# [pdftex] pdflatex/epstopdf questions

Richard Sperling rsperling at boo.net
Mon Sep 10 09:59:53 CEST 2001

```I'm having some trouble with pdflatex 1.00a. I run MikTeX 2.1 with WinEDT
5.2 on Windows 2000 service pack 1. There are [three] \includegraphics
declarations in my .tex document. The graphics are eps files named
fig1.eps, fig2.eps, and fig3.eps.  I'm trying to use the epstopdf package
written by Heiko Oberdiek to do the graphics conversion. The document
preamble contains:

\usepackage[pdftex]{graphicx}
\usepackage{epstopdf}

When I try to compile the .tex file using pdflatex I get the following
sequence of errors:

Error: May not be a PDF file (continuing anyway)
Error (0): PDF file is damaged - attempting to reconstruct xref
table...
Error: Couldn't find trailer dictionary
Error: Couldn't find xref table

Error: pdflatex (file fig1.pdf): xpdf: reading PDF image failed
**********end*********

I have a few questions. First, I noticed that Sebastien Rahtz reported
similar errors when using version 0.12n of pdftex. The text of his message
follows:

***begin***

having just tried pdftex 0.12n under Windows (was Linux before), I can see
why Hans gets depressed (see appended). does it work for anyone under Windows?

sebastian

This is pdfTeX, Version 3.14159-0.12n (Web2c 7.2)
**\relax
isc.map><monotype.map><xy.map><hoek.map><uca.map><usb.map><mt.map><pjn.map><uty
.map><em.map><ptg.map> *\pdfimage 8-7-1
pdf Error (0): PDF file is damaged - attempting to reconstruct xref table...
Error: Couldn't find trailer dictionary
pdfTeX error (ext1): invalid image dimension.
<*> \pdfimage 8-7-1.pdf

! ==> Fatal error occurred, the output PDF file not finished!
Transcript written on texput.log.

***end***

Is there a way around this problem? I could not find a follow-up to this
message.

I can get pdflatex to create the pdf file, but it takes 5 runs. The first
run produces fig1.pdf and gives the error messages above. The
second run produces fig2.pdf with the error messages above, and the third
run produces fig3.pdf. The final 2 runs resolve the
bibliographic references. That seems like an inordinate number of pdflatex
runs. Obviously this process becomes a nightmare with more than a few
graphics files.  Hence, my second question. Assuming the aforementioned
errors can be fixed, shouldn't epstopdf.sty/epstopdf.exe be doing the
graphics conversions in a single run? BTW, I am invoking epstopdf.exe with
-shell-escape on the command line.