I'm caught in the crossfire between an author and a publisher with an unpleasant sty file :-) The publisher redefines \normalsize to be 9.5pt and changes the math sizes correspondingly. Using lmodern this is not a problem for most symbols, but there are issues with bbm and mathrsfs. The code
\documentclass{article}
\usepackage[T1]{fontenc}
\usepackage{lmodern}
\usepackage{mathrsfs,bbm}
\makeatletter
% taken from the sty file I must use
\renewcommand\normalsize{%
\@setfontsize\normalsize{9.5}{13}%
\abovedisplayskip9.5\p@\@plus2\p@\@minus5\p@
\abovedisplayshortskip\z@\@plus3\p@
\belowdisplayshortskip6\p@\@plus3\p@\@minus3\p@
\belowdisplayskip\abovedisplayskip
\let\@listi\@listI}
\DeclareMathSizes{9.5}{9.5}{\@viipt}{\@vpt}
\makeatother
\begin{document}
Hello world $\mathbbm{N}$, $\mathscr{B}$
\end{document}
returns the understandable warnings
LaTeX Font Warning: Font shape `U/rsfs/m/n' in size <9.5> not available
(Font) size <9> substituted on input line 22.
LaTeX Font Warning: Font shape `U/bbm/m/n' in size <9.5> not available
(Font) size <9> substituted on input line 22.
LaTeX Font Warning: Size substitutions with differences
(Font) up to 0.5pt have occurred.
Assuming that I cannot change the font and considering that optically both \mathbbm{N} and $\mathscr{B} look fine (the difference between 9pt and 9.5pt isn't overwhelming), is there a way to tell TeX to use the 9pt size for \mathbbm and \mathscr? Or is there a way to suppress the warnings for these two math alphabets?
LaTeXhandles fonts): I tried and it works also without adding the line<9.5> bbm9but just adding<9.5>between<9>and<10>(I hope you understand what I mean). Is it the same? – campa Jan 19 '18 at 16:28<9.5>bbm9 <10.95>bbm10, the 9-points fontbbm9will be used for a size request of 9.5 points, scaling it up a tiny bit; on the other hand, if you say<9.5> <10.95>bbm10, the 10-points fontbbm10(the same being used for a request of 10.95 points) will be used for a size request of 9.5 points, scaling it down a tiny bit. You can take your pick: both alternatives are perfectly viable. – GuM Jan 20 '18 at 00:16<9>and<10>, the one withgen * bbm. What would that do? – campa Jan 22 '18 at 08:33genfunction, since a TFM file name likebbm9.5.tfmis not allowed. It turns out thebbm9.tfmit used and scaledat 9.5all the same, but I wouldn’t rely on this and I would not recommend your solution. – GuM Jan 22 '18 at 23:28