[pdftex] PDFX3 Standard

Siep Kroonenberg siepo at cybercomm.nl
Mon Oct 27 12:38:19 CET 2003

On Sun, Oct 26, 2003 at 05:46:28PM +0100, Verwalter Dr.Betz wrote:
> I recently did a job application using TeXLive7 (6/2002). I was quite
> disappointed, how the included graphics (jpegs from my digicam) came out
> of the hightech printer in the copy shop. Printing out the JPEGs in Corel
> Draw (WindowsXP) was on the other hand no problem. It seemed a problem of
> the PDF format, as I tried several methods of PDF generation without
> success.

What was the problem? Bad color? Low resolution? Did the pdf look ok
in Acrobat Reader?

> Now I recently read of the PDFX3 standard (www.pdfx3.org). There is a
> plugin available for acrobat writer > 4.05 to test for X3-compatibility
> and even to convert to this format. The resulting PDF is considered to
> include every information necessary for any printer: fonts included, ICC
> and lots of funny techstuff.
> Scribus (a new linux DTP) boasts of building PDFX3 files, although I had
> no more success with it than with pdflatex (even when trying to convert to
> X3 using the above acrobat plugin).
> So what about PDFX3 compatibility? Should it not be the default in future
> TeXLive distributions? Already inclusion of ALL fonts needs some tweaking
> (especially when running TeXLive from CDROM). What about including ICC and
> all the other stuff? I did not find much about PDFX3 on the pdftex list.

PDF/X is specifically for prepress. It is rather restrictive,
because it is intented to minimize the likelihood of accidents in a
prepress workflow. You don't want it for general use.

On top of that, I understand that in much of Europe the graphics
industry uses certified pdf, which is based on pdf/x but is not an
open standard.

I myself spent a couple of days getting the latest pdftex (1.11b) to
generate valid PDF/X-3. But if you want color to come out right then
you need the _right_ ICC profile.

In order to get Acrobat 6 Pro to accept pdftex output as valid
pdf/x-3, I put the following code in the preamble:

  \ifnum \month < 10 0\the\month \else \the\month \fi
  \ifnum \day < 10 0\the\day \else \the\day \fi}
  /OutputIntents [ <<
  /Info (none)
  /Type /OutputIntent
  /OutputConditionIdentifier (OFCOM_PO_P1_F60_95)
  /RegistryName (http://www.color.org/)
  >> ]
  /CreationDate (D:\pdfdate)
  /ModDate (D:\pdfdate)
  /Trapped (False)
  /GTS_PDFXVersion (PDF/X-3)
  /Title (\jobname)
\pdfpageattr{/TrimBox [ 0 0 442.08 663.12] }

Without the \pdfcatalog{/OuputIntents... stuff, Acrobat was willing
to convert the pdf to pdf/x, asking me for an OutputIntent. In real
life, it would make more sense to let the printshop do this than to
specify a random OutputIntent just to satisfy the standard. I also
tried the free pdfx3inspector from Callas on a Windows NT box, but
it crashed on me.

It might be a good idea if the pdftex documentation elaborated a bit
on pdf/x.

By the way, does anybody know of some sort of `neutral' or 'don't
care' output intent that still produces valid pdf/x?

> Is there a PDFX3 compatible TeX dialect, e.g. Context?
> I hope, this would spare me to look for a copy shop compatible with the
> momentary TeXLive or Acrobat Writer?
> Yours,
> H. Betz

Siep Kroonenberg
siepo at cybercomm.nl

More information about the pdftex mailing list