2

The following output pdftex > dvi > ps shows the problem: enlarged font problem

This comes about as follows. First, I use the file

\pdfoutput=0
\hsize=5truein
\tenrm

\pdffontexpand\tenrm 20 20 20
\pdfadjustspacing=2

Our {\it major topics} include random walks and their intimate connection to
electrical networks; uniform spanning trees, their limiting forests, and
their marvelous relationships with random walks and electrical networks;
branching processes; percolation and the powerful, elegant mass-transport
technique; isoperimetric inequalities and how they relate to both random
walks and percolation; minimal spanning trees and forests, and their
connections to percolation; Hausdorff dimension, capacity, and how to
understand them via trees; and random walks on Galton-Watson trees.
Connections among our topics are pervasive and rich, making for surprising
and enjoyable proofs.

\bye

Second, I made two new versions of cmr10 as in this answer. Third, I ran the above file with pdftex and then dvips. I viewed it with gv (using the default antialias) and zoomed in on a part to show above. Adding a line to use \pdfpkresolution = 72 (or larger) made no difference. How can I correct this so that the new fonts are like the old in quality?

  • 1
    If you've created expanded source fonts (.mf), then they are not scalable and certain readers (especially a certain reader) is known to handle these fonts very badly with precisely the results you show. If you wish to avoid this, whatever fonts you use (expanded or otherwise) must be scalable. It doesn't matter whether you're creating the fonts of they are pre-packaged. You'll get this effect unless you stick to scalable fonts. – cfr Jan 12 '16 at 23:26
  • Although you are viewing postscript with gv, I assume something similar is at work, given the contrast. That is, at least with the settings you're using, it doesn't handle .mf well. I'm guessing that it is using scalable fonts for the unexpanded ones. However, you'd need to confirm these suppositions. – cfr Jan 12 '16 at 23:28
  • @cfr I see. So once I create scalable fonts, they will have to be installed like other Type 1 fonts; is that right? – Russ Lyons Jan 12 '16 at 23:48
  • P.S. The expanded fonts are Type 3, but I don't think this says whether they are scalable. They don't look good even not zoomed. – Russ Lyons Jan 12 '16 at 23:55
  • You can tell from the console output precisely what it is using. But those are not scalable. In any case, if the fonts you produced involved .mf and .pk files, then they are not scalable. You know how you produced them. If you used the first method in that question for source fonts, you'll have non-scalable. If you used the second for type1 fonts, you'll have scalable. – cfr Jan 13 '16 at 00:40
  • Try typesetting `\documentclass{article} \usepackage{humanist} \begin{document}

    \hminfamily ABC

    \end{document}` using (pdf)LaTeX. Or change it into TeX if you prefer. This font is only available as source. You'll see precisely the same effect if you zoom a bit and just the same 'type' in the properties for the document listing the font as type3.

    – cfr Jan 13 '16 at 00:44
  • @cfr Yes, I used .mf and .pk. Your example indeed is similar. – Russ Lyons Jan 13 '16 at 01:46
  • Why do you think these are low-quality? The whole TeX+Metafont+CM project of Knuth was created for producing output for a digital typesetter, so it went all the way and produced the actual pixels for the output device (printer). Have you tried printing it out and seeing the output? Or looking on your screen at a normal zoom level? Most screens are just 100–200 dpi and MF was written to target a typesetter at 5333.33 dpi so you should be able to get great output without much trouble. Instead of something awful like \pdfpkresolution = 72, try 2400 or 8000. – ShreevatsaR Apr 20 '17 at 05:53
  • @ShreevatsaR I just tried what you suggest. As in my original post, it made no difference. I did not print it out; it is important that it look good on screen. As I wrote in a comment earlier, it does not look good at normal zoom. – Russ Lyons Apr 20 '17 at 16:44

0 Answers0