Profiling Targets

Bill W

Well-known member
Greetings,

I am a narrow web flexo printer and need some advice on profiling targets. I am presently using IDEALink and the G7 methodology to set up the proper curves for a number of our locations. Most of the presses I will be working with have a 10" wide web. I really like the ECI2002 random target and have printed some in the past experiments.

The only problem with this target is that it takes up most of my available web real estate, leaves almost no room for images and I can only read it in spot mode using my iO. Reading in spot mode takes almost 1 hour to read 1 target and I will be reading a minimum of 20 targets per press for each of the 8 presses I am trying to profile. I am trying to get all this done before the end of the year including 5 more trips to plants in Texas and Illinois to run and measure linear testing, run and measure G7 curve for accuracy and run the profile target.

The non random ECI2002 target is not an option because of how it formats on our web. Unlike the P2P target the ramps would be running across the web which means they can be dramatically affected by gear marking that we get in Flexo printing. A gear mark in a ramp would render it useless. I am able to mitigate gear mark problems with the P2P by running it with it ramps around the press, placing 2 of them staggered and by running a second run with the plates rotated 180 °. When I measure 10 sheets (20 P2P targets) I get a relatively smooth curve - at least something I can work with.

I am looking for suggestions of another target that is smaller but will give me enough values to make a good profile, or, a way to read the ECI2002 random target in strip mode.

I have checked out all the available targets in MeasureTool and I do not think any will work. I need my target to be on one set of plates - I have found by experience that splitting a target over multiple plants is a sure path to insanity.

Thanks to all that respond.

-Bill-
 
Re: Profiling Targets

Lets see....8 times 20 equal...Wow! Thats a lot of measurement to do in spot mode. What would be your ideal size for a test chart , and have you tried generating your own target with MeasureTool's Test Chart Generator?
 
Re: Profiling Targets

Bill

I would have to think that if you are traveling to multiple locations
and reading that many targets, it would be cost effective for your
company to buy a chart reader (like the X-Rite i1 iSis). My Isis can
read an EC12002 chart in about 10 minutes, and I don't have to baby-
sit it. I know your question was about a better chart, and perhaps
someone else will have a suggestion, but if you can save 50 minutes
on each of 160 readings, you are looking at over 130 man hours!

-Todd Shirley
 
Re: Profiling Targets

Greetings,

Michael - web width is 10.25 inches, of which I can use 10 inches including bearers. Our bearers take up about 1/4 inch. I would like to keep the repeat length under 12 inches. I have made a few attempts to make a custom chart, but am not sure if what I have made will give me enough numbers to make a good profile and, it seems all my attempts have made and "ordered" chart, not a random one which I am finding is really the way to go when profiling flexo presses, especially old ones.

Todd - yes on paper it looks like the purchase of the iSis would certainly be justified. Unfortunately the "paper look" does not always get upper management to open their purse. I purchased the iO because of the different thickness of stock we will eventually be profiling and I was told that readers like the iSis may not be of use with thicker stocks.

Is your iSis reading a random chart or an "ordered chart"? The random chart appears to not have well defined division lines between the color blocks and what little space there is, is often invaded upon by our press wander.

-Bill-
 
Re: Profiling Targets

>web width is 10.25 inches, of which I can use 10 inches including bearers. Our bearers take up about 1/4 inch. I would like to keep the repeat length under 12 inches. I have made a few attempts to make a custom chart, but am not sure if what I have made will give me enough numbers to make a good profile

Well, TestChart Generator will give you 1376 patches with a 12"x10" test chart formatted for the IO. Not quite the count of an ECI (about 1485), but pretty close. That should get you a decent profile I would think.

> it seems all my attempts have made and "ordered" chart, not a random one which I am finding is really the way to go when profiling flexo presses

Yep, you're right...can't believe there's no "scramble" button in the test chart generator. Seems odd. Anyway, you could save out the "ordered" reference file, open it up and randomize the order in excel, then create yet another new test chart from this reference file and save out new reference and tiff files. Seems to work ok here, but I don't have an IO, so if you try this, might want to output a proof and test measure it.

You can randomize your reference file in excel by doing the following:
1. open your "ordered" reference file in Excel
2. Copy and paste the CMYK values into another document
3. in an adjacent column, select all empty cells next to your data
4. type in "=RAND()" and then press Ctrl-Enter (this will randomly assign values to those selected cells)
5. select the top cell of your random values and sort by ascending (or descending...doesn't matter really)
6. Copy and paste the newly ordered CMYK values over your old values and re-save your text file (Text tab delimited).
7. use that new reference file to generate a new target based on the size you need.
 
Re: Profiling Targets

download xrites colorport 154

this will allow you to make custom targets with custom page size, from any standard target, eci, it874 etc, also supports direct import into monaco and profilemaker

Edited by: mike egloff on Oct 5, 2007 5:02 PM
 
Re: Profiling Targets

Greetings,

Mike - Thanks for reminding that I have ColorPort 1.5 on my computer. I had only played with a previous version, but it looks like 1.5 will help me in that I can make a randomized IT8.7.3 extended that will fit my format. Unfortunately any other larger target (more patches) goes to 2 or more pages.

When you say "supports direct import into ..." do you mean the measurement I created in ColorPort can be read by "....". I thought maybe I could create the target and export the reference file so I could just use MeasureTool to measure the target, but I can find no way to export the reference file and it also does not appear in the ColorPort folder. Does it exist somewhere, can I export it? Or, just be satisfied to measure in ColorPort and open in MeasureTool.

Michael - thanks for the excel education. Unfortunately my dongle is at work and I am at home. Maybe I will pick it up on the way to the airport on Monday and try your trick.

Happy weekend to all.

-Bill-
 
Re: Profiling Targets

>When you say "supports direct import into ..." do you mean the measurement I created in ColorPort can be read by "....". I thought maybe I could create the target and export the reference file so I could just use MeasureTool to measure the target, but I can find no way to export the reference file and it also does not appear in the ColorPort folder. Does it exist somewhere, can I export it? Or, just be satisfied to measure in ColorPort and open in MeasureTool.

Colorport generates xml for the reference file that isn't compatable with ProfileMaker. So you have to measure in Colorport, then export ProfileMaker compatable data for profile creation. Your also a bit limited in utilizing the maximum space available of your chart. If you end up using Colorport, test it out first. I have it installed on 4 computers here and only one will run it without serious bugs (one yeilds totally currupt measurement data). Others run it without issue though.
 
Re: Profiling Targets

Sorry the direct import only works with Monaco, If you use one of the standard targets , ECI 2002, it873, it874, read the target with color port save the data as profilemaker compatable , launch profilemaker spec the reference data and the newly created measured data.
 
Re: Profiling Targets

Greetings Michael,

I was able to try your "randomize" trick while flying to Dallas and it worked great. In order to save some room for images the max number of patches turned out to be 992. How does one know if there enough? The highlights appear to be covered which is important to us in flexo.

Thanks also for you info regarding Colorport. I really prefer to stay in MeasureTool for the entire process so I doubt that I will be doing much with Cololport.

-Bill-
 
Re: Profiling Targets

>In order to save some room for images the max number of patches turned out to be 992. How does one know if there enough? The highlights appear to be covered which is important to us in flexo.

How many patchs is enough depends on the stability of the device. The general rule is the closer the device is to being linear and gray balanced, the less patches you'll need, and the less work your profile will have to do. With a stable device, more patches can be needless, or even detrimental and introduce noise/banding. Since you're dealing with a fairly variable process with flexo, undergoing calibration toward G7 will help, as will the amount of averaging you'll be doing. I'm not really in a position to say how stable the devices are or how effective the process control is a keeping things in check and you'll have to use your own judgement there. Errors will likely be more noticable in highlight areas, and yes, a target generated through MeasureTools Test Chart Generator should have much better representation of highlight data than that derived from an IT87/3, and uses fractional percentages rather than whole numbers (less calc. errors).
 
Re: Profiling Targets

Anyone care to chime in on the importance of having an individual gradient ramp for C, M, Y, and K included in a profile press target.

I used the Excel randomize trick Michael pointed out and will be doing a quick test between my 992 patch target and the ECI2002 1485 patch target next week on press to see if the 992 target will be useable in my final profile run. Suggestions on how to evaluate the two would be greatly appreciated.

I noticed that even though the ECI target does not have ramps that run the total tint from highlite to 100%, it does have abbreviated ramps. There are no ramps on my 992 target.

Thanks to all that answer.

-Bill-
 
Re: Profiling Targets

Juergen Roesch of GMG UK sent me a 'GMG_CMYK_smoothCheck_V1.tif' and 'GMG_RGB_smoothCheck_V1_02.tif' to check the smoothness of my profile. Very good test that will quickly show problems with smoothness in the profile. You can try contacting him.

Don
 
Re: Profiling Targets

Greetings Don,

Juergen was very helpful with information we shared a few months ago regarding the unique challenges of profiling a flexo press. Juergen, if you are lurking, thanks again.

He supplied me with some files but I do not believe the files you listed were included. I will contact him to see if he is willing to share a few more files.

-Bill-
 
Re: Profiling Targets

Looks like my last post was blank for some reason....

>I noticed that even though the ECI target does not have ramps that run the total tint from highlite to 100%, it does have abbreviated ramps. There are no ramps on my 992 target.

Are you sure there are no pure ramps in your target? When I create a 992 patch target in Gretag's Target Gernerator, I get individual CMYK ramps in 10% increments.

>Suggestions on how to evaluate the two would be greatly appreciated.

The smoothcheck file is a good suggestion. You could also do objective round trip tests and the like. If you have ColorThink Pro, this is a lot more fun.
 

PressWise

A 30-day Fix for Managed Chaos

As any print professional knows, printing can be managed chaos. Software that solves multiple problems and provides measurable and monetizable value has a direct impact on the bottom-line.

“We reduced order entry costs by about 40%.” Significant savings in a shop that turns about 500 jobs a month.


Learn how…….

   
Back
Top