GMG Flexoproof Iterations

[Cris]

I have been using GMG's Flexoproof for my company's Epson 9900 proofer, have been experiencing some issues with the software and was wondering if anyone else has been having or had the same problem and found a solution or solutions.
What is going on is that during the profile set up after I have scanned in the press test target the manual states that it should only take 2 to 3 iterations of printing out and scanning the same target type to get the Epson 9900 in gamut with the press sample. On average I have been having to do 3 times that many iterations and a few times 4 - 8 times that many iterations. That's just getting the mx4 profile set up. It can take just as many iterations to get my mx5 profile set up for dot proofs. I do exactly as I was told to do profiles by the trainer, go to the letter of the PDF manuals and used the help section of the software as well. I was told by GMG support it shouldn't take that long and it may be a fluke. So I thought I would reach out to my peers on here and see if there are any suggestions or solutions. All my other requests for info and help on here have been very useful.
I will greatly appreciate any help because I hate using the excess ink and paper for a profile when according to the manual and GMG that I should only be doing 2 - 3 iterations at the most to get a profile in gamut with the press sample.
If it takes me a couple of days to reply to any questions for further info if anyone needs more clarification that replies to me or I don't say thanks it's because the area I live in is going to get God only knows how much more snow tonight and tomorrow. I've heard anywhere from 6 - 24 additional inches on top of the 26+ inches we got this past weekend. The state doesn't want anyone on the roads tomorrow except in the event of an emergency so we will be closed. I will attempt to answer from home barring power loss. But in advance, I appreciate any help.

 

[TerryWyse]

How are you judging whether it's "in-gamut" or not? What sort of dE values are you getting during the iteration process?

A few suggestions/comments:
* If the target values are coming directly from your press measurements, as opposed to using a standard data set such as GRACoL/SWOP, it could be a problem with your data. I would suggest you average as many press sheet samples as you have time for (5 or more) to help smooth the data. When you print your first calculation from the MX4, use a chart with a "visual" layout such as the ECI2002 Visual. Look at the patches to see if they are smooth. If the patches kind of jump around and look "chunky", your data may need smoothing. You can do this in GMG by selecting "Smooth Target Values" under the Tools menu (I think). Smoothness is perhaps more important than accuracy, especially if your data is not very linear.

* Use a delta e tolerance of "0" when doing your iterations to force GMG to calculate the entire MX4. To monitor my progression, in the "Stats" window I jot down the average dE and also the number of patches greater than 1, 2 and 3 dE. When I start to see that the number of patches greater than about 2 dE stop improving (or get worse) it's time to stop iterating even if the average dE is still improving.

* Try creating a new "limited" gamut file using a high patch count (>2,000 patches). Generally, more patches than the standard ECI chart will take fewer iterations to get to the same result.

If I had to point to one thing though, my guess would be that your custom press data is very non-linear (UNsmooth). If you're not able to get under about 1 dE in 2-3 iterations (using a good media of course), I would say it's your data. Good data such as GRACoL will get well under .5 dE under good conditions in only a couple of iterations.

Regards,
Terry

 

[meddington]

Could your proofing paper be playing a role? Less than optimal stock can be less accepting of ink , and lead to higher max values, yoyo-ing and in general cause GMG to have trouble nailing the profile in the first few iterations. If your max values are high, or bounce around, your paper might require a bit extra effort to dial in.

* Use a delta e tolerance of "0" when doing your iterations to force GMG to calculate the entire MX4. To monitor my progression, in the "Stats" window I jot down the average dE and also the number of patches greater than 1, 2 and 3 dE. When I start to see that the number of patches greater than about 2 dE stop improving (or get worse) it's time to stop iterating even if the average dE is still improving

Far be it from me to say anything contrary to Terry's advice...but I will anyway. its been my experience that repeatedly calculating the entire mx4 can sometimes cause yoyo-ing of measurements that would othewise remain unchanged with a higher tolerance, particularly with less than ideal paper. I typically start with 0 or 0.5 as the tolerance, then will raise the tolerance to 1 or 1.5 or even 2 or more after a couple iterations. I've seen patches bounce from 1-3 delta e to 4-8 delta e between iterations, and have sometimes had to manually set problematic patches so GMG wouldn't kick them higher. Again, this is usually with third party paper stock.

 

[TerryWyse]

Far (farther?) be it from me to disparage Sir Eddington's sage advice....but as long as a) you let the print dry for a consistent amount of time (10 min. is my rule) and b) you have a consistent/repeatable measurement device, i.e. DTP70 or iSis, it should be fine to use 0 dE. On the other hand, if I were using an EyeOne handheld or EyeOne iO table, I'd jack that dE tolerance up a bit. I don't find either of those instruments consistent enough to be used with something as precise as GMG...but that's just my opinion and I'm darn well entitled to it....as long as my wife says it's OK with her.

It's the DATA Mike, the DATA!



Regards,
Terry

 

[Cris]

Sorry took a a couple days to reply to everyone. I was snowbound and then could not get my car out of my driveway. Oh the joys of moving farther north and all the snow storms.

Terry - What I meant by In Gamut was that I had the software set to 0 delta E and according to the manual the max delta E should be below 3. I have been trying to follow the instructions to the letter as I said in my previous post. My data has been coming from my press sheets. I have been averaging the data of 2 - 3 press sheets but I will up the average to 5 Terry to see if that helps.
I continued trying to get the delta E lower even when the 2 or 3 delta Es started yo-yo-ing, getting better, same, then worse and then better again.
What do you mean by creating a limited gamut file? I was told that it had to be the ECI charts GMG supplied and nothing else for profiling. Do you mean there is a way to set up my own chart?
You are right. My data sets have been pretty unsmooth. I need to start smoothing the data to possibly make it faster and easier.
I am using and i1 and i0 table. That is what our sales rep suggested that we get because I am also reading in corrugated targets. I have found a difference in readings between delta E, color readings, dot gains and densities with the i1 and our X-Rite 530.
I do let the print dry for a minimum of 10 minutes. Sometimes I will let it dry for 15 minutes.

Eddington - The stock on the samples I have been reading in has been a corrugated stock that seems to absorb ink pretty well. That very well could be part of the problem. The proofing paper I am using is the GMG semimatte 250.
It has been a while since I have done the color work I am doing now and have been learning by studying theories again and working, working, working on the device. How would you suggest that I manually set the patches without throwing off the profile?

Everything that ya'll have given me so far has given me a lot to think about and been very helpful. I appreciate it a lot. I look forward to reading the answers to the other questions that I have asked.

 

[caveman]

In the GMG manual they describe a procedure called length feed adjustment (or something similar) which may solve similar problems to yours. It requires a tool which can be downloaded from the GMG site. I believe it worths a try.