LaTeX + chemscheme + bkchem = Ugly?
I just finished up my first polished draft on my prospectus, and I was unhappy with how my schemes looked. I was using the chemscheme package to automatically manage referencing and numbering my molecules, but a requirement for this amazing feature is postscript files with plain text markers. I use bkchem for my chemical structure drawing, and while awsome, it’s postscript export option gave me two-ish choices: Cairo or Piddle/built-in. Cairo results in amazing looking output, however this is because it strokes everything, which leaves no text markers for chemscheme to pickup on. Bkchem can also use an internal version of piddle (now called SPING) or the built-in Tk functions, which properly uses regular text for chemscheme to find, however the output in both is very ugly.
Some recent experimenting has finally led me to a solution. First, while piddle and “built-in” both had text spacing issues, they were far less noticeable in piddle (when I started, piddle would throw up erros on export, so I was using built-in, however this has somehow resolved itself). Switching from “built-in” to piddle made things look better spacing wise. Secondly, it seems both piddle and “built-in” default to using ‘Times’ as a postscript font (trust me, Serif fonts are ugly in structures…). I’m not sure how configurable this may be at any level (certainly no options in bkchem), but my quick and dirty solution was to simply sed out “Times” and replace it with “Helvetica” (the default bkchem font, this way the output looks close to what I see when I am drawing). This is certainly a fine font that doesn’t make me cringe when looking at my schemes. I only wish I figured this out before I gave the draft to my advisor…
Anyway, a summary of how I use these schemes in LaTeX/chemscheme and a general workflow will follow (including pitfalls, and problems I had to learn to fix along the way)