QTR-Quadtone RIP

On 27 Feb 2009, at 8:11 PM, Dileepa Wijesundera wrote:

> I am using measuring tool in the demo mode. I have struggled through
> this but cannot find in the documentation how to do this. Any
> suggestions?

I tried it myself (also in the demo mode) and couldn't get anywhere  
either.  I did notice that if I turned on the "Spectral" checkbox in  
the Configuration dialog, then selecting Spot Measurement at least  
showed a possibility of "Densitometric" mode (in addition to  
"Colormetric") -- however, it was grayed out, and nothing I did would  
enable it.  I wonder if this is a limit of the demo?

> Interesting to know about your software, is it due to come out soon.
> Anything to make the proces easier would be helpful.

My main goal is to automate the time-consuming process of analyzing  
the stepcharts to determine the relative ink levels.  I dislike  
depending on the (ancient) MeasureTool.  More importantly, I find the  
calibration process to be very manually-oriented and prone to errors.   
It seems a perfect thing to build into a program.  There's also a  
wonderful opportunity to visualize how QTR will print an image, and  
allow the printmaker to really tune their printing to their own tastes.

As of now, this is what I have:

	- A program that simply lets me measure patches on a printed chart,  
like a very simple version of MeasureTool.  My program communicates  
with the EyeOne, then writes the resulting Lab values out to a text  
file; I use it by measuring prints of the "inkseparation6.tif" chart  
that comes with QTR.  In this way, I get the range of Lab values of  
the inks I'm using.  I'd like to expand this to generate the charts  
themselves (for reliability purposes), as well as to read the charts  
in strip mode rather than single-patch mode.

	- A program that reads those text files containing the inks' Lab  
values, and performs the same process as outlined in the "QuadToneRIP  
2.5 Calibration" document.  The program generates a profile file with  
both the LIMIT_* and the GRAY_INK_*/GRAY_VAL_* values, according to  
the analyzed inkset.  (The cool thing about the algorithm I've come up  
with is that it doesn't care what inks are what, nor what order  
they're in.  That's great when you have a clogged K nozzle, like I do  
right now!)

	- A program that shows a particular grayscale image, and given a  
particular (.quad) profile, shows not only the curves for the inks in  
the profile (like QTR-CurveView already does), but also renders  
several additional versions of the image, one per ink, showing exactly  
what ink coverage would be used for that image.  It's not perfect --  
it doesn't account for things like dot gain -- but I've found it  
useful to visualize how a particular image is going to print using a  
particular profile.

Another idea I had was to take a final step-chart of a fully  
linearized profile, and produce a chart showing the color shift (that  
is, the a/b components of each step) as well as the overall density.

I have to warn you that all these programs are quite rough, and not  
ready for general use.  However, I would like to do the work to get  
them into a form that other people could use.  I should mention that  
all of these are designed to work on a Mac (running Leopard/10.5).   
Parts of the core analysis could be made to run on Windows (it's all  
programmed in Ruby), but I don't use Windows, so I have no plans for  
anything but Mac versions.

--John