Yahoo Groups archive

QTR-Quadtone RIP

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

Thread

Invalid Linearize curve message

Invalid Linearize curve message

2007-10-13 by Louis Dina

I am using QTR 2.5.2.1 on a Windows XP machine.  I keep getting an 
error message when trying to linearize that reads:

"Invalid linearize curve -- not constantly increaseing.  Curve 
Creation failed."

I export data to a tab delimited text format and drag it onto the QTR-
Linearize-Data.exe file and I get a nice smooth ramp without any 
reversals reversals in the density or luminance data.  Then I copy 
the "linearize" data line in the bottom of the generated text file 
and paste it into the Linearization window of the curve creation 
module.  The original data looks fine to me.  

What the heck am I doing wrong?  Below is the linearize data for this 
file:

QTR-Linearize-Data version 2.5.2.0

LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 63.64 58.9 
53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"

Thanks, Lou Dina

Re: [QuadtoneRIP] Invalid Linearize curve message

2007-10-13 by Jim Thyer

Hi Diana,

Remembering when I did my linearization the numbers were in ascending order, so try reversing them.

Jim Thyer
Show quoted textHide quoted text
  ----- Original Message ----- 
  From: Louis Dina 
  To: QuadtoneRIP@yahoogroups.com 
  Sent: Saturday, October 13, 2007 3:37 PM
  Subject: [QuadtoneRIP] Invalid Linearize curve message


  I am using QTR 2.5.2.1 on a Windows XP machine. I keep getting an 
  error message when trying to linearize that reads:

  "Invalid linearize curve -- not constantly increaseing. Curve 
  Creation failed."

  I export data to a tab delimited text format and drag it onto the QTR-
  Linearize-Data.exe file and I get a nice smooth ramp without any 
  reversals reversals in the density or luminance data. Then I copy 
  the "linearize" data line in the bottom of the generated text file 
  and paste it into the Linearization window of the curve creation 
  module. The original data looks fine to me. 

  What the heck am I doing wrong? Below is the linearize data for this 
  file:

  QTR-Linearize-Data version 2.5.2.0

  LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 63.64 58.9 
  53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"

  Thanks, Lou Dina



   

[Non-text portions of this message have been removed]

Re: Invalid Linearize curve message

2007-10-13 by Louis Dina

Thanks, Jim, but I don't think that is the problem.  I have other 
files that linearized just fine, and they list the L* data in 
descending order exactly as presented below.  

I did just upgrade to QTR 2.5.2.1, so maybe this is a bug?  It's 
probably something I am doing.  Perhaps the two or three darkest 
patches are too close together.  I did notice that the QTR-Linearize-
Data version is 2.5.2.0.

Lou 

--- In QuadtoneRIP@yahoogroups.com, "Jim Thyer" <jimth@...> wrote:
>
> Hi Diana,
> 
> Remembering when I did my linearization the numbers were in 
ascending order, so try reversing them.
> 
> Jim Thyer
> 
>   ----- Original Message ----- 
>   From: Louis Dina 
>   To: QuadtoneRIP@yahoogroups.com 
>   Sent: Saturday, October 13, 2007 3:37 PM
>   Subject: [QuadtoneRIP] Invalid Linearize curve message
> 
> 
>   I am using QTR 2.5.2.1 on a Windows XP machine. I keep getting an 
>   error message when trying to linearize that reads:
> 
>   "Invalid linearize curve -- not constantly increaseing. Curve 
>   Creation failed."
> 
>   I export data to a tab delimited text format and drag it onto the 
QTR-
>   Linearize-Data.exe file and I get a nice smooth ramp without any 
>   reversals reversals in the density or luminance data. Then I copy 
>   the "linearize" data line in the bottom of the generated text 
file 
>   and paste it into the Linearization window of the curve creation 
>   module. The original data looks fine to me. 
> 
>   What the heck am I doing wrong? Below is the linearize data for 
this 
>   file:
> 
>   QTR-Linearize-Data version 2.5.2.0
> 
>   LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 63.64 
58.9 
Show quoted textHide quoted text
>   53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"
> 
>   Thanks, Lou Dina
> 
> 
> 
>    
> 
> [Non-text portions of this message have been removed]
>

Re: Invalid Linearize curve message - The Answer

2007-10-13 by Louis Dina

I think I figured out what I was doing wrong and wanted to share it 
in case others are experiencing the same thing.  Not sure if it is 
the best solution, but it seemed to work and give me a better profile.

I had my default ink limit set fairly high initially, where the curve 
was getting very flat.  My Eye One showed very slight increases in 
density at this point, but by eye it looked black a few steps 
before.  The ink limit I chose loaded up the shadow areas with a lot 
of ink, and while there was 'a little separation' between tones, it 
wasn't much.  The 90, 95 and 100% tones were bunched up pretty 
close.  So, I lowered the ink limit (in this case from 65 to 45, then 
bumped the black boost up to about 70.  This changed the shape of the 
black curve substantially and gave me a lot more separation between 
shadow values (and a better Dmax).  The program accepted these values 
without any error messages.

So, even though there was a slight separation between tones, I am 
guessing the program has some default parameters for accepting or 
rejecting linearization data, and my points were just too close 
together to get a passing grade.  Just as well, since I am sure I 
ended up with a better profile.

If this seems like a convoluted approach, please feel free to set me 
straight.  

Thanks, Lou

--- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@...> wrote:
>
> Thanks, Jim, but I don't think that is the problem.  I have other 
> files that linearized just fine, and they list the L* data in 
> descending order exactly as presented below.  
> 
> I did just upgrade to QTR 2.5.2.1, so maybe this is a bug?  It's 
> probably something I am doing.  Perhaps the two or three darkest 
> patches are too close together.  I did notice that the QTR-
Linearize-
> Data version is 2.5.2.0.
> 
> Lou 
> 
> --- In QuadtoneRIP@yahoogroups.com, "Jim Thyer" <jimth@> wrote:
> >
> > Hi Diana,
> > 
> > Remembering when I did my linearization the numbers were in 
> ascending order, so try reversing them.
> > 
> > Jim Thyer
> > 
> >   ----- Original Message ----- 
> >   From: Louis Dina 
> >   To: QuadtoneRIP@yahoogroups.com 
> >   Sent: Saturday, October 13, 2007 3:37 PM
> >   Subject: [QuadtoneRIP] Invalid Linearize curve message
> > 
> > 
> >   I am using QTR 2.5.2.1 on a Windows XP machine. I keep getting 
an 
> >   error message when trying to linearize that reads:
> > 
> >   "Invalid linearize curve -- not constantly increaseing. Curve 
> >   Creation failed."
> > 
> >   I export data to a tab delimited text format and drag it onto 
the 
> QTR-
> >   Linearize-Data.exe file and I get a nice smooth ramp without 
any 
> >   reversals reversals in the density or luminance data. Then I 
copy 
> >   the "linearize" data line in the bottom of the generated text 
> file 
> >   and paste it into the Linearization window of the curve 
creation 
> >   module. The original data looks fine to me. 
> > 
> >   What the heck am I doing wrong? Below is the linearize data for 
> this 
> >   file:
> > 
> >   QTR-Linearize-Data version 2.5.2.0
> > 
> >   LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 
63.64 
Show quoted textHide quoted text
> 58.9 
> >   53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"
> > 
> >   Thanks, Lou Dina
> > 
> > 
> > 
> >    
> > 
> > [Non-text portions of this message have been removed]
> >
>

Re: Invalid Linearize curve message - The Answer

2007-10-13 by joemulligan_68

I've had this problem before when trying to use the Create ICC program when trying to 
create a proflle for an older printer that wasn't laying down the inks linearly.  However, 
there was truly a reversal rather than just the small differences that you point out.

My other question is about your method of linearization.  I have never used the Linearize 
Data exe file to linearize as you mention.  I read the densities through Gretag Measure tool  
and simply type in the densities in the Curve Creation tab.  Once you save the curve 
Linearization happens automatically.  I get the impression that you are taking Linearized 
data and throwing it in curve creation tab and then saving this file.  You could be 
linearizing twice.  Correct me if I didn't interpret you correctly.  

--- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@...> wrote:
Show quoted textHide quoted text
>
> I think I figured out what I was doing wrong and wanted to share it 
> in case others are experiencing the same thing.  Not sure if it is 
> the best solution, but it seemed to work and give me a better profile.
> 
> I had my default ink limit set fairly high initially, where the curve 
> was getting very flat.  My Eye One showed very slight increases in 
> density at this point, but by eye it looked black a few steps 
> before.  The ink limit I chose loaded up the shadow areas with a lot 
> of ink, and while there was 'a little separation' between tones, it 
> wasn't much.  The 90, 95 and 100% tones were bunched up pretty 
> close.  So, I lowered the ink limit (in this case from 65 to 45, then 
> bumped the black boost up to about 70.  This changed the shape of the 
> black curve substantially and gave me a lot more separation between 
> shadow values (and a better Dmax).  The program accepted these values 
> without any error messages.
> 
> So, even though there was a slight separation between tones, I am 
> guessing the program has some default parameters for accepting or 
> rejecting linearization data, and my points were just too close 
> together to get a passing grade.  Just as well, since I am sure I 
> ended up with a better profile.
> 
> If this seems like a convoluted approach, please feel free to set me 
> straight.  
> 
> Thanks, Lou
> 
> --- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@> wrote:
> >
> > Thanks, Jim, but I don't think that is the problem.  I have other 
> > files that linearized just fine, and they list the L* data in 
> > descending order exactly as presented below.  
> > 
> > I did just upgrade to QTR 2.5.2.1, so maybe this is a bug?  It's 
> > probably something I am doing.  Perhaps the two or three darkest 
> > patches are too close together.  I did notice that the QTR-
> Linearize-
> > Data version is 2.5.2.0.
> > 
> > Lou 
> > 
> > --- In QuadtoneRIP@yahoogroups.com, "Jim Thyer" <jimth@> wrote:
> > >
> > > Hi Diana,
> > > 
> > > Remembering when I did my linearization the numbers were in 
> > ascending order, so try reversing them.
> > > 
> > > Jim Thyer
> > > 
> > >   ----- Original Message ----- 
> > >   From: Louis Dina 
> > >   To: QuadtoneRIP@yahoogroups.com 
> > >   Sent: Saturday, October 13, 2007 3:37 PM
> > >   Subject: [QuadtoneRIP] Invalid Linearize curve message
> > > 
> > > 
> > >   I am using QTR 2.5.2.1 on a Windows XP machine. I keep getting 
> an 
> > >   error message when trying to linearize that reads:
> > > 
> > >   "Invalid linearize curve -- not constantly increaseing. Curve 
> > >   Creation failed."
> > > 
> > >   I export data to a tab delimited text format and drag it onto 
> the 
> > QTR-
> > >   Linearize-Data.exe file and I get a nice smooth ramp without 
> any 
> > >   reversals reversals in the density or luminance data. Then I 
> copy 
> > >   the "linearize" data line in the bottom of the generated text 
> > file 
> > >   and paste it into the Linearization window of the curve 
> creation 
> > >   module. The original data looks fine to me. 
> > > 
> > >   What the heck am I doing wrong? Below is the linearize data for 
> > this 
> > >   file:
> > > 
> > >   QTR-Linearize-Data version 2.5.2.0
> > > 
> > >   LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 
> 63.64 
> > 58.9 
> > >   53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"
> > > 
> > >   Thanks, Lou Dina
> > > 
> > > 
> > > 
> > >    
> > > 
> > > [Non-text portions of this message have been removed]
> > >
> >
>

Re: Invalid Linearize curve message - The Answer

2007-10-13 by Louis Dina

Joe,

The whole problem was having a few points on the curve, that while 
not showing reversal, were too close to each other, and then rejected 
by the linearization routine.

Roy created a droplet program that allows you to simply drop a text 
file with the scanned Lab readings on it, and the droplet puts it in 
sorted order.  All you need to do is highlight the raw sorted data, 
copy it, then paste it into QTR's linearization screen.  Works great.


Thanks for your thoughts, Lou.


--- In QuadtoneRIP@yahoogroups.com, "joemulligan_68" 
<joemulligan_68@...> wrote:
>
> I've had this problem before when trying to use the Create ICC 
program when trying to 
> create a proflle for an older printer that wasn't laying down the 
inks linearly.  However, 
> there was truly a reversal rather than just the small differences 
that you point out.
> 
> My other question is about your method of linearization.  I have 
never used the Linearize 
> Data exe file to linearize as you mention.  I read the densities 
through Gretag Measure tool  
> and simply type in the densities in the Curve Creation tab.  Once 
you save the curve 
> Linearization happens automatically.  I get the impression that you 
are taking Linearized 
> data and throwing it in curve creation tab and then saving this 
file.  You could be 
> linearizing twice.  Correct me if I didn't interpret you 
correctly.  
> 
> --- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@> wrote:
> >
> > I think I figured out what I was doing wrong and wanted to share 
it 
> > in case others are experiencing the same thing.  Not sure if it 
is 
> > the best solution, but it seemed to work and give me a better 
profile.
> > 
> > I had my default ink limit set fairly high initially, where the 
curve 
> > was getting very flat.  My Eye One showed very slight increases 
in 
> > density at this point, but by eye it looked black a few steps 
> > before.  The ink limit I chose loaded up the shadow areas with a 
lot 
> > of ink, and while there was 'a little separation' between tones, 
it 
> > wasn't much.  The 90, 95 and 100% tones were bunched up pretty 
> > close.  So, I lowered the ink limit (in this case from 65 to 45, 
then 
> > bumped the black boost up to about 70.  This changed the shape of 
the 
> > black curve substantially and gave me a lot more separation 
between 
> > shadow values (and a better Dmax).  The program accepted these 
values 
> > without any error messages.
> > 
> > So, even though there was a slight separation between tones, I am 
> > guessing the program has some default parameters for accepting or 
> > rejecting linearization data, and my points were just too close 
> > together to get a passing grade.  Just as well, since I am sure I 
> > ended up with a better profile.
> > 
> > If this seems like a convoluted approach, please feel free to set 
me 
> > straight.  
> > 
> > Thanks, Lou
> > 
> > --- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@> wrote:
> > >
> > > Thanks, Jim, but I don't think that is the problem.  I have 
other 
> > > files that linearized just fine, and they list the L* data in 
> > > descending order exactly as presented below.  
> > > 
> > > I did just upgrade to QTR 2.5.2.1, so maybe this is a bug?  
It's 
> > > probably something I am doing.  Perhaps the two or three 
darkest 
> > > patches are too close together.  I did notice that the QTR-
> > Linearize-
> > > Data version is 2.5.2.0.
> > > 
> > > Lou 
> > > 
> > > --- In QuadtoneRIP@yahoogroups.com, "Jim Thyer" <jimth@> wrote:
> > > >
> > > > Hi Diana,
> > > > 
> > > > Remembering when I did my linearization the numbers were in 
> > > ascending order, so try reversing them.
> > > > 
> > > > Jim Thyer
> > > > 
> > > >   ----- Original Message ----- 
> > > >   From: Louis Dina 
> > > >   To: QuadtoneRIP@yahoogroups.com 
> > > >   Sent: Saturday, October 13, 2007 3:37 PM
> > > >   Subject: [QuadtoneRIP] Invalid Linearize curve message
> > > > 
> > > > 
> > > >   I am using QTR 2.5.2.1 on a Windows XP machine. I keep 
getting 
> > an 
> > > >   error message when trying to linearize that reads:
> > > > 
> > > >   "Invalid linearize curve -- not constantly increaseing. 
Curve 
> > > >   Creation failed."
> > > > 
> > > >   I export data to a tab delimited text format and drag it 
onto 
> > the 
> > > QTR-
> > > >   Linearize-Data.exe file and I get a nice smooth ramp 
without 
> > any 
> > > >   reversals reversals in the density or luminance data. Then 
I 
> > copy 
> > > >   the "linearize" data line in the bottom of the generated 
text 
> > > file 
> > > >   and paste it into the Linearization window of the curve 
> > creation 
> > > >   module. The original data looks fine to me. 
> > > > 
> > > >   What the heck am I doing wrong? Below is the linearize data 
for 
> > > this 
> > > >   file:
> > > > 
> > > >   QTR-Linearize-Data version 2.5.2.0
> > > > 
> > > >   LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03 
> > 63.64 
> > > 58.9 
> > > >   53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 
6.39"
Show quoted textHide quoted text
> > > > 
> > > >   Thanks, Lou Dina
> > > > 
> > > > 
> > > > 
> > > >    
> > > > 
> > > > [Non-text portions of this message have been removed]
> > > >
> > >
> >
>

Re: [QuadtoneRIP] Re: Invalid Linearize curve message - The Answer

2007-10-15 by James Haney

Louis,

I struggled with the same thing for a couple of years until I finally  
realized the same thing you did. It is better to start with a more  
conservative max density for building curves, and after you have  
everything working right, then you can mess with BOOST K, OVERLAP, etc.

James Haney



On Oct 13, 2007, at 9:58 AM, Louis Dina wrote:

> I think I figured out what I was doing wrong and wanted to share it
> in case others are experiencing the same thing. Not sure if it is
> the best solution, but it seemed to work and give me a better profile.
>
> I had my default ink limit set fairly high initially, where the curve
> was getting very flat. My Eye One showed very slight increases in
> density at this point, but by eye it looked black a few steps
> before. The ink limit I chose loaded up the shadow areas with a lot
> of ink, and while there was 'a little separation' between tones, it
> wasn't much. The 90, 95 and 100% tones were bunched up pretty
> close. So, I lowered the ink limit (in this case from 65 to 45, then
> bumped the black boost up to about 70. This changed the shape of the
> black curve substantially and gave me a lot more separation between
> shadow values (and a better Dmax). The program accepted these values
> without any error messages.
>
> So, even though there was a slight separation between tones, I am
> guessing the program has some default parameters for accepting or
> rejecting linearization data, and my points were just too close
> together to get a passing grade. Just as well, since I am sure I
> ended up with a better profile.
>
> If this seems like a convoluted approach, please feel free to set me
> straight.
>
> Thanks, Lou
>
> --- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@...> wrote:
> >
> > Thanks, Jim, but I don't think that is the problem. I have other
> > files that linearized just fine, and they list the L* data in
> > descending order exactly as presented below.
> >
> > I did just upgrade to QTR 2.5.2.1, so maybe this is a bug? It's
> > probably something I am doing. Perhaps the two or three darkest
> > patches are too close together. I did notice that the QTR-
> Linearize-
> > Data version is 2.5.2.0.
> >
> > Lou
> >
> > --- In QuadtoneRIP@yahoogroups.com, "Jim Thyer" <jimth@> wrote:
> > >
> > > Hi Diana,
> > >
> > > Remembering when I did my linearization the numbers were in
> > ascending order, so try reversing them.
> > >
> > > Jim Thyer
> > >
> > > ----- Original Message -----
> > > From: Louis Dina
> > > To: QuadtoneRIP@yahoogroups.com
> > > Sent: Saturday, October 13, 2007 3:37 PM
> > > Subject: [QuadtoneRIP] Invalid Linearize curve message
> > >
> > >
> > > I am using QTR 2.5.2.1 on a Windows XP machine. I keep getting
> an
> > > error message when trying to linearize that reads:
> > >
> > > "Invalid linearize curve -- not constantly increaseing. Curve
> > > Creation failed."
> > >
> > > I export data to a tab delimited text format and drag it onto
> the
> > QTR-
> > > Linearize-Data.exe file and I get a nice smooth ramp without
> any
> > > reversals reversals in the density or luminance data. Then I
> copy
> > > the "linearize" data line in the bottom of the generated text
> > file
> > > and paste it into the Linearization window of the curve
> creation
> > > module. The original data looks fine to me.
> > >
> > > What the heck am I doing wrong? Below is the linearize data for
> > this
> > > file:
> > >
> > > QTR-Linearize-Data version 2.5.2.0
> > >
> > > LINEARIZE="95.63 93.09 88.98 85.03 81.23 76.85 71.99 68.03
> 63.64
> > 58.9
> > > 53.9 49.73 44.41 39.95 34.64 28.78 22.71 15.46 10.57 6.85 6.39"
> > >
> > > Thanks, Lou Dina
> > >
> > >
> > >
> > >
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
>
>
> 



[Non-text portions of this message have been removed]

RE: [QuadtoneRIP] Re: Invalid Linearize curve message - The Answer

2007-10-15 by Paul Roark

I think my experiences are consistent with the views of this thread.


>> I had my default ink limit set fairly high initially, 
>> where the curve was getting very flat. 
>>... The 90, 95 and 100% tones were bunched up pretty close. 

Some set their dmax not to the point of absolute darkest, but to that point
where the curve becomes so flat the noise and variables might cause future
posterization.   

Usually the higher the load of an ink (more pigment particles relative to
base), the sooner it will reach the point where the curve's slope is too low
to be useful or reliable.

The MK is the highest load ink.  So, I use it for the default ink load.
Usually you can then simply ignore the black boost and other ink limits.  I
set this with a single 21-step curve printed with a straight line curve
profile.  I make the curve by entering the coordinates (0, 0) and (100, 100)
in QTR's "Point List."  (Tools>Curve Creation>Curve>Point List tab)  In
short, on 21-step print takes care of dmax and all ink limit (and black
boost) issues in making the profile.

>... It is better to start with a more 
>conservative max density for building curves, and after you have 
>everything working right, then you can mess with BOOST K, OVERLAP, etc.

Among other things, I prefer to have a reasonable good curve going into the
linearization.   With respect to the adjustments in the tabs between the Ink
Setup and Linearization tabs, I leave adjustments in their default
positions.  As one expert once commented about Gimp print in another
context, it has lots of controls, but most of them are not worth bothering
with.

Paul
www.PaulRoark.com

Re: [QuadtoneRIP] Re: Invalid Linearize curve message - The Answer

2007-10-15 by James Haney

I couldn't agree more.

In fact, after backing off my K ink limit, generating curves and  
linearizing, I haven't needed to "mess" with BOOST K or OVERLAP.

I am living quite happily with a very simple set of curves created  
with the most minimal set of options available and producing better  
prints than the efforts I have previously made using all of the  
whistles and bells to get the absolute darkest max density. I have  
found that I have usually ended up with unsatisfactory shadow  
gradation when I have tried too hard.

James




On Oct 15, 2007, at 1:41 PM, Paul Roark wrote:

> I think my experiences are consistent with the views of this thread.
>
> >> I had my default ink limit set fairly high initially,
> >> where the curve was getting very flat.
> >>... The 90, 95 and 100% tones were bunched up pretty close.
>
> Some set their dmax not to the point of absolute darkest, but to  
> that point
> where the curve becomes so flat the noise and variables might cause  
> future
> posterization.
>
> Usually the higher the load of an ink (more pigment particles  
> relative to
> base), the sooner it will reach the point where the curve's slope  
> is too low
> to be useful or reliable.
>
> The MK is the highest load ink. So, I use it for the default ink load.
> Usually you can then simply ignore the black boost and other ink  
> limits. I
> set this with a single 21-step curve printed with a straight line  
> curve
> profile. I make the curve by entering the coordinates (0, 0) and  
> (100, 100)
> in QTR's "Point List." (Tools>Curve Creation>Curve>Point List tab) In
> short, on 21-step print takes care of dmax and all ink limit (and  
> black
> boost) issues in making the profile.
>
> >... It is better to start with a more
> >conservative max density for building curves, and after you have
> >everything working right, then you can mess with BOOST K, OVERLAP,  
> etc.
>
> Among other things, I prefer to have a reasonable good curve going  
> into the
> linearization. With respect to the adjustments in the tabs between  
> the Ink
> Setup and Linearization tabs, I leave adjustments in their default
> positions. As one expert once commented about Gimp print in another
> context, it has lots of controls, but most of them are not worth  
> bothering
> with.
>
> Paul
> www.PaulRoark.com
>
>
>
> 



[Non-text portions of this message have been removed]

Re: Invalid Linearize curve message - The Answer

2007-10-16 by Louis Dina

Hi Paul.

You lost me with this statement:

"I set this with a single 21-step curve printed with a straight line 
curve profile.  I make the curve by entering the coordinates (0, 0) 
and (100, 100) in QTR's "Point List."  (Tools>Curve 
Creation>Curve>Point List tab)  In short, on 21-step print takes care 
of dmax and all ink limit (and black > boost) issues in making the 
profile."

I understand the concept of setting the default ink limit just before 
it reaches the unreliable flat zone, but my copy of QTR doesn't seem 
to have the features you refer to (at least I can't find it).  I am 
using v2.5.2.0 on a Windows XP machine.

When I go to Tools>Curve Creation, I see no "Curve" or "Point List" 
tabs to select.  I'd love to understand exactly what you are doing 
and recommending, but I am confused at this point.

Thanks, Lou


--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@...> 
wrote:
>
> I think my experiences are consistent with the views of this thread.
> 
> 
> >> I had my default ink limit set fairly high initially, 
> >> where the curve was getting very flat. 
> >>... The 90, 95 and 100% tones were bunched up pretty close. 
> 
> Some set their dmax not to the point of absolute darkest, but to 
that point
> where the curve becomes so flat the noise and variables might cause 
future
> posterization.   
> 
> Usually the higher the load of an ink (more pigment particles 
relative to
> base), the sooner it will reach the point where the curve's slope 
is too low
> to be useful or reliable.
> 
> The MK is the highest load ink.  So, I use it for the default ink 
load.
> Usually you can then simply ignore the black boost and other ink 
limits.  I
> set this with a single 21-step curve printed with a straight line 
curve
> profile.  I make the curve by entering the coordinates (0, 0) and 
(100, 100)
> in QTR's "Point List."  (Tools>Curve Creation>Curve>Point List 
tab)  In
> short, on 21-step print takes care of dmax and all ink limit (and 
black
> boost) issues in making the profile.
> 
> >... It is better to start with a more 
> >conservative max density for building curves, and after you have 
> >everything working right, then you can mess with BOOST K, OVERLAP, 
etc.
> 
> Among other things, I prefer to have a reasonable good curve going 
into the
> linearization.   With respect to the adjustments in the tabs 
between the Ink
> Setup and Linearization tabs, I leave adjustments in their default
> positions.  As one expert once commented about Gimp print in another
> context, it has lots of controls, but most of them are not worth 
bothering
Show quoted textHide quoted text
> with.
> 
> Paul
> www.PaulRoark.com
>

Re: Invalid Linearize curve message - The Answer

2007-10-16 by Louis Dina

Hi Paul.

I finally found it....I have to use the "Load Curve" option for the 
ink, then click the curve button.  Not sure I totally understand the 
procedure yet, but I found these features in QTR.

Lou

--- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@...> wrote:
>
> Hi Paul.
> 
> You lost me with this statement:
> 
> "I set this with a single 21-step curve printed with a straight 
line 
> curve profile.  I make the curve by entering the coordinates (0, 0) 
> and (100, 100) in QTR's "Point List."  (Tools>Curve 
> Creation>Curve>Point List tab)  In short, on 21-step print takes 
care 
> of dmax and all ink limit (and black > boost) issues in making the 
> profile."
> 
> I understand the concept of setting the default ink limit just 
before 
> it reaches the unreliable flat zone, but my copy of QTR doesn't 
seem 
> to have the features you refer to (at least I can't find it).  I am 
> using v2.5.2.0 on a Windows XP machine.
> 
> When I go to Tools>Curve Creation, I see no "Curve" or "Point List" 
> tabs to select.  I'd love to understand exactly what you are doing 
> and recommending, but I am confused at this point.
> 
> Thanks, Lou
> 
> 
> --- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@> 
> wrote:
> >
> > I think my experiences are consistent with the views of this 
thread.
> > 
> > 
> > >> I had my default ink limit set fairly high initially, 
> > >> where the curve was getting very flat. 
> > >>... The 90, 95 and 100% tones were bunched up pretty close. 
> > 
> > Some set their dmax not to the point of absolute darkest, but to 
> that point
> > where the curve becomes so flat the noise and variables might 
cause 
> future
> > posterization.   
> > 
> > Usually the higher the load of an ink (more pigment particles 
> relative to
> > base), the sooner it will reach the point where the curve's slope 
> is too low
> > to be useful or reliable.
> > 
> > The MK is the highest load ink.  So, I use it for the default ink 
> load.
> > Usually you can then simply ignore the black boost and other ink 
> limits.  I
> > set this with a single 21-step curve printed with a straight line 
> curve
> > profile.  I make the curve by entering the coordinates (0, 0) and 
> (100, 100)
> > in QTR's "Point List."  (Tools>Curve Creation>Curve>Point List 
> tab)  In
> > short, on 21-step print takes care of dmax and all ink limit (and 
> black
> > boost) issues in making the profile.
> > 
> > >... It is better to start with a more 
> > >conservative max density for building curves, and after you have 
> > >everything working right, then you can mess with BOOST K, 
OVERLAP, 
> etc.
> > 
> > Among other things, I prefer to have a reasonable good curve 
going 
> into the
> > linearization.   With respect to the adjustments in the tabs 
> between the Ink
> > Setup and Linearization tabs, I leave adjustments in their default
> > positions.  As one expert once commented about Gimp print in 
another
Show quoted textHide quoted text
> > context, it has lots of controls, but most of them are not worth 
> bothering
> > with.
> > 
> > Paul
> > www.PaulRoark.com
> >
>

Re: Invalid Linearize curve message - The Answer

2007-10-16 by Louis Dina

--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@...> 
wrote:

> The MK is the highest load ink.  So, I use it for the default ink 
load.
> Usually you can then simply ignore the black boost and other ink 
limits.  I
> set this with a single 21-step curve printed with a straight line 
curve
> profile.  I make the curve by entering the coordinates (0, 0) and 
(100, 100)
> in QTR's "Point List."  (Tools>Curve Creation>Curve>Point List 
tab)  In
> short, on 21-step print takes care of dmax and all ink limit (and 
black
> boost) issues in making the profile.


Paul,

Now that I have found the Load Curve and Point List tabs, can you 
elaborate a little.  I amstill unclear on what to do.  Do you use 
just those two points above?  And only for the black ink or for all 
inks.  I was under the impression that printing without any 
linearization data entere was the same as printing with 0,0 and 
100,100.  I guess that is incorrect?

I have read the following description on your website and it appears 
similar in many respects, but I didn't see a reference to the above.  

"QTR, 2200, 4K+, Matte BW, Carbon & general monotone inkset curve 
profiling"

Just a little confused, but your approach does seem to elimiate 
significant iterations and give a better result, so I'd like to 
understand it.  Is this only used for Quad inks, or can it also be 
used for a color inkset like the UC inks?

Thanks, Lou Dina

RE: [QuadtoneRIP] Re: Invalid Linearize curve message - The Answer

2007-10-16 by Paul Roark

Hi Lou


>--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@...> 
wrote:

>> The MK is the highest load ink. So, I use it for the 
>>default ink load.  Usually you can then simply ignore 
>>the black boost and other ink limits. I
>> set this with a single 21-step curve printed with a 
>>straight line curve profile. I make the curve by 
>>entering the coordinates (0, 0) and (100, 100)
>> in QTR's "Point List." (Tools>Curve Creation>Curve>
>>Point List tab) In short, on 21-step print takes 
>>care of dmax and all ink limit (and black
>> boost) issues in making the profile.


>Now that I have found the Load Curve and Point List tabs,
> can you elaborate a little... Do you use 
>just those two points above?

Yes, those 2 end point define a straight line curve from (0, 0) to (100,
100). I set the ink load for the black ink at 100.  All other inks are
turned off.  I then make a profile named, in effect, as a K straight line 0
- 100.  Then, I print a 21-step file with that profile.  The x% label on the
21-step test print where the dmax is reached defines the default ink limit.

> And only for the black ink or for all inks.

I just print the black ink for the default in limit.

> I was under the impression that printing without any 
>linearization data entered was the same as printing 
>with 0,0 and 100,100. I guess that is incorrect?

There needs to be some underlying profile.  The linearization just makes an
overlaying correction curve.  (Not that I know the details of the
programming.  You'll need to ask Roy for that.)

>I have read the following description on your website and it 
>appears similar in many respects, but I didn't see a reference 
>to the above. 

>"QTR, 2200, 4K+, Matte BW, Carbon & general monotone inkset curve 
>profiling"

Those were some preliminary experiments and notes.  I think as you gain
experience you develop your own style as to how you do things.  I've
increasingly tried to minimize the variables but at also directly describing
the curves in the point lists.  As an example, for partitioning, the
automatic system is a good start and most would stop there.  I, however, do
a screen grab of it and turn the curve into point lists.  Then I directly
manipulate that to get a fairly straight carbon core.  This allows for
simpler toner curves.

> ... your approach does seem to elimiate 
>significant iterations and give a better result, 

That's the goal.  I hope it's going in that direction.  I'm really not an
expert in QTR, however.

> Is this only used for Quad inks, or can it also be 
> used for a color inkset like the UC inks?

See what I've done with the R260 at
http://www.paulroark.com/BW-Info/R260-Color.pdf  I'd love for QTR to be able
to pass RGB information through it, but at this time it converts an RGB file
into a GS file.  The color profiling I've been able to do just with
neutralizing a 21-step is surprisingly good.  I just printed some color
photos using this plus an increase in the saturation (which I do for most
MIS color pigs to match the Epson colors) that are about the best color I've
ever printed.  If QTR could do this, it would allow me/us much more
flexibility to, for example, have an 1800 with CMY in it that we drive with
QTR for color, and then have the remaining slots filled with more
interesting B&W oriented inks. 

Paul
www.PaulRoark.com

Re: Invalid Linearize curve message - The Answer

2007-10-17 by Louis Dina

Thanks, Paul.

I think I understand what you are doing to set the black ink limit.  
Printing the 21 step grayscale with the straight line profile 
eliminates the need to print the initial calibration target at 100%.  
Do you pick the absolute blackest black as read with a spectro for 
your black Dmax or do you back off a little bit?

I assume that once you have established the black ink limit using 
this approach, you then print the calibration target (with 
calibration mode now turned on) at the above determined black ink 
limit and proceed with partitioning the other inks.

It also appears you are still trying to stick with QTR default 
settings for all settings on the Gray Curve and Toner Curve tabs.

Sorry for all the questions, but it does get a little complex with so 
many options and approaches.  I guess no matter how you get there, 
the key is to get smooth curves and transitions, no unwanted color 
crossovers or reversals, a linear tone scale, and the best Dmax 
possible (without getting into unstable territory).  Having built 
some pretty good profiles (with a LOT of iterations) I am looking for 
a simpler, faster, more effective way.    

Regards,
Lou

--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@...> 
wrote:
>
> Hi Lou
> 
> 
> >--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@> 
> wrote:
> 
> >> The MK is the highest load ink. So, I use it for the 
> >>default ink load.  Usually you can then simply ignore 
> >>the black boost and other ink limits. I
> >> set this with a single 21-step curve printed with a 
> >>straight line curve profile. I make the curve by 
> >>entering the coordinates (0, 0) and (100, 100)
> >> in QTR's "Point List." (Tools>Curve Creation>Curve>
> >>Point List tab) In short, on 21-step print takes 
> >>care of dmax and all ink limit (and black
> >> boost) issues in making the profile.
> 
> 
> >Now that I have found the Load Curve and Point List tabs,
> > can you elaborate a little... Do you use 
> >just those two points above?
> 
> Yes, those 2 end point define a straight line curve from (0, 0) to 
(100,
> 100). I set the ink load for the black ink at 100.  All other inks 
are
> turned off.  I then make a profile named, in effect, as a K 
straight line 0
> - 100.  Then, I print a 21-step file with that profile.  The x% 
label on the
> 21-step test print where the dmax is reached defines the default 
ink limit.
> 
> > And only for the black ink or for all inks.
> 
> I just print the black ink for the default in limit.
> 
> > I was under the impression that printing without any 
> >linearization data entered was the same as printing 
> >with 0,0 and 100,100. I guess that is incorrect?
> 
> There needs to be some underlying profile.  The linearization just 
makes an
> overlaying correction curve.  (Not that I know the details of the
> programming.  You'll need to ask Roy for that.)
> 
> >I have read the following description on your website and it 
> >appears similar in many respects, but I didn't see a reference 
> >to the above. 
> 
> >"QTR, 2200, 4K+, Matte BW, Carbon & general monotone inkset curve 
> >profiling"
> 
> Those were some preliminary experiments and notes.  I think as you 
gain
> experience you develop your own style as to how you do things.  I've
> increasingly tried to minimize the variables but at also directly 
describing
> the curves in the point lists.  As an example, for partitioning, the
> automatic system is a good start and most would stop there.  I, 
however, do
> a screen grab of it and turn the curve into point lists.  Then I 
directly
> manipulate that to get a fairly straight carbon core.  This allows 
for
> simpler toner curves.
> 
> > ... your approach does seem to elimiate 
> >significant iterations and give a better result, 
> 
> That's the goal.  I hope it's going in that direction.  I'm really 
not an
> expert in QTR, however.
> 
> > Is this only used for Quad inks, or can it also be 
> > used for a color inkset like the UC inks?
> 
> See what I've done with the R260 at
> http://www.paulroark.com/BW-Info/R260-Color.pdf  I'd love for QTR 
to be able
> to pass RGB information through it, but at this time it converts an 
RGB file
> into a GS file.  The color profiling I've been able to do just with
> neutralizing a 21-step is surprisingly good.  I just printed some 
color
> photos using this plus an increase in the saturation (which I do 
for most
> MIS color pigs to match the Epson colors) that are about the best 
color I've
> ever printed.  If QTR could do this, it would allow me/us much more
> flexibility to, for example, have an 1800 with CMY in it that we 
drive with
Show quoted textHide quoted text
> QTR for color, and then have the remaining slots filled with more
> interesting B&W oriented inks. 
> 
> Paul
> www.PaulRoark.com
>

RE: [QuadtoneRIP] Re: Invalid Linearize curve message - The Answer

2007-10-17 by Paul Roark

Hi Lou,

>...
>Printing the 21 step grayscale with the straight line 
>profile eliminates the need to print the initial 
>calibration target at 100%. 

Yes.  It's just faster.

>Do you pick the absolute blackest black as read with a 
>spectro for your black Dmax or do you back off a little bit?

I want to be sure there is continuously decreasing density below the dmax
point I pick. I try to pick a point that is at the edge of the "plateau" of
the ink density, even if there is a slightly higher point somewhere on the
plateau.

>I assume that once you have established the black ink limit 
>using this approach, you then print the calibration target 
>(with calibration mode now turned on) at the above determined 
>black ink limit and proceed with partitioning the other inks.

Yes (probably), if I were doing that type of partitioning.  

Note that recently my workflows have not used traditional partitioning. The
1800 3-MK, 100% Eboni workflow (http://www.paulroark.com/BW-Info/R1800.htm)
uses a "staged" beginning of the channels, but all continue with a positive
slope to the same 100% point.  However, this is a very different approach
than one would use with, for example, partitioning of a k3 printer.

The reason I added the "probably" above is that ink densities can probably
be just assumed.  The standard dilution ratio is about 0.3.  That is, each
light ink has about 30% as much pigment in it as the darker one.  The ink
load to density relationship is not linear, however, particularly at darkest
end.  So, the LK is not going to be just 30% of the K.  It'll probably be
more like 35% to 40%.  At the light end it'll be more linear.  That is, the
LLK will probably be about 30 - 33% of the LK.  I wonder if one just took
those numbers and plugged them in if that would be good enough.  Given what
I do to perfect the partitioning curve manually, I suspect I might just
ignore the second calibration curve also.  However, I don't want to confuse
everyone with my initial shortcuts and then additional manual steps that
most will not want to do.  So, most might be best advised to ignore this.

>It also appears you are still trying to stick with QTR default 
>settings for all settings on the Gray Curve and Toner Curve tabs.

Yes.  I think those tabs/controls might be useful to people who are stuck
with curves they can't otherwise alter, but I vastly prefer directly
correcting the underlying curves if there is a problem, as opposed to using
one of these tools.  But, again, if these tools work for people, then use
them.

>... it does get a little complex with so many options and approaches.

Yes.  On the one hand I have tried to simplify, but on the other hand, my
more manual approach to perfecting the curves may just confuse people more.
So, I've been reluctant to talk much about it on this forum.

At the risk of getting too far off the specific subject, I might add that
much of what I've done with the 1800 3-MK approach came out of not only my
desire for the most stable B&W prints but also after seeing many eyes glaze
over as I described profiling, partitioning, toning, Lab color approaches,
etc. to groups of B&W photographers.  I finally just dumped my notes at one
such talk and decided to turn the talk into a focus group session to see
what people wanted.  After doing this with a few groups of traditional B&W
printers and then getting their feedback on the perceived quality of
different types of sample prints I bring to such events, I concluded most
don't want anything to do with all this profiling.  They don't want to learn
the Lab color system - or any color system.  It they wanted that, they'd do
color printing.

I concluded that if black-only printing was acceptably smooth and neutral,
most B&W photographs would do that.  So, I bought an R1800 to see if the 1.5
pl dot would do the trick.  It did -- if and only if multiple channels were
used -- and the R1800 11 x 14 prints are now the bulk of my sales and the
ones I prefer to do.  

With the R1800 multi-channel Eboni MK workflow, all of the profiles use the
same basic, underlying curves.  So, profiling a new paper is a very fast
2-step process -- determine the dmax with a 21-step test print (and a
special profile I put in the package), and linearize.  That's pretty much
all users have to do, and a flatbed scanner is all they need.  Frankly, from
what I've seen in groups of traditional B&W printers, that's about the most
they really want to do.  And, while the people on these forums might be able
to detect a fine grained structure in the test strip midtones, none of the
B&W printers who have seen actual prints, and no prospective customer at the
gallery has every noticed the image structure.  

>I guess no matter how you get there, 
>the key is to get smooth curves and transitions, no unwanted 
>color crossovers or reversals, a linear tone scale, and the 
>best Dmax possible (without getting into unstable territory). 
>Having built some pretty good profiles (with a LOT of iterations) 
>I am looking for a simpler, faster, more effective way. 

See the above.  Join the club.

Then of course, add the frustrations of inevitable ink batch variances, not
to mention less than perfect mixing and dilution, and the reality of having
to re-do profiles starts to get a bit frustrating.  I personally, have found
happiness in an approach that has no transitions, no crossovers, no color,
no mixed inks and no dilutions, but is also the most lightfast and cheapest.


Paul
www.PaulRoark.com

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.