I get e-mails now and then about typography related things, so I figured I would try putting my responses (or revised versions) here for general reference.
Is there a way to determine how much each particular font embedded in a PDF file contributes to the file's size?
There are a few tools that will tell you information about fonts in a PDF. The
pdffonts, which is part of xpdf, seems like the most informative but it does not tell you about how much space an embedded font uses up. Acrobat Reader will also tell you a little bit about the fonts that are embedded in a PDF, but again I couldn't find anything about space used. Even the profession version did not seem offer this information, which seems kind of strange given the various optimization tools it provides.
pdffonts does tell you the object number of each font in the PDF, so if you could extract a specific object from a PDF it would be possible to write wrapper that would tell you how much space the embedded fonts are using. However, I spent a few minutes investigating and did not find any really easy simple way to extract PDF objects. By really easy I mean by a command-line tool. I have to imagine that there is probably something in CPAN that has an API for extracting PDF objects. There is
pdftosrc, part of most modern TeX distributions, that will extract PDF »stream objects«, but these seem to be different than just plain »objects«.