White boxes instead of characters

[frailer]

Would someone mind refreshing me on this one. Yonks since I've seen it, white boxes being substituted. Colleague used supplied PDFs; Indy, used 'em before, all fonts embedded, all TTF.

Imposed in PREPS 5.1.2, distilled in watched folder...normal procedure.
Have a recollection it's to do with glyphs. I did change the Distiller Compatibility from Acro 4 to 6 recently, but that may have nothing to do with it. Got a hunch had I suggested it before they were dumped, that if the PostScript files had been Distilled via Acro 7 Pro Distiller on one of our other Macs, they may have been OK. Colleague's solution was to go back to source Indy doc +fonts/pics.. Fine, but curious anyway. Maybe get Acrobat on that machine up to 7.

[JanZ]

We have seen the same thing in our workflow. Customer supplied or in-house written Indy pdf impossed back into Indesign and sent to the Xitron rip. When we resend the same impossed file to the rip, without redoing the pdf, we get a different result. (Everything looks great or a different page has the squares instead of type.)

The only thing we can come up with is that every one of the jobs used to be sent to us as a PageMaker file...

Jan

[jimking]

99.9% of the time when I've run across this same problem it's Times. It seems it's a font substitution problem. For example if the job was created using TimesNewRomanPSMT and for some reason or another Times TT is embeded into the job instead, our little white boxes appear hear and there.
Thats been my experience.

[zox]

It's subset font issue where two subsets of the same font are in the same PDF and are named the same.
RIP does bad thing of using only one subset, hence the dropped characters.

I know it is known limitation of the PS workflow such as Brisque.
I also know that Prinergy has similar problem but they have script in place that checks these things and warn you about potential problem (some users find it annoying).

In real PDF workflow (without back conversion to PS) it should not happen and it should work ok.

I know our workaround on Brisque is to flatten it prior to ripping.

[frailer]

Will have time (will make time), to back-track the job next week. We got DynaStrip+dongle with our new RIP but not fully up to speed yet. Good experiment to drag it through Dyna as full PDF in/out, then no PostScript created, as happens in PREPS no matter what.
After these posts I'm thinking PREPS PostScripting may be the baddie.
Thanks for the input.

[local1]

Zox,

This same problem shows up from time to time for me also. I am using a route that involves postscript/preps for now.

What changes are made to the PDF when you "flatten" on your Brisque?
Would you be able to perform the same changes outside of your system, via acrobat/pitstop? Do you know what it does to the fonts to make them better behaved? Are you aware of what the Prinergy script does or if it works outside of Prinergy? Or does it just report the issue only?

Thanks for any input

[zox]

As far as I know Prinergy script just reports on potential issue, but I am not running Prinergy so I can not confirm 100%, we are Brisque shop.

As far as I know is that for one reason or another you can have two different subsets of the same font but with same name.
This should not be the case but apparently sometimes you get not unique ID of the subset.

Some rips (possibly Adobe) will consider these two subsets to be the same and will be use only first subset throughout the document.
In such case if you did not have all needed characters in second subset, they will be dropped or replaced with space.

This is in layman terms.
I am pretty sure there is more technical explanation related to PDF technology and how it handles subsets.

e.g.
In first subset you have characters I K E.
In second subset you have characters M I K E.

If the rip takes first subset to be your "main" font and you have word "Mike" in your PDF, it will be ripped as " ike" without M.

I hope I did not confuse you.
I am not 100% right if this is exactly how it happens, but something in that line.
I am not developer but prepress guy.

I imagine if you replace one of the subsets with another or in other words rename the subset, it will be fine.
Maybe replacing subset with actual font and re embedding it would do it, I am not sure.
Our Brisque fixes it if you use flattening so I didn't have to go further.

I hope I didn't confuse you and it helped a bit.

[local1]

Zox,

Thanks. That does help to clear it up for me. Thanks.