[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Orgmode] a bit offtopic, fonts in exported PDF documents

From: Nick Dokos
Subject: Re: [Orgmode] a bit offtopic, fonts in exported PDF documents
Date: Thu, 26 Aug 2010 18:44:20 -0400

Erik Iverson <address@hidden> wrote:

> For those that don't want to read all that follows, please note my
> conclusion, given here:
> Since orgmode is automatically telling latex to use T1 encoding,
> perhaps we should somewhere document to the user that Type 1
> fonts should be available to get the best looking PDF possible.
> Otherwise, type 3 fonts will be substituted. I got suitable
> Type-1 fonts by installing the texlive-fonts-extra package
> under Ubuntu.

[With apologies for the length (and the further off-topic direction) of
this post, I hope the following is of interest to a few people. Those
of you who have no interest in the finer points of TeX fontological history
can safely hit "Delete" now.]

I think the conclusion is substantively correct, but there are a couple
of minor nits: I would leave out the first line of the second paragraph
above ("Since orgmode...T1 encoding,"). I would also note that the problem
is a viewer problem, not a Type3 problem (see below). Here's why:

* Type1 fonts (e.g. the Adobe Postscript standard fonts) are implemented
  as scalable outlines, basically programs that describe what the glyph
  will look like. It's easy to tell the program: draw this glyph at
  1.41421 the size (or whatever other scaling factor you want), hence

* Type3 fonts are usually bitmaps (that's actually an oversimplification, but 
  true in the vast majority of cases), so you need a multiplicity of
  them at different sizes to cover the needs of a document.

Both of these (and presumably the missing Type2 as well, although I have
never seen one of those) were defined by Adobe.

T1-encoding has nothing to do with Type1 fonts except for the
unfortunate similarity of the names: an encoding is just the table that
translates from numbers to glyphs in a font. Knuth used a very peculiar
encoding originally (now called OT1) and in 1990, the TeX people got
together at Cork, Ireland, and hashed out this new encoding (it was
called Cork encoding for a long time, but at some point it became
"T1"). The T1-encoding *is* pretty much the encoding that Adobe used for
*its* fonts (whether Type1 or Type3). In particular, there were
T1-encoded Type3 fonts (the so called EC fonts, in their original
incarnation, before they were auto-traced and made into Type1 fonts) and
there are non-T1 encoded Type1 fonts (in particular, T1-encoding deals
with text, mostly European languages that use the Latin alphabet or
slight variations thereof, so Cyrillic, Greek, non-European alphabets
and symbols of various kinds have their own encodings: whether a font
for any of these would be Type1 or Type3 is a matter of implementation.)

Here are some links of interest from the TeX FAQ:

"What are encodings?"   http://www.tex.ac.uk/cgi-bin/texfaq2html?label=whatenc

"What are the EC fonts?" http://www.tex.ac.uk/cgi-bin/texfaq2html?label=ECfonts

"Adobe font formats" http://www.tex.ac.uk/cgi-bin/texfaq2html?label=adobetypen

This last one (and a link therein to PDF quality) is of particular
interest, because it describes the situation when Acrobar Reader
(version < 6) exhibited exactly the problem that you encountered with
Evince. Acrobat Reader has been fixed since then so it deals with Type3
fonts reasonably, but I guess Evince still mistreats Type3 fonts (they
may believe in the motto that "All the fonts (that matter) are Type1".)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]