Bug 691617 - Missing glyphs in subsetted T42 fonts
Summary: Missing glyphs in subsetted T42 fonts
Status: RESOLVED INVALID
Alias: None
Product: Ghostscript
Classification: Unclassified
Component: PDF Writer (show other bugs)
Version: 8.71
Hardware: All All
: P4 normal
Assignee: Ken Sharp
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-09-16 12:35 UTC by Rolf Becker
Modified: 2010-09-27 14:26 UTC (History)
0 users

See Also:
Customer:
Word Size: ---


Attachments
EPS with subsetted T42 fonts (425.30 KB, application/postscript)
2010-09-16 12:35 UTC, Rolf Becker
Details
PDF created by GS (47.75 KB, application/pdf)
2010-09-16 12:37 UTC, Rolf Becker
Details
Preflight screenshot (of similar file) (27.11 KB, image/png)
2010-09-16 12:40 UTC, Rolf Becker
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Rolf Becker 2010-09-16 12:35:44 UTC
Created attachment 6726 [details]
EPS with subsetted T42 fonts

An EPS with subsetted T42 fonts (sudoku.eps) is converted to PDF (sudoku.pdf):
gs -sDEVICE=pdfwrite -sOutputFile=sudoku.pdf -dEPSCrop sudoku.eps.
The PDF is displayed well in Acrobat Pro, but the prefight complains about the missing glyphs adieresis, odieresis, and udieresis in the font AEEJHI+Arial.
On a Harlequin RIP, these glyphs are visible if the EPS is ripped, and invisible if the PDF is ripped.
Comment 1 Rolf Becker 2010-09-16 12:37:28 UTC
Created attachment 6727 [details]
PDF created by GS
Comment 2 Rolf Becker 2010-09-16 12:40:04 UTC
Created attachment 6728 [details]
Preflight screenshot (of similar file)
Comment 3 Rolf Becker 2010-09-16 12:43:51 UTC
Sorry, the preflight complains, not the prefight.
Comment 4 Ken Sharp 2010-09-16 13:24:37 UTC
(In reply to comment #0)
> Created an attachment (id=6726) [details]
> EPS with subsetted T42 fonts
> 
> An EPS with subsetted T42 fonts (sudoku.eps) is converted to PDF (sudoku.pdf):
> gs -sDEVICE=pdfwrite -sOutputFile=sudoku.pdf -dEPSCrop sudoku.eps.
> The PDF is displayed well in Acrobat Pro, but the prefight complains about the
> missing glyphs adieresis, odieresis, and udieresis in the font AEEJHI+Arial.

Which preflight profile ? You appear to be using a PDF/X profile to preflight a file which has not been prepared for PDF/X. Also the preflight screen capture is of a 'similar file', I'm not sure what relevance that has.

The preflight supplied with Acrobat is not 100% reliable in any event.


> On a Harlequin RIP, these glyphs are visible if the EPS is ripped, and
> invisible if the PDF is ripped.

I have opened/rendered the supplied PDF file with Ghostscript, Acrobat Pro and a JAWS-based rip. In every case the dieresis glyphs are present and there are no font substitution messages.

I see the words "Füllen", "Sudoku-Rätsel", "Kästchen", "Vergnügen" and "Auflösung" using the accented characters, should there be others ?

Finally I extracted the embedded TrueType font and decoded it. I see from the CMAP that 0xFC (ü) maps to glyph ID 129. Glyph 129 is at offset 0x4700 in the font, it is a composite glyph (presumably u and dieresis) but is present in the font. So this glyph is not missing (despite teh complaint of the preflight)

In short, I can't see a problem with the supplied file. Perhaps you have uncovered a problem in the Harlequin rip ?
Comment 5 Rolf Becker 2010-09-16 14:08:00 UTC
The similar file was a PDF created from a 80 MB Postscript file. I extracted the EPS from that file, created a PDF and made a preflight, but not a new screenshot of that. I did not care of the profile, I just saw the missing glyphs.
I tried the PDF with Ghostscript, Acrobat, Foxit, Poppler without problems.
A PDF created by Distiller was rendered correctly on the Harlequin.
I suppose there is something in the font that does not conform 100% to the spec which causes some font engines to complain.
Comment 6 Rolf Becker 2010-09-27 14:10:55 UTC
Ken, it was indeed a bug in the Harlequin rip which is fixed in the meantime.
I regret to have caused your investigations, but I was confused about the Acrobat preflight results which I thought to be reliable.
Thank you for your efforts.
Comment 7 Ken Sharp 2010-09-27 14:26:34 UTC
(In reply to comment #6)
> Ken, it was indeed a bug in the Harlequin rip which is fixed in the meantime.
> I regret to have caused your investigations, but I was confused about the
> Acrobat preflight results which I thought to be reliable.
> Thank you for your efforts.

Thanks for the update, its great to know for sure what the problem was.