PostScript Oddness

[rick6502]

http://home.swbell.net/burn3deg/ebay/BarCodeProblem/UPC%20A.73413621251.EPS
A little background. The company I work for produces labels for animal feed company. They have many different labels and we produce them on demand. The UPC bar codes are eps files created in a old os 9 bar code program. They never change so we have been using the same EPS files for years. We set each label 1 up in InDesign CS2, place the proper bar code in the file and export as a PDF. Then depending on what they order we import each PDF into a new InDesign file and print from that file to our RIP.
Now the problem. Suddenly, we have one set of labels that were unable to be scanned by one of the animal feeds distributors. The first image is what it came out as. The second it what happens now. The third is the actual EPS.
We have never had a problem before. I have made several test runs and I am unable to duplicate the problem. I think it was a PS fluke, that may never happen again. My boss is convinced it's something bigger. But I also would like to not have this happen again. Very stressfull. Has anyone seen anything like this before?

 

[Blue Velvet]

No, I haven't seen anything like that before.

I also deal with bar-codes often for our publications. The production of those bar-codes is done externally but they need certain information stripped out and a bit of tweaking before they go to page-layout, especially if it's a two-colour job. But they go into Quark as EPSs straight from Illustrator...

Just looking at the numerals shows some messing with the paths. Maybe it might be wise to look at your PDF creation settings to see whether the vector balance is right but that's just a wild guess...

I'm curious as to why you go through the intermediate PDF process.

Still, if you can't recreate the problem that can be a good sign. Maybe some form of flight-checking the original EPSs and intermediate PDFs might be wise for a while until if and when you can get to the bottom of this mystery.

 

[rick6502]

Just looking at the numerals shows some messing with the paths. Maybe it might be wise to look at your PDF creation settings to see whether the vector balance is right but that's just a wild guess...

I'm curious as to why you go through the intermediate PDF process.

My best guess at this point is the RIP received too much data and had to compromise the quality of the output. I have seen that on other printers that use PS clones. You send a page of tiff's with a lot of transparency effects to my HP and either the tiff's become very pixelated or it just blows the file off. This is Harlequin RIP so I expect better.

As to why we go through the multistep process: These labels have been setup over many years. They were originally laid out on a Compugraphic. They don't use a whole lot of different fonts. But they are several different colors used. Most are one or two color, but those two colors may not be the same. They all use different artwork. They all use different paper. And they are orientated both landscape and portriat. They also vary the quantities they order at a time. There are only about 3 sizes. We tired to print some in advance, but government regulations cause changes in the wording to be made fairly often. We also used to just print them directly from PageMaker. But when we combined two different labels on one plate we had to open both files, make sure all of the links are up to date, copy paste from one of the two files, sometimes one has to be rotated, sometimes the rotation screwed up the formatting, sometimes not, etc. This way when the order 1000 of two different ones, we don't have to make sure we have any fonts on the particular machine we are on, we just open a blank Indesign page, import the two PDFs we need and print. It makes some more work on the front end, but speed is often an issue with these labels, so all of the help on that end is good.