1

I think the spacing of some math letter combination is wrong with this minimal document using xelatex, unicode-math and its default font. At least, it looks odd to me. $R_1$ looks ok but $v_f$ and $U_R$ look wrong, so much so that $v_f R_1$ looks like f is bound to R instead of v.

Is this a bug?

Edit: I add below a sample without unicode-math. Admittedly the difference is small, but I find it significant. Is it just me?

\documentclass{minimal}
\usepackage{fontspec}
\usepackage{xunicode}
\usepackage{unicode-math}
%\setmathfont{latinmodern-math.otf} (no change with this command as this is the default font)
\begin{document}
$v_f$, $U_R$, $R_1$, $v_f R_1$
\end{document}

examples of math with odd and adequate spacing

examples of math with correct spacing (the second sample is obtained with no \usepackage; images obtained compiling with xelatex or pdflatex in that case are visually identical, as far as I can tell).

Olivier Cailloux
  • 836
  • Related questions, though different: Placement of subscripts using unicode-math, but it is about vertical positions of subscripts (not the problem here, I think). Also Fixing mathit spacing with unicode-math but it is about \mathit, not involved here. – Olivier Cailloux Oct 26 '15 at 10:31
  • Actually, if you draw a vertical line from the right hand side of the v you'll see that the left side of the fis correctly positioned. Maybe subjective, but I don't feel the spacing is off... But you can always add a negative space, \! after v. – Fredrik Johansson Oct 26 '15 at 10:37
  • Comparing both samples (with and without unicode-math) it appears the vertical positions of subscripts is very different. Perhaps related to the subscript xetex bug but it has been marked solved in 2012. I use “XeTeX 3.14159265-2.6-0.99992 (TeX Live 2015/dev/Debian)” – Olivier Cailloux Oct 26 '15 at 10:53
  • @FredrikJohansson: what do you mean with the f being “correctly” positioned, is there some accepted standard? On the second picture (added after your comment), f is a bit closer to v (about 2/3 distance compared to the first picture, considering the right edge of v and the left edge of f). I think the second picture looks better, especially considering $v_f R_1$. – Olivier Cailloux Oct 26 '15 at 11:09
  • @FredrikJohansson (about adding a negative space): thanks for the tip, but I’d prefer not manually “correcting” latex’s typesetting (especially if it is not incorrect!). I want my document to be considered good looking and easy to read by my readers, my own opinion should not count. Hence I’d like to know whether the current spacing is considered wrong or ok. In the second case I’ll be happy to let (xe)latex do its job. – Olivier Cailloux Oct 26 '15 at 11:11
  • No, afaik there isn't an accepted standard - hence my saying it's subjective, but then English is far from being my first language. What I mean is that to me, the fact that the f is as far to the left as it can get without intruding on the position of the v, gives it a good placement. I believe math characters aren't kerned in the same way as text characters, but I may be wrong - I often am. – Fredrik Johansson Oct 26 '15 at 11:33
  • @FredrikJohansson: your appeal to subjectivity was clear. My point however is precisely that f could be moved to the left while staying vertically disjoint from the v vertical position. That’s what’s done in the second picture for example, which I find more satisfactory (and it could be done even more, as zooming in the pictures reveals, but I guess some minimal amount of horizontal space should be used). – Olivier Cailloux Oct 26 '15 at 18:00
  • I am also seeing a very bad spacing using unicode-math with thebeamer class, for example in \beta^i – oibaFox Mar 10 '20 at 18:17

0 Answers0