QTR-Quadtone RIP

I apparently was making a mound out of a molehill...I have a reasonable understanding of 
QTR profiling now.  Pretty simple now that I have created my first profiles.    I was 
confused by reading the scattered documentations, which to my mind makes this more 
complicated than it really is.

I hope someone will answer a few more questions.  I did not get any answers to previous 
questions...apparently they were just too stupid!!!  Sorry if my questions were/are foolish.

I am using OSX.  Here are my questions.

1.  There is a droplet, DropQuadProfile.  Current version does not seem to function on 
Leopard.  Do I need to use this?  What I have been doing is simply editing the text file, 
putting in the appropriate profile folder, and running the terminal script to reinstall the 
profile.  Appears to work fine. Am I missing something?

2.  My assumption on linearization is that if I make some changes, ie add a toner curve, 
that I need to print the modified curve/step wedge  without any linearization values in the 
installed text file.  Is this correct?  I assume I would not get a correct linearization if the 
modified curve has any linearization text.  Please confirm that I am correct.  Once I have 
read the 21 step, I simply paste in the linearization values, right?...and reinstall to have a 
linearized curve.

3. I am still wondering what .quad files are?  They are in the Cone Ink areas.   Just curious. 
I would like to look at the text values, but have not found a way to view these curves.


Thanks for your patience with me.

Phil

On Thu, Jul 3, 2008 at 10:11 AM, brouwerkent <philip@...> wrote:
> I apparently was making a mound out of a molehill...I have a reasonable understanding of
> QTR profiling now.  Pretty simple now that I have created my first profiles.    I was
> confused by reading the scattered documentations, which to my mind makes this more
> complicated than it really is.
>
> I hope someone will answer a few more questions.  I did not get any answers to previous
> questions...apparently they were just too stupid!!!  Sorry if my questions were/are foolish.
>
> I am using OSX.  Here are my questions.
>
> 1.  There is a droplet, DropQuadProfile.  Current version does not seem to function on
> Leopard.  Do I need to use this?  What I have been doing is simply editing the text file,
> putting in the appropriate profile folder, and running the terminal script to reinstall the
> profile.  Appears to work fine. Am I missing something?

The install script is fine -- it's what I always use.

>
> 2.  My assumption on linearization is that if I make some changes, ie add a toner curve,
> that I need to print the modified curve/step wedge  without any linearization values in the
> installed text file.  Is this correct?  I assume I would not get a correct linearization if the
> modified curve has any linearization text.  Please confirm that I am correct.  Once I have
> read the 21 step, I simply paste in the linearization values, right?...and reinstall to have a
> linearized curve.

That's correct -- print without linearize values, measure, and the put
the values into curve.

>
> 3. I am still wondering what .quad files are?  They are in the Cone Ink areas.   Just curious.
> I would like to look at the text values, but have not found a way to view these curves.

The .quad files are the actual curves.  They are just a big table of
numbers.  Use TextEdit if
you really want to see what's there.  Better is the new program called
QTR-CurveView.
Drop a bunch of .quad files onto it and it shows nice graphs of the inks.

Roy

>
>
> Thanks for your patience with me.
>
> Phil
>

"pretty simple" huh? I have, with the help of folks in the Digital BW the
Print group, figured out how to make grayscale .icc profiles for use with
the Epson 3800 driver (Windows Vista), but I haven't figured out how to
"print without linearize values, measure, and put the values into a curve."
And that's why I'm not using QTR..

 

Anyone have a "for dummies" step by step explanation, pretty please with
sugar on top?

 

Much thanks in advance,

 

  - susan

 

>
> 2. My assumption on linearization is that if I make some changes, ie add a
toner curve,
> that I need to print the modified curve/step wedge without any
linearization values in the
> installed text file. Is this correct? I assume I would not get a correct
linearization if the
> modified curve has any linearization text. Please confirm that I am
correct. Once I have
> read the 21 step, I simply paste in the linearization values, right?...and
reinstall to have a
> linearized curve.

That's correct -- print without linearize values, measure, and the put
the values into curve.





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

> but I haven't figured out how to
> "print without linearize values, measure, and 
> put the values into a curve."
> And that's why I'm not using QTR..
> 

That assumes that you have followed the first couple of steps in 
setting up a new curve.  Basically you should have already done the 
following:

1) Print out the the step chart for all of the inks, with no ink 
limit set, using the QTR "Calibration Mode." Then,

2) Determine the black ink limit from that chart.  Set the limit in 
QTR. Print out the step charts a second time with that limit set.

3) Go to "Curve Creation" in QTR. Use the 1st and 2nd chart to 
determine the "Density" and "Limit" for each of the other black/grey 
inks.  Save the curve set that you are working on.  (Paul Roark has 
some pretty good instructions for all of this on his web site.)


Once you are at that point, go into QTR and open up the 21 Step 
chart.  

4) Select the curve set that you are working on.  Set the dpi, etc. 
that you want to use later for your prints.  Print out the 21 Step 
chart with that curve.

5) Open the Gretag ProfileMaker software. Use the "Measure" function. 
Use a spectrophotometer to read the "L" values of each patch on the 
21 step chart that you just printed.

6) Export the "L" values to a text file.  Open the text file, cut and 
paste the values to Excel.  Copy just the "L" values from Exzcel.

7) Paste the L values on the "Linerization" tab of teh curve that you 
are making in QTR.  Save the file.  Generate the curves.


A few points. If you have gotten to the stage of making your own ICC 
profile using the QTR "Create ICC" tool, you should already be 
through step #6?  You should have created a text file from the 21 
step chart that you dragged to the "Create ICC" tool?  Then it is 
just a matter of entering those values in the curve that you are 
creating in QTR.

One question though: what papers are you using? Which inks? Aren't 
there existng profiles in QTR for the 3800 and a paper close to teh 
one tht you are using? The curves don't vary a lot from paper to 
paper.

Hope that helps.  No time to re-read - hope it all makes sense 
too. ;>) Post questions. Good luck!

Best,
Michael

Thanks for this clear - as far as it goes - explanation. And despite
what follows, I am really truly grateful for your answer to my question.

Why do I say "as far as it goes"? I think what I am missing is a good
overview and under-the-hood-view of the subject, something I would get
in a published book (I don't know of any) or an advanced course on
digital black and white printing (not offered in my community
college). I'm sure with this under-the-hood knowledge all would come
clear.

I know in color printing what I am accomplishing when I create a .icc
profile: I'm recording the response of my
printer/ink/paper/resolution/etc. combination to a request to print a
certain RGB value and making that information available to Photoshop
so it can compensate when it sends a file to the printer. 

Now, what am I accomplishing when I create an Epson ABW profile?
Presumably the same thing, except I am measuring shades of gray
instead of colors.

That's overview. Under the hood, I know the spectrophotometer measures
the RGB values actually printed. For color, the Eye-one software
somehow magically turns those numbers into a .icc profile. For
grayscale, the MeasureTool software somehow magically creates a
mysterious text file full of mysterious numbers, which QTR CreateICC
magically turns into a .icc profile. Photoshop reads the .icc profile
and adjusts the numbers sent to the printer so the printer prints the
colors or shades of gray Photoshop expects the original numbers to
represent.

Now we come to QTR curves/profiles, non-Epson ink sets, mysterious
terms like "linearization," "ink limit," "density," "L values," etc.
And unclear actions such as entering values into a curve in QTR. All
the information I can find on the Web and in news groups and in QTR
documentation appears to take for granted that the reader has an
overview and under-the-hood-view of what QTR is doing. I don't; that's
why I asked for a "For dummies" explanation. 

I know I'm asking too much for any individual to give me a semester's
tutorial for free -- but can anyone point me to a book or on-line
class or, I suppose, tutorial video, or even an in-person workshop
that explains this material, with particular reference to QTR and the
terms used? 

 - confused susan trying to figure it all out

--- In QuadtoneRIP@yahoogroups.com, "Michael T. Murphy"
<uriel_bear@...> wrote:

>
> > but I haven't figured out how to
> > "print without linearize values, measure, and 
> > put the values into a curve."
> > And that's why I'm not using QTR..
> > 
> 
> That assumes that you have followed the first couple of steps in 
> setting up a new curve.  Basically you should have already done the 
> following:
> 
> 1) Print out the the step chart for all of the inks, with no ink 
> limit set, using the QTR "Calibration Mode." Then,
> 
> 2) Determine the black ink limit from that chart.  Set the limit in 
> QTR. Print out the step charts a second time with that limit set.
> 
> 3) Go to "Curve Creation" in QTR. Use the 1st and 2nd chart to 
> determine the "Density" and "Limit" for each of the other black/grey 
> inks.  Save the curve set that you are working on.  (Paul Roark has 
> some pretty good instructions for all of this on his web site.)
> 
> 
> Once you are at that point, go into QTR and open up the 21 Step 
> chart.  
> 
> 4) Select the curve set that you are working on.  Set the dpi, etc. 
> that you want to use later for your prints.  Print out the 21 Step 
> chart with that curve.
> 
> 5) Open the Gretag ProfileMaker software. Use the "Measure" function. 
> Use a spectrophotometer to read the "L" values of each patch on the 
> 21 step chart that you just printed.
> 
> 6) Export the "L" values to a text file.  Open the text file, cut and 
> paste the values to Excel.  Copy just the "L" values from Exzcel.
> 
> 7) Paste the L values on the "Linerization" tab of teh curve that you 
> are making in QTR.  Save the file.  Generate the curves.
> 
> 
> A few points. If you have gotten to the stage of making your own ICC 
> profile using the QTR "Create ICC" tool, you should already be 
> through step #6?  You should have created a text file from the 21 
> step chart that you dragged to the "Create ICC" tool?  Then it is 
> just a matter of entering those values in the curve that you are 
> creating in QTR.
> 
> One question though: what papers are you using? Which inks? Aren't 
> there existng profiles in QTR for the 3800 and a paper close to teh 
> one tht you are using? The curves don't vary a lot from paper to 
> paper.
> 
> Hope that helps.  No time to re-read - hope it all makes sense 
> too. ;>) Post questions. Good luck!
> 
> Best,
> Michael
>

Not a lot of time, but here is a quick response.

What you are setting in QTR is at a more "detailed" level than an ICC 
profile. You atre really setting up a generic "driver" the program 
taht puts the ink on the paper. 

Imagine you work for Epson. They ask you to write a driver to make 
prints on their next printer. Grey inks only - no color - to make it 
a little easier for your first driver. ;>)

Your first question: How many inks? 

That corresponds to the number of inks that you set in QTR as "Grey 
ink." You might tell it, for example, that the Black (K), Light Black 
(LK), and Cyan (C) positions have an ink cartridge in them.  The M, 
LC, LM, and Y positions have no ink (or a color ink) so you leave 
those unused.

The next question: "Which ink is darkest?"  "By how much?"

The "L" value ranges from 0 to 100. 0 is the darkest, 100 is the 
lightest.

You use the "Density" value to enter the "L" value that you read from 
a patch of ink to tell QTR how dark an ink is. First you set the 
value of the darkest ink, K.  

The values you enter for the other inks are relative, in terms of the 
darkest ink - usually K.  They tell QTR how much "lighter" those inks 
are than K.

So you enter the denisty of "LK" as a 40, that tells QTR that the 
darkest that LK gets in the printout is the same as K at the 40 
level, a litte less than halfway to black on K's "white" to "black" 
scale. 

And then you enter a 13 for C. That says that C is lighter than K or 
LK, that the blackest it gets in the test printout is equal to K at 
the 13 level.  

So C is more useful for highlights than shadows. LK will be useful 
for midtones. And when you need a dark black, K will be used for the 
shadows.  

Next question: "How do I know when to stop adding more ink?"

The answer: at the point of diminishing returns for any given 
dilution.  That is, as you put more ink on the paper, the patch 
usually becomes darker - which is good. But there comes a point where 
you keep adding ink, anbd the patch becomes just barely darker, or 
actually becomes a little lighter. That is the "Limit", the point at 
which you stop adding ink (and probably switch to your next darker 
ink.)  If you didn't have a limit, you could keep addinmg ink - 
trying to get a draker black - until you had puddles on the paper 
taht smeared and created a holy mess.

OK, you enter all of that. QTR makes a print, but there are jumps and 
gaps in the values it outputs, especially when it switches from one 
ink to the next. So you "linerize" it - make the output appear 
smooth, or continous, to the eye, so that there aren't weird things 
going on when it switches from the C to LK ink - like suddenly 
getting darker by 15 units, instead of 1 or 2 "L" values.

I hope that helps.   More questions?

Best,
michael

Hello Susan. One of the things you might want to try is to simply type
the specific terms you don't understand into the search box above the
group messages. Do this one at a time and then follow and read all the
postings related to it from the archives of the group. You're likely
to spend some time doing this but you will probably get a lot more
information than you expected. It's a veritable treasure trove but,
like all treasures, you have to dig for it ;)

Regards.
Duane

--- In QuadtoneRIP@yahoogroups.com, "slcphoto73" <slchapin@...> wrote:

>

> 
> I know I'm asking too much for any individual to give me a semester's
> tutorial for free -- but can anyone point me to a book or on-line
> class or, I suppose, tutorial video, or even an in-person workshop
> that explains this material, with particular reference to QTR and the
> terms used? 
>

That is helpful! Thank you. I'll be studying your message, and following up
on the resources suggested by a couple of other people. If I still can't
figure it out on my own I might try to get to a Piezography workshop at Cone
in Vermont, which was one suggestion.

 

 - grateful susan



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