19

In a 2011 EuroBachoTeX talk, Ulrik Vieth discusses TeX unicode math fonts and their state of completeness.

He lists the following 4 fonts as (more or less) production-ready:

  • Cambria Math
  • Asana Math (Apostolos Syropoulos)
  • XITS Math (Khaled Hosny)
  • Latin Modern Virtual (Hans Hagen) [ConTeXt only]

And the following 4 fonts as "under development":

  • Neo Euler (Hermann Zapf, Khaled Hosny)
  • Lucida Math (Bigelow & Holmes, Khaled Hosny)
  • Latin Modern Math (GUST Foundry)
  • TeX Gyre Math (GUST Foundry)

As far as I can tell, Neo Euler development looks like it has stalled. The last update to the Neo Euler GitHub repository seems to have been 4 years ago.

Is the future of Neo Euler known? What about the development of the other unicode-math fonts?

emacsomancer
  • 1,795
  • 1
    Latin Modern Math is good, I think. Similarly, XITS Math. No idea about the others. – cfr Dec 24 '14 at 23:45
  • This question has a more up-to-date community wiki answer here: https://tex.stackexchange.com/questions/425098/which-opentype-math-fonts-are-available/425099#425099 I’d redirect this as a dupe, but your question was first. – Davislor Sep 05 '18 at 02:11

1 Answers1

17

The Latin Modern Math and TeX Gyre Math support projects are complete. Together, these fonts provide a total of 5 fonts for typesetting mathematics, all produced by the GUST e-foundry. Details can be found here.

That page lists an additional six fonts, produced by other foundries, which support the mathematics opentype extension, including 3 available from CTAN:

  • Asana Math;
  • Stix Math (v.2 expected 2016-12-31 according to barbara beeton;
  • Xits Math (contains a bold version);

and 3 proprietary:

  • Cambria Math;
  • Lucida Math (contains a demi bold version);
  • Minion Math by Johannes Küster (contains a bold version).

GUST also provide this comparison document which covers all of the above fonts except for Minion Math.

That information is dated May 2014. What about new fonts? How can you evaluate their suitability?

  • I wrote this answer a while back to explain why a particular font combination did not work with unicode-math. In that answer, I compared that font combination with Latin Modern Math, which provides proper support. You can use my answer to assess the suitability of other fonts for use with unicode-math since I explain, basically, what a font needs in order to support typesetting maths using the interface this package provides. (The answer is: quite a lot.)

EDIT

Here's a mini-demo of the 5 maths fonts provided by GUST:

maths font demo

\documentclass{article}
\usepackage{amsmath}
\usepackage{unicode-math}
\NewDocumentCommand \testme { o }{%
  \IfValueT{#1}{%
    \setmainfont{TeX Gyre #1}%
    \setmathfont{TeX Gyre #1 Math}%
    TeX Gyre #1 \& TeX Gyre #1 Math:
  }%
  \begin{align*}
    p(D_k/T)  &       = \frac{p(T/D_k) p(D_k)}{p(T)} \\
    &       = \frac{p(T/D_k) p(D_k)}{ \sum_{i=1}^n  p(T/D_i) p(D_i)}  & \text{where } \sum_{i=1}^n p(D_i) = 1 \\
  \end{align*}%
}
\begin{document}
Latin Modern Roman \& Latin Modern Math:
\testme

\testme[Bonum]

\testme[Pagella]

\testme[Schola]

\testme[Termes]
\end{document}
cfr
  • 198,882
  • 1
    That is useful information. I'm still most curious about Neo Euler - which is rather different in appearance from the other fonts. I'm looking for something which works well for on-screen viewing (projected lecture slides), and Neo Euler seems nice, particularly in combination with the Ubuntu font. – emacsomancer Dec 29 '14 at 00:29
  • 1
    @emacsomancer The last commit is a bit less than a year old. Have you tried contacting the authors? – cfr Dec 29 '14 at 00:40
  • Nowadays what opentype font with nice math do you suggest?. Asana, NewPX, Neo-Euler or TeX Gyre?. Preferably something with text similar to Palatino (because of university restrictions). I'm going to use it with LuaTeX. – skan Nov 25 '16 at 13:21
  • @skan I'm not the right person to ask. I don't use unicode-math. I'm waiting for other people to find all the bugs :-). Also, I don't typeset much maths period. What I do use is not projected on screen (except logic). But, in any case, I use only pdfLaTeX for real documents. I want the reliability, speed and stability. For text, I can always install a font if I really want it. (Unless the licence prohibits conversion of an .otf.) As for maths, the options are certainly not more restrictive than they would be with unicode-math. – cfr Nov 25 '16 at 18:04
  • @skan I don't know what NewPX is. TeX Gyre is not one font but several. The obvious choice for Palatino is obviously TeX Gyre Pagella Math. Is there some reason you don't want to use that? – cfr Nov 25 '16 at 18:06
  • "it is impossible to typeset “real” mathematical papers using only TeX Gyre fonts, because so far math symbols are scarcely represented", http://www.gust.org.pl/projects/e-foundry/math/index_html – skan Nov 25 '16 at 21:25
  • @skan During the BachoTeX 2014 meeting, the Project was announced complete: the math companions for four TeX Gyre serif fonts (Bonum, Pagella, Schola, Termes) were released. You are taking a historical document as current. That's explaining why the project was needed and what it aimed to achieve. For TG Pagella, it is complete. – cfr Nov 25 '16 at 21:32
  • @skan http://www.gust.org.pl/projects/e-foundry/tg-math/gust_e-foundry-math_fonts-presentation.pdf But you aren't obligated to use unicode-math. Personally, I wouldn't. But it can hardly be said that TeX Gyre Pagella Math doesn't provide at least a reasonable proportion of what is needed for maths - enough, I'd assume, for many 'real' maths documents. Obviously, some will need specialist symbols. But that's true whether you use traditional maths fonts or unicode maths fonts. – cfr Nov 25 '16 at 21:38
  • @skan Please see edit above. – cfr Nov 25 '16 at 22:44
  • @cfr OK, thanks. I just want to know your opinion about different alternatives. – skan Nov 25 '16 at 23:33
  • 1
    @skan Well, I wouldn't use unicode-math. But if you do and you're using Palatino, Pagella is the obvious choice. That's really all there is to it. You won't get a closer match to Palantino than a Palantino clone, unless you have Palantino Maths also, which almost certainly doesn't exist. – cfr Nov 25 '16 at 23:42
  • 1
    release of stix version 2 is expected approximately december 31 (2016). this version has been tuned for acceptable appearance with text as well as with math. – barbara beeton Nov 26 '16 at 01:11
  • @barbarabeeton Thanks. :-) I'm a bit confused by the discussion I'm having in these comments, to be honest :|. – cfr Nov 26 '16 at 01:45
  • @cfr I'm also confused. Where is Euler, exactly? I see above references to the euler-otf repo, whose tagline says it is abandoned. Is there another, more recent, effort? Or was that it? – Jim Hefferon Nov 26 '16 at 12:15
  • 1
    @JimHefferon As far as I know, that's it. It is an abandoned project which remains 'for archaeological purposes' according to the description. barbara might know if any later efforts exist. – cfr Nov 26 '16 at 13:53