QTR-Quadtone RIP

>
>  rossfmj  wrote:
> 
> > **
> >
> >
> > ... from Paul Roark's website;
> > http://home1.gte.net/res09aij/21-Step_GG22_L5.tif (marked for Gloss
> > Paper).
> >
>  I use the one at http://www.paulroark.com/BW-Info/21-step.jpg as it is
> easier to read from left to right.
> 
> 
> > Both of these show to not be linear in values for the graduations. Values
> > I get (% and on 255 8Bit Luminance Scale) are;
> >
> > Black 100% is 100% (0), 95% is 98% (5), 90% is 92% (19), 85% is 87% (33),
> > 80% is 81% (48), 75% is 76% ((62). The Luminance values do not match the
> > figures on Paul's test strip.
> >
> If you look at the "Grayscale" values, which is an option in the Info
> panel, do the values match the % markings on the test strips?  They do on
> my system.  The RGB numbers of the test strip you noted also match the RGB
> value in my Photoshop Info panel (except for 95% showing 13 instead of 12).
> 
> 
> >
> > Have these wedges been created with a conversion factor to open up the
> > blacks in linearisation?
> >
> 
> No.
> 
> Paul
> www.PaulRoark.com
> 
> 
> [Non-text portions of this message have been removed]
>

>
> The Info palette can be deceiving if you are not careful.   It only
> reports the contents of the file
> where the file type matches.  In other words grayscale file => look at
> K, RGB file => look at RGB.
> The others are all color managed conversions based on the default
> working profile set in Color Settings...
> 
> So what you are seeing in the RGB values is the conversion of working
> gray to working RGB.
> And the Lab values I think you are looking at are based on the working gray.
> 
> All step wedges ought to be untagged and evenly spaced.  (and I'd bet
> any you find on major site are).
> If you are on a Mac, printing targets with No color management is a
> bit tricky unless you use Print-Tool.
> 
> Roy
> 
> 
> On Tue, Feb 12, 2013 at 9:46 AM, rossfmj  wrote:
> > I am still having problems profiling UT14 to Epson Premium Glossy Photo Paper, particularly with setting the default ink limit and getting blocking up in the blacks.  I'm thinking I may be having a problem linearising the profiles.  In order to get my head round what is going on I have been checking the 21 Step Wedges I was using, one is a tif in the Eye One folder, and the other was from Paul Roark's website; http://home1.gte.net/res09aij/21-Step_GG22_L5.tif (marked for Gloss Paper).
> >
> > Both of these show to not be linear in values for the graduations.  Values I get (% and on 255 8Bit Luminance Scale) are;
> >
> > Black 100% is 100% (0), 95% is 98% (5), 90% is 92% (19),  85% is 87% (33),  80% is 81% (48), 75% is 76% ((62).  The Luminance values do not match the figures on Paul's test strip.
> >
> > I have also checked a 21 step wedge from Keith Cooper's site, this does match the % values.
> >
> > Have these wedges been created with a conversion factor to open up the blacks in linearisation?
> > Have I been using the wrong wedges?
> > Which wedge should I use for linearisation?
> > Is there a problem occurring in Elements when I open the images?
> > Might there be something else I am doing wrong?
> > Should I get my camera, enlarger and darkroom kit back from whom I gave it to?
> >
> > regards
> >
> > Ross
> >
> >
> 
> 
> -- 
> Roy Harrington
> roy@...
> www.harrington.com
>

>
> Ross,
> 
> As a general rule, whenever you are printing of step wedges to "see
> what the system does"
> you want to do it with No Color Management.
> 
> The inkseparation.tif file is an extra special case with bit values.
> 
> For QTR linearization the idea is printing a target using a curve with
> no-linearization, no-color-management.
> Reading the values, running QTR-Linearize-Data, and putting the values
> into the curve produces
> a correction that makes the steps linear in Lab value space.
> 
> For ICC profiling its the next level up.  Print a target using a
> linearized curve(s) (blends etc) with
> no-color-management.  When you run these values through QTR-Create-ICC
> you get a industry
> standard ICC profile that allows the system's color management to
> match up the luminosity (L)
> of your image file to the luminosity of your printing method.  Color
> Management of your screen is already
> doing this so adding this to your print workflow is what gives you
> screen-to-print matching.
> 
> In some sense QTR linearization and ICC profiling duplicate some of
> the functionality so that's
> why some generic ICC's can be and are provided.  QTR-Gray-Matte-Paper
> assumes a linear Lab curve with
> a lower dMax of matte paper, QTR-Gray-Photo-Paper assumes same with higher dMax.
> 
> BTW, the standard color ICC profiling works the same way -- print a
> target with no-cm, read it
> and make a profile.
> 
> If you have Print-Tool and can use it for all image printing its
> easier to get consistency -- for
> both QTR use and color printing thru Epson driver.  If you do custom
> color ICC's this makes
> it easier.
> 
> Roy
> 
> On Wed, Feb 13, 2013 at 12:45 AM, rossfmj <ross.jarvis@...> wrote:
> > Thank you Roy
> >
> > This highlights another issue which is confusing me.
> >
> > My understanding is that the ink separation. tif needs to be printed without colour management, so is in a way an "absolute" test.  For this I use QTR print tool.  However I thought that the step wedges were "subjective" and prints for linearisation were to be done through the process most like that of my normal workflow, so I've been doing these through Elements.  As you mention "No Color Management" appears not to be an option through Elements 11 on a Mac.
> >
> > Should I be printing step wedges for linearisation through QTR Print Tool?
> >
> > As mentioned earlier, I am not yet happy with the printing of shadows/blacks, particularly at the 90%+ values, and also the linearisation "bulls eyes" show rings and unevenness.
> >
> > Might I be getting too involved in the detail here?
> > Do these tests show some of the limitations of UT14, particularly with profiles only using 2 inks?
> >
> > regards
> >
> > Ross
> >
> 
> -- 
> Roy Harrington
> roy@...
> www.harrington.com
>

>
> What is the trick to getting the QTR Linearize Data app to actually linearize the data? I can't get beyond "Data not linear and can't be linearized" error. 
> 
> I print the step wedge, measure the L values, and graph them in exel. The graph shows a smooth (albeit not a straight line but pretty close) curve. When I drop the resulting .quad file into "QTR Linearize Data', I get the above error and nothing I do to make it more linear seems to correct the error. Very, very frustrating as I test this thing.
> 
> The Excel graph shows what appears to be a high-gamma curve with low contrast in the shadows and high contrast in the highlights, but each step is increasing in the L value over the entire curve. 
> 
> Thanks,
> JIm
> 
> 
> --- In QuadtoneRIP@yahoogroups.com, Roy Harrington <roy@> wrote:
> >
> > Ross,
> > 
> > As a general rule, whenever you are printing of step wedges to "see
> > what the system does"
> > you want to do it with No Color Management.
> > 
> > The inkseparation.tif file is an extra special case with bit values.
> > 
> > For QTR linearization the idea is printing a target using a curve with
> > no-linearization, no-color-management.
> > Reading the values, running QTR-Linearize-Data, and putting the values
> > into the curve produces
> > a correction that makes the steps linear in Lab value space.
> > 
> > For ICC profiling its the next level up.  Print a target using a
> > linearized curve(s) (blends etc) with
> > no-color-management.  When you run these values through QTR-Create-ICC
> > you get a industry
> > standard ICC profile that allows the system's color management to
> > match up the luminosity (L)
> > of your image file to the luminosity of your printing method.  Color
> > Management of your screen is already
> > doing this so adding this to your print workflow is what gives you
> > screen-to-print matching.
> > 
> > In some sense QTR linearization and ICC profiling duplicate some of
> > the functionality so that's
> > why some generic ICC's can be and are provided.  QTR-Gray-Matte-Paper
> > assumes a linear Lab curve with
> > a lower dMax of matte paper, QTR-Gray-Photo-Paper assumes same with higher dMax.
> > 
> > BTW, the standard color ICC profiling works the same way -- print a
> > target with no-cm, read it
> > and make a profile.
> > 
> > If you have Print-Tool and can use it for all image printing its
> > easier to get consistency -- for
> > both QTR use and color printing thru Epson driver.  If you do custom
> > color ICC's this makes
> > it easier.
> > 
> > Roy
> > 
> > On Wed, Feb 13, 2013 at 12:45 AM, rossfmj <ross.jarvis@> wrote:
> > > Thank you Roy
> > >
> > > This highlights another issue which is confusing me.
> > >
> > > My understanding is that the ink separation. tif needs to be printed without colour management, so is in a way an "absolute" test.  For this I use QTR print tool.  However I thought that the step wedges were "subjective" and prints for linearisation were to be done through the process most like that of my normal workflow, so I've been doing these through Elements.  As you mention "No Color Management" appears not to be an option through Elements 11 on a Mac.
> > >
> > > Should I be printing step wedges for linearisation through QTR Print Tool?
> > >
> > > As mentioned earlier, I am not yet happy with the printing of shadows/blacks, particularly at the 90%+ values, and also the linearisation "bulls eyes" show rings and unevenness.
> > >
> > > Might I be getting too involved in the detail here?
> > > Do these tests show some of the limitations of UT14, particularly with profiles only using 2 inks?
> > >
> > > regards
> > >
> > > Ross
> > >
> > 
> > -- 
> > Roy Harrington
> > roy@
> > www.harrington.com
> >
>

>
> Jim
> 
> I am by no means an expert.  I have had a number of these linearisation issues and the only way to get the thing to work is by changing the ink descriptor, usually by adjusting the default ink limit or tray value numbers, usually downward, until I can get the thing to linearise.  I put the linearise numbers in at the bottom of the ink descriptor and use install scripts.  I have not used QTR Linearize data.
> 
> I have found following instructions to print and read the inkseparation calibration prints do not necessarily lead to successful profile creation.
> 
> I haven't a clue why a steadily increasing curve is not recognised as such through the system, but I have had had a message saying something along the lines "cannot linearise as data not constantly increasing" even though it definately is!?  Maybe whatever happens during the linearisation process can go outside set parameters if the rate of increase is too high or low.
> 
> You may need to adjust some of the values by guesstimation before printing the step wedges.
> 
> Ross  
> 
> --- In QuadtoneRIP@yahoogroups.com, "James Stewart" <8jstewart@> wrote:
> >
> > What is the trick to getting the QTR Linearize Data app to actually linearize the data? I can't get beyond "Data not linear and can't be linearized" error. 
> > 
> > I print the step wedge, measure the L values, and graph them in exel. The graph shows a smooth (albeit not a straight line but pretty close) curve. When I drop the resulting .quad file into "QTR Linearize Data', I get the above error and nothing I do to make it more linear seems to correct the error. Very, very frustrating as I test this thing.
> > 
> > The Excel graph shows what appears to be a high-gamma curve with low contrast in the shadows and high contrast in the highlights, but each step is increasing in the L value over the entire curve. 
> > 
> > Thanks,
> > JIm
> > 
> > 
> > --- In QuadtoneRIP@yahoogroups.com, Roy Harrington <roy@> wrote:
> > >
> > > Ross,
> > > 
> > > As a general rule, whenever you are printing of step wedges to "see
> > > what the system does"
> > > you want to do it with No Color Management.
> > > 
> > > The inkseparation.tif file is an extra special case with bit values.
> > > 
> > > For QTR linearization the idea is printing a target using a curve with
> > > no-linearization, no-color-management.
> > > Reading the values, running QTR-Linearize-Data, and putting the values
> > > into the curve produces
> > > a correction that makes the steps linear in Lab value space.
> > > 
> > > For ICC profiling its the next level up.  Print a target using a
> > > linearized curve(s) (blends etc) with
> > > no-color-management.  When you run these values through QTR-Create-ICC
> > > you get a industry
> > > standard ICC profile that allows the system's color management to
> > > match up the luminosity (L)
> > > of your image file to the luminosity of your printing method.  Color
> > > Management of your screen is already
> > > doing this so adding this to your print workflow is what gives you
> > > screen-to-print matching.
> > > 
> > > In some sense QTR linearization and ICC profiling duplicate some of
> > > the functionality so that's
> > > why some generic ICC's can be and are provided.  QTR-Gray-Matte-Paper
> > > assumes a linear Lab curve with
> > > a lower dMax of matte paper, QTR-Gray-Photo-Paper assumes same with higher dMax.
> > > 
> > > BTW, the standard color ICC profiling works the same way -- print a
> > > target with no-cm, read it
> > > and make a profile.
> > > 
> > > If you have Print-Tool and can use it for all image printing its
> > > easier to get consistency -- for
> > > both QTR use and color printing thru Epson driver.  If you do custom
> > > color ICC's this makes
> > > it easier.
> > > 
> > > Roy
> > > 
> > > On Wed, Feb 13, 2013 at 12:45 AM, rossfmj <ross.jarvis@> wrote:
> > > > Thank you Roy
> > > >
> > > > This highlights another issue which is confusing me.
> > > >
> > > > My understanding is that the ink separation. tif needs to be printed without colour management, so is in a way an "absolute" test.  For this I use QTR print tool.  However I thought that the step wedges were "subjective" and prints for linearisation were to be done through the process most like that of my normal workflow, so I've been doing these through Elements.  As you mention "No Color Management" appears not to be an option through Elements 11 on a Mac.
> > > >
> > > > Should I be printing step wedges for linearisation through QTR Print Tool?
> > > >
> > > > As mentioned earlier, I am not yet happy with the printing of shadows/blacks, particularly at the 90%+ values, and also the linearisation "bulls eyes" show rings and unevenness.
> > > >
> > > > Might I be getting too involved in the detail here?
> > > > Do these tests show some of the limitations of UT14, particularly with profiles only using 2 inks?
> > > >
> > > > regards
> > > >
> > > > Ross
> > > >
> > > 
> > > -- 
> > > Roy Harrington
> > > roy@
> > > www.harrington.com
> > >
> >
>