:
> > > 
> > > After the curve is setup on the Ink Setup tab and I am happy with
> > the limits
> > > etc on that tab, I print a step wedge, measure the values and plot
> > them in
> > > Excel. I compare this non linearized curve with the so called ideal
> > curve
> > > for the dmax I am achieving with the ink/paper I'm using. If there
> is a
> > > "big" discrepancy between the measure curve and ideal, I'll
adjust the
> > > Highlight, Shadow and Gamma parameters on the Gray Curve tab and
print
> > > another step wedge, measure and plot it. You may also have to
> adjust the
> > > same parameters on the Toner tabs if appropriate.
> > > 
> > > I don't have any suggestion as to how to judge "big". All I can
> > suggest is
> > > that if you're not happy with the results of linearizing your
> curve, you
> > > could try making your "big" smaller. Also, the Gray Curve
> parameters are
> > > sometimes useful in eliminating flat spots or small reversals that
> make
> > > linearization impossible.
> > > 
> > > I may adjust these parameters, print, measure and plot the results
> > several
> > > times before I feel the curve is close enough to get QTR to
> linearize it
> > > satisfactorily. After doing several curves the number of iterations
> > required
> > > drops off. Of course once you've done one curve for an ink/paper
> > combo you
> > > can usually use the same values for other curves for that paper - at
> > least
> > > as a starting point.
> > > 
> > > As an indication of how much adjustment can be required, the
> default and
> > > final values for these parameters for a curve I created for Premier
> > Fine Art
> > > are:
> > > Parameter       Default         Final		
> > > ---------       -------         -----
> > > Highlight         10              6
> > > Shadow            10              4
> > > Gamma              1             1.6
> > > 
> > > The Gamma adjusts the mid-tones and the other parameters adjust the
> > obvious
> > > parts of the curve. In general, "L"arger values "L"ighten the
> curve. As
> > > noted in the User Guide, the Gamma adjustments are dramatic and
> > should be
> > > small compared to the other 2 parameters. Once you reach acceptable
> > values
> > > for these parameters, you can use the measured values from the last
> > > iteration as the linearization values to create your final curve.
> If you
> > > need to adjust these parameters again, you must clear the
> linearization
> > > array and print and measure a step wedge so you can enter the new
> > > linearization values (and regenerate a new curve).
> > > 
> > > I recall that adjusting these parameters was suggested in a post
> by Roy
> > > because the linearization algorithm is not able to do an effective
> > job on
> > > curves that require a large adjustment. I don't recall him stating
> > how close
> > > you need to be to get a good linearization.
> > > 
> > > I created the curveval2.xls spreadsheet (available in the files
> > section) as
> > > a tool to aid in visualizing the curve. I use Xkey, a key-stroke
> grabber
> > > from X-Rite, to place the values from my 810 directly into the
> > spreadsheet.
> > > I find I can make a quite good (in my humble opinion) curve using
> the 21
> > > step wedge. The 51 step wedge is more work and, because of the
> > smaller step
> > > is more prone to reversals.
> > > 
> > > Hope this helps. Good luck.
> > > 
> > > Tom Moore
> > > 
> > > > -----Original Message-----
> > > > From: QuadtoneRIP@yahoogroups.com
> > [mailto:QuadtoneRIP@yahoogroups.com] On
> > > > Behalf Of Philip Schwartz
> > > > Sent: Wednesday, December 20, 2006 7:19 PM
> > > > To: QuadtoneRIP@yahoogroups.com
> > > > Subject: [QuadtoneRIP] Re:can't properly linearize ink profile
> in QTR
> > > > 2.4.3.10
> > > > 
> > > > Hi Jeff. I am familiar with Keith's image, but I didn't
actually try
> > > > it. My reasoning: I need to create curves that will produce
> > > > (reasonably) identical prints on different papers. If I can't
> reliably
> > > > linearize my profiles this will be an exercise in futility. I
> may get
> > > > good prints, but none will match. I did also try the 51-step
wedge,
> > > > but I kept getting QTR errors complaining that the linearization
> array
> > > > sequence was invalid because it was not constantly increasing so
> curve
> > > > generation failed. To be honest, I got this message most of
the time
> > > > using 21 values, but did manage to generate a couple of curves. It
> > > > seems like QTR doesn't like multiple steps with the same value, or
> > > > steps that are too close together, or, horrors, a value that
> actually
> > > > decreases. All of these things can happen in my world, and I don't
> > > > know how to fool QTR without invalidating my measurement.
> > > > I also tried to load a Photoshop CS acv I use all the time, but
> got a
> > > > QTR error about invalid values. Sorry, I'm not in front of that
> > > > computer at the moment but I have captured all these errors on
> my home
> > > >  PC.
> > > > --- In QuadtoneRIP@yahoogroups.com, "Jeff Randall" <jrandall@>
> wrote:
> > > > >
> > > > > Phil: Have you printed Keith Coopers test image
> > > > > <http://www.northlight-
> > > > > images.co.uk/article_pages/black_and_white_test.html>
> > > > > and looked at the circular gradient and printed the
> offsetgrad8.jpg
> > > > > file located in this group's Files section to see if you can
> > > > > actually see the non-linearities you are measuring?
> > > > >
> > > > > Another thing to try is a 51 step calibration approach using
> the 51-
> > > > > step file in the OneEye folder of the QTRip folder.  I had
> problems
> > > > > getting a linear profile using the 21-step approach, but the
> 51-step
> > > > > approach worked very well for me.  YMMV.
> > > > >
> > > > > BTW, I stretched the 51step image to print landscape -- bigger
> > > > > patches for my clumsy hands.
> > > > >
> > > > >
> > > > > --- In QuadtoneRIP@yahoogroups.com, "Philip Schwartz"
<pschwart@>
> > > > > wrote:
> > > > > >
> > > > > > Thanks Terence. I am a new to QTR but have lots of experience
> > > > > printing
> > > > > > fine art B&W. QTR has lots of controls, but they are not well
> > > > > > documented so results are often not predictable, making the
> > > > > profiling
> > > > > > process incredibly tedious. I am spending far too much time
> > > > > profiling,
> > > > > > and not enough time creating and printing images :( Hoping
this
> > > > > will
> > > > > > change ...
> > > > > >
> > > > > > --- In QuadtoneRIP@yahoogroups.com, Terence Lowe <televe@>
> wrote:
> > > > > > >
> > > > > > > I, too struggled with the QTR linearization process.  My
> > approach
> > > > > > was to
> > > > > > > try to get the best density measurements by modifying
the ink
> > > > > setup and
> > > > > > > grey curve tabs, but with the linearization array clear.
 This
> > > > > took
> > > > > > many
> > > > > > > iterations.  Eventually, when I thought that the measured
> > > > > density curve
> > > > > > > was close to the ideal density curve (I plotted both in
Excel
Show quoted textHide quoted text
> > > > > for
> > > > > > > comparison purposes), I linearized, printed the step tablet
> > > > > again and
> > > > > > > checked the new, linearized density measurements to
> confirm (or
> > > > > > > otherwise) that my final curve was good enough.
> > > > > > >
> > > > > > > I made the assumption that it was incorrect to
> re-linearize the
> > > > > > > linearized curve during the many iterations and so only
> > > > > linearized once
> > > > > > > as the final step in curve creation..  I hope this was the
> > > > > correct
> > > > > > > assumption.
> > > > > > >
> > > > > > > Cheers,
> > > > > > >
> > > > > > > Terence Lowe.
> > > > > > >
> > > > > >
> > > > >
> > > > 
> > > > 
> > > > 
> > > > 
> > > > 
> > > > Yahoo! Groups Links
> > > > 
> > > > 
> > > >
> > >
> >
>

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.