PDF-X1a is dropping some text. Why?

[guyfawkes]

The client has a layer file that is about 60 megs.
The EPS file is 135 megs due to the many layers of text.

When I made a PDF-X1a two lines of text disapeared
from different paragraphs.

Is this due to the many layers of text saved as Vector Data?
Is there info on what the limit is?
Does Adobe warn against the extreme use of their product?
How much Vector Data is too much?

Thanks,

Guy Fawkes

[lokki]

Welcome to PsT , guyfawkes - make yourself at home

I'm not sure exactly if there are stated limits to the amount of text you can deal with, but is it worth considering using another application, like PageMaker or Quark?

Also, have you tried importing your PDF into Acrobat Pro and doing a preflight on it? That may give you some clues as to why text is dropping out. Another thing to check might be the validity of your font files, and see if the client properly embedded the data.

135M is just rediculously large for text within a graphics app... way too much shape information.

[guyfawkes]

Thank you Lokki,

Yes, using text in Quark or Illustator is much better.
(no one uses Pagemaker)

The text that dropped off was the same font and
within the same sentence and on the same layer
as the text that printed.

I would love to comfirm that it IS the result of too much
Vector Data and understand why so I can tell the client why.

33 layers of text is too much. Right? But why?
And why isn't there a warning lable on the software box?

:)

:)

[lokki]

- we still use RageMaker quite a bit at work (not me personally... after I found I had nothing to actually strangle while working with it...), but I wholeheartedly understand why it's not popular ;}

I'll check around, but I think most of the problem has to do with how PS handles memory. We know it's a resource hog, and handling *any* huge document may cause problems. In the case of text/vectors, it can get even worse because it's not a native vector app.

33 layers of text, says you? xx|

Oh, one other thing - have you tried to open and close the file since then, with the same data missing, or does it randomly drop something each time, with previously dropped stuff coming back? Or (better yet), when it drops the text, does it leave the space there, or just close the gap?

[lokki]

OK, here's the first thing I found at the Adobe user forum:


Topic
Text is cut off or disappear in large images
Issue

Text is cut off or does not appear on a type layer in large images (e.g.,
400 MB) in Adobe Photoshop 5.x>.

Solutions:
Do one of the following:

Solution 1

Change anti-aliasing in the type layer:

1 Double-click the type layer.
2. In Photoshop 5.5>, select None, Crisp, or Strong from the Anti-Alias
pop-up menu in the Type Tool dialog box. In Photoshop 5.0.2, deselect
Anti-Aliased in the Type Tool dialog box.
3. Click OK.

Solution 2

Move the type layer, open the Type Tool dialog box, click OK, and then move
the layer back:

1. Select the move tool.
2. Move the type layer to the left until all the text is off the image.
3. Double-click the type layer.
4. Click OK in the Type Tool dialog box.
5. Move the type layer back with the move tool. Do not move it back using
the History palette.

Solution 3

Create the text using the type mask tool and fill the selection:

1. Select the type mask tool and click in your image.
2. Enter your text in the Type Tool dialog box. Click OK.
3. Move the type selection to the desired location. If you can see part of
the text on the existing type layer, you can use it as a guide to position
the type selection.
4. Select the layer on which you want the text to appear.
5. Select a foreground color for the type.
6. Choose Edit > Fill.
7. Chose Select > Deselect.
8. Remove the original type layer.

Solution 4

Create the text in Photoshop 5.0 or earlier, or in another application.
Then, import the text into Photoshop 5.x and remove the original type layer.

Additional Information

The type tool cannot create anti-aliased text larger than 8192 pixels on the
image canvas in Photoshop 5.x, although it can in earlier versions. The type
mask tool, however, can create selections that exceed this limit. You can
fill type mask selections to create pixels in the shape of type, but these
pixels are not editable in the Type Tool dialog box and do not appear on a
new Type layer.


This brings up additional questions I should have asked the first time... What platform, version of PS, and type of machine (speed/memory) are you running? What kind of project is it - single or multipage, colourspace, etc...

I'd look at using the type mask tool solution. I'll put up other stuff if I can find it.

[guyfawkes]

The Postscript is always fine. It does take long to RIP (of course)
than files put together properly.

But each time the file is Distilled the same text drops.
I don't know if the EPS was saved again. I'll have to try that.

We opened the bad PDF in Illustrator and viewed it in outline
mode. We can see a box around where the missing text is suposed to be.
But no text outline to work with.

Adobe has had some success in pushing InDesign onto
schools and ad agencies and now us since we have to
have everything. We like it for creative work. But it is a
replacement to Phtoshop Vector Data work and or Illustrator.

Not a replacement to QUARK as they are pushing it.

:)

[lokki]

Are you sure there were no special characters or formatting on the text that fell out? I've noticed in the past that some special ASCII characters can cause bizarre things like this, but it's not always consistent.

I'm about at the end of my ability to help with this problem without sitting down in front of the doc/app... I'll drop a note to some of the other mods - they might be able to help.

[Shaggy]

Hi guyfawkes...

I'm hoping I missed something but is there any reason that you can't flatten the image in Photoshop before you distill the PostScript? Does the text drop when you flatten all of the layers?

Originally posted by guyfawkes
Not a replacement to QUARK as they are pushing it.

... and on a sidenote: %0 You really think so!?

[Covenant]

I was thinking the same thing.

I can only assume your trying to make a PDF to send to an outside printer? I wouldnt be suprised if they would easily take the EPS or a PSD.

One Idea might be to completely recreate the layer. Might be a pain in the butt if the text is formatted a certain way or has any effects on it, but I've had to do this before.

Also look for Faux bold/italics on text layers. I've had this cause issues with printing.

Hope you find the solution to your problem.



... and on a sidenote: %0 You really think so!? [/B]

Oh I know I dont think so. Every agency I deal with on a design/print basis now has Indesign. My dept uses it exclusively. Infact we've gone through all of our older projects and have converted or recreated quark files in Indesign. It is simply the best page layout program available.

I also do not agree with it being a replacement for vector work. It does not have near the functionality of Illustrator. It is definatly a quark replacement. ;)

But I completely understand about being comfortable with an application. Its hard to try to get your daily work done AND learn new software.

[Shaggy]

Originally posted by Covenant
I can only assume your trying to make a PDF to send to an outside printer? I wouldnt be suprised if they would easily take the EPS or a PSD.

Most printers will accept an EPS or PSD if it is nescessary to the workflow. In this case it sounds as if it would only push the problem forward. Chances aren't very good that the file would RIP correctly if the PDF doesn't work. A printer that is asking for PDF/X1a is specifically trying to avoid these problems.

[David Anders]

((OF course the Distiller version is updated to the current version- if there has been an update))
In this case the Professional Win version is 6.01...
http://www.adobe.com/support/downloads/detail.jsp?ftpID=2366


When having problems with a file that is

1. Very Large (in any dimension - size, imported images or objects, resolution, pages, etc)
AND has worked (printed, saved, pdfed, etc) during the production process and now does not.

Make duplicates of the file and selectively delete percentages of the document copy until it works again.

If you have a problem with one piece - copy and paste it into a new document and test.

ALL FILES have two parts -
the file structure (which exists in a blank document)
and the contents (created in the app or imported, placed, linked)

The File Structure can become damaged and the normal solution is to copy and paste all contents into a new blank document.

If the contents are damaged (imported pictures, linked documents, fonts, etc) selectively removing portions to eliminate pieces is the normal approach.

A damaged font can cause havoc even if it only has a single space located somewhere in the document.

A site that goes into detail about these types of problems with Pagemaker is found at http://www.makingpages.org/pagemaker/tspmdoc.html

These sorts of problems are common in applications that allow outside file import, layers, etc - such as Pagemaker, Quark, Freehand, Illustrator, Word, etc.

Another problem is that some of these apps allow fast saves which save not a document as it is, but the last "saved as" version with a list of changes appended to it.
(I do not believe this is the case with PS, but do not know fer sure)

The RIP in Adobe Acrobat is of course very good, but can be overwhelmed by complex documents and become confused.
Can you try distilling with another version, computer or resolution setting - probably, will not make a difference.

There are sites specifically concerned with PDF - www.pdfzone.com fer instance. 8/


Edit
Did a google search and came up with these on the first page...
PDF/X-3 - Start
Updated PDF/X FAQ, PDF/X FAQ aktualisiert. November 2003 · Martin
Bailey has just finished updating his PDF/X FAQ – answering many ...
www.pdfx1a.com/

WhatTheyThink.com - Print's Home Page
... PDF). PDF/X1a stands for PDF eXchange 1a. ... 2001. PDF/X1a meets the following
specifications: - All fonts and images are embedded. ...
http://members.whattheythink.com/home/pdfx1a.cfm

Communication Arts
... PDF/X1a Specifications. Careful attention must be paid to the proper creation
of PDF/X1a files to ensure that they will reproduce correctly. ...
http://www.commarts.com/advertising/mag_specs_pdf.html
All high-resolution images and fonts must be included when the Postscript file is saved. Only use Type1 fonts-no TrueType fonts or font substitutions.
No RGB or JPEG images. Do not nest EPS file into other EPS files. Do not embed ICC profiles within images.

[guyfawkes]

Hello everyone,
And thank you all for your replies.

PDF x1a are for all Times mags. There is a lot of them.
Not just Time. But the same problem occurs with any very hi rez setting.
Many pubs request only PDF or Tiffit files so they won't have problems.

The flattened EPS file contains the Vector Data.
This files is good...so far. The fonts are good. No fake Italic or bold.
:) We check.

The EPS is put in Quark. Sent to RIP station. A postscipt is created.
This file is good. (We never create PDF from Quark) bad idea.
The postscript file converts to a CT and LW file. This is for proofing.
The proof is OK.

We now make the PDF-X1a first then make the proofs from the PDF-X1a.
This also converts to a CT and LW file on the RIP station. THIS is what the pubs want.

It is the Distiller that can't handle all the Vector Data.
I think??? I assume. What do you all think?

======
Quark / InDesign.
We do love InDesign CS. BUT.....for simple ads, 4 pages or more,
and some other work, QUARK is KING. It's fast, it's problem free.
(Not talking about 6.0)
There are some disturbing bugs with InDesign and multi pages
are a problem. Some of our clients have converted to using
InDesign too. That's fine. But some have have adjusted their thinking.
And I am not suggesting using InDesign exclusively to replace
PS or ILL. But to use InDesign for some PARTS of a project
instead of say using only one app for the entire project.

============
You converted all your Quark files to InDesign?????
Wow. Is this so you don't have to use OS9 inside OSX???
======

Thanks again

:)

[Shaggy]

How exactly are you creating your PDF/X1a? You skipped this over in the last post. It seems that if sending postscript to the RIP works fine then the problem lies in the distiller settings or within the file itself, but we already knew that. Have you tried a twist on David's suggestion: Turn off all text layers, copy the raster information to a new Photoshop document. Then drag each individual text (vector) layer over to the new document (hold shift while dragging to preserve the positioning). Save as an EPS and try again.

Like Lokki, without being there this is even harder to troubleshoot. I could keep tossing out ideas and workarounds but each one leads to a new batch of questions.

If you print just the black plate from Quark and distill that, does the text still drop?

Do your CTs look normal in PhotoShop? (no ghosted text, etc). Can you open your LW files in PressTouch? Is the text all there each time?

What happens when you export an EPS from Quark and place that to a new document?

Isn't printing FUN!?

[frayedknot]

Droppping text:
We had a similar problem.
Parts of a paragraph in our 86 page catalog would just not print, on 5 pages. We use InDesign CS, Photoshop CS, Illustrator CS.
(On Mac OS 10.2.8, I'm on Panther (dept guina pig))

We were thinking it was a font issue and our printer also did what trouble shooting they could. The printer got the error during the RIP process. All paper printed proofs show the text from our's and the printer's paper printers.

I spent a couple of days trying to figure it out and couldn't. The printer didn't find a solution either. It didn't repeat on the next catalog and for the most part the 5 affected pages didn't change that much.

I've noticed PDF's in OS X are not as stable as they used to be. I have to send a PDF and a Stuff'd PDF to my vendor. Sometimes they can't open the PDF (get an error), so they use the Stuffed PDF. Othertimes it's reversed and the Stuffed PDF has the error. I use the same process I used in OS 9. Distilled PS file from Quark. In OS X distilled PS file from InDesign. (Also exporting as PDF from InDesign.) I'm not confident when I send PDF's these days. We also use tiffit files for some vendors and have not had a problem with them.

=====
Quark vs InDesign:
We converted our dept to InDesign. I personally like Quark better. (even with it's bugs) But we moved to OS X and Quark wasn't ready and they took too long to get where they are now and it still isn't ready. InDesign just got the Find/Replace to use Superscript in the Replace in the CS version. (Something we use ALOT.) InDesign 2 was so painful to work with because it couldn't do that. (It couldn't do alot...) The CS version improved alot and is much better to work with. I use it now because I have to. I like Quark better.

Just thought I'd share some... enjoy your day.

[guyfawkes]

Thank you Freyedknot, for your experiences.

And to all, I have yet to take everything apart and test, but, here is
what has happened lately.

I have been swamped with jobs that needed to all ship within hours.
All these jobs had text set in Photoshop. Some had not very much text.
.ps files where made.
PDF files where made.
But, when we went to RIP from the PDF to make .job folder (LW & CT)
the RIP would stop in the CT resolve process. After a half hour we stopped
the RIP.

Not having much time to fix the problem, I opened the Layer files.
The simple looking ones had duplicate white text with drop shadow effects
to get a darker shadow. I Created Layers of the layer effects.
This is in OSX 10.3 and Photoshop CS.
Then changed Type to Shape Layer.
Turned off one of the white layers leaving double layers of shadows.

I did the same to the more complex jobs and copied the Shapes from
one white layer and pasted them onto another white layer to merge the
shapes onto one layer. (same color layers to same color layers.)
These happened to be white. I made sure that the opacity was the same.
I no longer had text. Just shapes. The files where still very large with
the many layers of Vector Data that I could not merge but the jobs worked
and got out on time.
We still sent to Brisque (RIP station) to make .ps files.
Then Distilled to PDF.
Then RIPed the PDF.
Why? to make sure the proof is made from the PDF because this
is what the printer will do with the PDF.

=====
We don't use qxd 6.0 yet. It doesn't work with our system.

I'll let you all know about any testing I do / we do.

Thank you

[sPECtre]

guy, on the first file, do you use the font, or did you convert the text to vector data? I can imagine that converting a whole page of text to a vector layer might choke a RIP.

I suggest also to post in the Adobe Acrobat forum, and of course report here the answers that you receive...

[guyfawkes]

The Text was put in Photoshop by the designer.
This is what did not work. On the original problem of missing type.

The problem with PDFs not RIPing was also Fonts in Photoshop.
When I converted Type to Shapes, that fixed it.
I also cleaned the files up a bit as I explained on the other posts.

Keep in mind that I am always saving with Vector Data.
Text or Shapes.

Good idea on posting to Adobe.

Thanks

[frayedknot]

Glad converting the text to vector worked. (Also surprised it didn't kill the RIP.)

Too bad that wouldn't be a solution to the problem we had. (All text that was a problem was in InDesign.)

I have to 'deal' with a few people in our department who continue to layout whole pages in Photoshop. They don't understand that there are other programs that handle text layout much better than Photohop. (But then they think that all text needs a drop shadow/emboss effect on it...)

Looking forward to any more details on this... hoping to prevent it in the future.

[sPECtre]

Did the designer always use the very same font that is choking the rip? did he use a truetype one? a T1? or an Opentype?
on the other works , is it differents fonts that are missing?