Possible Duplicate:
xfrac + siunitx gives me a font warning

This document (New Century Schoolbook font in \SI)






produces this warning

LaTeX Font Warning: Size substitutions with differences
(Font)              up to 2.01195pt have occurred.

Additional information: This does only happen with the NC font in \SI with \per ("power to the -1"). Outside of \SI this does not happen (for example `$a^{-1}$ does not produce the warning). This does only happen on the units, not on magnitudes (10^-1).

  • Why does this happen?
  • Does siunitx require another font size for powers? Why?
  • How to fix this warning?

marked as duplicate by lockstep, Claudio Fiandrino, zeroth, Marco Daniel, Thorsten Oct 10 '12 at 17:02

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 1
    Load the fix-cm package. See tex.stackexchange.com/questions/32378/… – lockstep Oct 10 '12 at 10:20
  • 1
    Thanks, this removes the warnings, but care to explain why this package works for the New Century font when it should fix problems in Computer Modern? – Foo Bar Oct 10 '12 at 10:23
  • Not sure myself, but siunitx seems to rely on Computer Modern for certain symbols even when other font packages are loaded. – lockstep Oct 10 '12 at 10:25
  • BTW, without fix-cm I do get a warning outside tabular. Please test again and, if true, edit your question. – lockstep Oct 10 '12 at 10:30
  • Thanks, you are right. Maybe I did something different on the first run. I'll edit the question. – Foo Bar Oct 10 '12 at 10:38

The siunitx package jumps through a lot of 'hoops' to give the correct appearance of output as far as possible. That means quite a bit of font detection and math/text mode switching. The warning can be generated by an example such as


where you'll note that there is no change of font at all (without \unboldmath the warning is slightly different). That I know of, there should be no change in the fonts actually used in the output: a quick check shows that everything looks OK.

  • 1
    Thanks, but how does this solve the problem? It's just another example, a bit more "low-level" however. Why does your code produces the warning and what do you suggest to fix it? – Foo Bar Oct 10 '12 at 12:47

Not the answer you're looking for? Browse other questions tagged or ask your own question.