QTR-Quadtone RIP

I have made several curves for other papers and this paper is giving me a  
problem that has not occurred before. I am running an Epson 2200 with K7  
inks. I printed the ink pattern page from the calibration mode with the  
ink calibration slider set to 100 (all the way to the right.) I looked at  
the steps and determined that I was getting the max black at about 45 and  
that above that there was no increase in black level. There was, however,  
no decrease either. I reprinted the page with the ink level setting at 45  
and then used that to determine the densities for the 7 channels and then  
printed the 21step.tif file. The last two steps at 95 and 100 are lighter  
than the 90 step and they are lighter than what the blackest ink was  
laying down in the ink pattern page. When printing the ink pattern, it  
would stay completely black beyond 45 all the way up to 100, so I don't  
think the 95 and 100 steps of the 21 step strip are lighter due to over  
inking of the blackest ink. When I look at the curves, I see that the next  
to darkest channel, Lt Black, has an overlap with the darkest black all  
the way up to 100. Actually, at a binary value of 255 it is at 0 but at  
254 an under, there is some Lt. Black being layed down along with the  
Black. I believe that is what is causeing my last two steps to be lighter  
than what they should be. How do I control the overlap? Any value entered  
in the overlap setting inceases the overlap. I need to decrease it. I have  
tried all kinds of variations, and I can't figure out a way to reduce the  
Lt. Black overlap with Black. I could go into the .quad file and edit the  
values directly, but there must be a more elegant solution that I am not  
seeing. Can anyone tell me what I am doing wrong and how to get the deeper  
blacks that I see on the ink patter page printouts.

Thanks,
Steve


-- 
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/

Well, I haven't gotten any replies on this so I'll just let you know what  
I did. I loaded the .quad file data into Excel, plotted the curves, and  
then plotted the resultant overall curve based on my ink density factors.  
Then I modified the LK  and K curves to reduce the LK and have it go to 0  
sooner and increased the K ink to compensate. Then I modified the .quad  
file with my new values and ran another 21step test strip print. After a  
couple of iterations, I got it looking quite nice. I was kind of flying in  
the dark but it wasn't as hard to do as I thought it might be.
Steve


On Wed, 16 Apr 2008 14:35:48 -0700, Steve and Ann Taylor  
<taylorfamily3688@...> wrote:

> I have made several curves for other papers and this paper is giving me a
> problem that has not occurred before. I am running an Epson 2200 with K7
> inks. I printed the ink pattern page from the calibration mode with the
> ink calibration slider set to 100 (all the way to the right.) I looked at
> the steps and determined that I was getting the max black at about 45 and
> that above that there was no increase in black level. There was, however,
> no decrease either. I reprinted the page with the ink level setting at 45
> and then used that to determine the densities for the 7 channels and then
> printed the 21step.tif file. The last two steps at 95 and 100 are lighter
> than the 90 step and they are lighter than what the blackest ink was
> laying down in the ink pattern page. When printing the ink pattern, it
> would stay completely black beyond 45 all the way up to 100, so I don't
> think the 95 and 100 steps of the 21 step strip are lighter due to over
> inking of the blackest ink. When I look at the curves, I see that the  
> next to darkest channel, Lt Black, has an overlap with the darkest black  
> all
> the way up to 100. Actually, at a binary value of 255 it is at 0 but at
> 254 an under, there is some Lt. Black being layed down along with the
> Black. I believe that is what is causeing my last two steps to be lighter
> than what they should be. How do I control the overlap? Any value entered
> in the overlap setting inceases the overlap. I need to decrease it. I  
> have tried all kinds of variations, and I can't figure out a way to  
> reduce the
> Lt. Black overlap with Black. I could go into the .quad file and edit the
> values directly, but there must be a more elegant solution that I am not
> seeing. Can anyone tell me what I am doing wrong and how to get the  
> deeper blacks that I see on the ink patter page printouts.
>
> Thanks,
> Steve
>
>



-- 
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/

Dear  Steve,

I just ran into QTR;- I have a 4800 pro epson printer branched on a mac pro
intel computer. So far I was not able to print nice BW prints on Hahnemule
rag brite papers. They look gray or are to dark. Detail is missing. On
glossy paper everything i fine in the black prints.

This weekend I started to use QTR. BW prints look a lot better though in the
deep shadow parts the blacks are far to saturated and look ugly greenisch.

I do not understand the system of calibrating. I am a gratuated enigeneer
and used to math. But the way things are explained is not lolgical for me.

I am curius wat steps you take to calibrate. Why are there no sliders to
ajust all the ink separately.

After calibration mesurement how can I tell the printer what to do.

Greetings Walther Walraven


Op 21-04-2008 19:57, Steve and Ann Taylor <taylorfamily3688@...>
schreef:

> Well, I haven't gotten any replies on this so I'll just let you know what
> I did. I loaded the .quad file data into Excel, plotted the curves, and
> then plotted the resultant overall curve based on my ink density factors.
> Then I modified the LK  and K curves to reduce the LK and have it go to 0
> sooner and increased the K ink to compensate. Then I modified the .quad
> file with my new values and ran another 21step test strip print. After a
> couple of iterations, I got it looking quite nice. I was kind of flying in
> the dark but it wasn't as hard to do as I thought it might be.
> Steve
> 
> 
> On Wed, 16 Apr 2008 14:35:48 -0700, Steve and Ann Taylor
> <taylorfamily3688@...> wrote:
> 
>> I have made several curves for other papers and this paper is giving me a
>> problem that has not occurred before. I am running an Epson 2200 with K7
>> inks. I printed the ink pattern page from the calibration mode with the
>> ink calibration slider set to 100 (all the way to the right.) I looked at
>> the steps and determined that I was getting the max black at about 45 and
>> that above that there was no increase in black level. There was, however,
>> no decrease either. I reprinted the page with the ink level setting at 45
>> and then used that to determine the densities for the 7 channels and then
>> printed the 21step.tif file. The last two steps at 95 and 100 are lighter
>> than the 90 step and they are lighter than what the blackest ink was
>> laying down in the ink pattern page. When printing the ink pattern, it
>> would stay completely black beyond 45 all the way up to 100, so I don't
>> think the 95 and 100 steps of the 21 step strip are lighter due to over
>> inking of the blackest ink. When I look at the curves, I see that the
>> next to darkest channel, Lt Black, has an overlap with the darkest black
>> all
>> the way up to 100. Actually, at a binary value of 255 it is at 0 but at
>> 254 an under, there is some Lt. Black being layed down along with the
>> Black. I believe that is what is causeing my last two steps to be lighter
>> than what they should be. How do I control the overlap? Any value entered
>> in the overlap setting inceases the overlap. I need to decrease it. I
>> have tried all kinds of variations, and I can't figure out a way to
>> reduce the
>> Lt. Black overlap with Black. I could go into the .quad file and edit the
>> values directly, but there must be a more elegant solution that I am not
>> seeing. Can anyone tell me what I am doing wrong and how to get the
>> deeper blacks that I see on the ink patter page printouts.
>> 
>> Thanks,
>> Steve
>> 
>> 
> 
>

Hi Walther,

I am not an expert with the QTR program but I can give you a quick  
explanation of what I think is happening. The underlying function for QTR  
is that of the .quad file. It divides the black levels printed onto the  
paper into 256 levels corresponding to an 8 bit integer. Each possible ink  
output is divided into 65536 levels corresponding to a 16 bit integer. At  
any given 8 bit input level, QTR references the .quad file to determine  
which inks and how much of each ink to dispense. There may be multiple  
inks being used because of the overlap in the transitions between inks and  
possible use of color inks to set the tone. How many inks there are is  
determined by the printer and ink set you are using. It is also further  
complicated by the fact that there are multiple ink dots printed for each  
actual image pixel and there is also dithering used to determine the how  
the dots get printed.

The .quad file is generated by the curve creation part of QTR. It uses  
some algorithm, which I don't have any information about, to derive the  
individual ink values for each 8 bit level from the information in the  
.qidf file.  This information is set during the calibration process by  
selecting the type of printer, the ink set being used, and then printing  
the ink pattern page, setting the ink limits and ink density values, and  
then printing the 21step file to determine the linearization levels. You  
can read the black levels from the test prints with a special instrument  
or with a scanner and photo editing program. If you are using a color ink  
set, you have to determine how much of the color inks you have to add to  
the basic black ink(s) to set the desired output tone, cool, neutral,  
warm, etc. All of this has to be done for each paper, printer, and ink set  
combination that you use since all these factors affect the finished print.

Jon Cone supplies his own .quad files for some paper/printer combinations  
for his K7 and other related ink sets. Some people also use the ICC curve  
generation capabilities in QTR, but I have not so I can't tell you much  
about them.

QTR is a great program at a great price and everything works pretty well  
and you have quite a bit of flexibility and control using it. It does not  
go so far as having sliders for each ink for you to play with and as I  
found, the built-in algorithms don't seem to work with some papers. Maybe  
there is something else that I am missing but that is my understanding at  
the moment. Hope this helps you get a better feel for what is going on.
Steve



On Mon, 21 Apr 2008 12:34:12 -0700, W.L. Walraven  
<walraveninnovation@...> wrote:

> Dear  Steve,
>
> I just ran into QTR;- I have a 4800 pro epson printer branched on a mac  
> pro
> intel computer. So far I was not able to print nice BW prints on  
> Hahnemule
> rag brite papers. They look gray or are to dark. Detail is missing. On
> glossy paper everything i fine in the black prints.
>
> This weekend I started to use QTR. BW prints look a lot better though in  
> the
> deep shadow parts the blacks are far to saturated and look ugly  
> greenisch.
>
> I do not understand the system of calibrating. I am a gratuated enigeneer
> and used to math. But the way things are explained is not lolgical for  
> me.
>
> I am curius wat steps you take to calibrate. Why are there no sliders to
> ajust all the ink separately.
>
> After calibration mesurement how can I tell the printer what to do.
>
> Greetings Walther Walraven
>
>
>

-- 
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/

Thanks a lot Steve,

Tomorow I am gooing to start with it.

Op 22 apr 2008, om 02:57 heeft Steve and Ann Taylor het volgende  
geschreven:

> Hi Walther,
>
> I am not an expert with the QTR program but I can give you a quick
> explanation of what I think is happening. The underlying function  
> for QTR
> is that of the .quad file. It divides the black levels printed onto  
> the
> paper into 256 levels corresponding to an 8 bit integer. Each  
> possible ink
> output is divided into 65536 levels corresponding to a 16 bit  
> integer. At
> any given 8 bit input level, QTR references the .quad file to  
> determine
> which inks and how much of each ink to dispense. There may be multiple
> inks being used because of the overlap in the transitions between  
> inks and
> possible use of color inks to set the tone. How many inks there are is
> determined by the printer and ink set you are using. It is also  
> further
> complicated by the fact that there are multiple ink dots printed for  
> each
> actual image pixel and there is also dithering used to determine the  
> how
> the dots get printed.
>
> The .quad file is generated by the curve creation part of QTR. It uses
> some algorithm, which I don't have any information about, to derive  
> the
> individual ink values for each 8 bit level from the information in the
> .qidf file.  This information is set during the calibration process by
> selecting the type of printer, the ink set being used, and then  
> printing
> the ink pattern page, setting the ink limits and ink density values,  
> and
> then printing the 21step file to determine the linearization levels.  
> You
> can read the black levels from the test prints with a special  
> instrument
> or with a scanner and photo editing program. If you are using a  
> color ink
> set, you have to determine how much of the color inks you have to  
> add to
> the basic black ink(s) to set the desired output tone, cool, neutral,
> warm, etc. All of this has to be done for each paper, printer, and  
> ink set
> combination that you use since all these factors affect the finished  
> print.
>
> Jon Cone supplies his own .quad files for some paper/printer  
> combinations
> for his K7 and other related ink sets. Some people also use the ICC  
> curve
> generation capabilities in QTR, but I have not so I can't tell you  
> much
> about them.
>
> QTR is a great program at a great price and everything works pretty  
> well
> and you have quite a bit of flexibility and control using it. It  
> does not
> go so far as having sliders for each ink for you to play with and as I
> found, the built-in algorithms don't seem to work with some papers.  
> Maybe
> there is something else that I am missing but that is my  
> understanding at
> the moment. Hope this helps you get a better feel for what is going  
> on.
> Steve
>
>
>
> On Mon, 21 Apr 2008 12:34:12 -0700, W.L. Walraven
> <walraveninnovation@...> wrote:
>
>> Dear  Steve,
>>
>> I just ran into QTR;- I have a 4800 pro epson printer branched on a  
>> mac
>> pro
>> intel computer. So far I was not able to print nice BW prints on
>> Hahnemule
>> rag brite papers. They look gray or are to dark. Detail is missing.  
>> On
>> glossy paper everything i fine in the black prints.
>>
>> This weekend I started to use QTR. BW prints look a lot better  
>> though in
>> the
>> deep shadow parts the blacks are far to saturated and look ugly
>> greenisch.
>>
>> I do not understand the system of calibrating. I am a gratuated  
>> enigeneer
>> and used to math. But the way things are explained is not lolgical  
>> for
>> me.
>>
>> I am curius wat steps you take to calibrate. Why are there no  
>> sliders to
>> ajust all the ink separately.
>>
>> After calibration mesurement how can I tell the printer what to do.
>>
>> Greetings Walther Walraven
>>
>>
>>
>
> -- 
> Using M2, Opera's revolutionary e-mail client: http://www.opera.com/ 
> m2/
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>

Met vriendelijke groet,

Ir. Walther L. Walraven

Walraven Innovation Management & Photography

T.P. Oosterhoffstraat 1,
9883 PJ  OLDEHOVE
Netherlands


(M) +31 	(0) 653 122 571
(T) +31 	(0) 594 5970 76
(F) +31 	(0) 594 5970 75
(Email) 	walraveninnovation@...