How to make published html look as good as the blogs on MATLAB Central?

Check out any of the posts on Cleve's Corner, for example this one. As with all blogs on MATLAB Central it is supposedly published from a MATLAB script using the built-in html publishing feature. But when I try to publish a script with LaTeX equations it looks nothing like this! Inline math is never aligned with the text, and display math is not centered. And all math is low-res images and looks rather bad. Are there some settings that I am missing, possibly undocumented, or is the blogs' html perhaps post-processed by professional tools? More generally, is there anything I can do to make math look better in published scripts?

Answer 1

[per isakson]

I don't think you have missed anything documented.

I had a look at the html-code of Cleve's blog and I think it uses Beautiful math in all browsers, whereas the standard Publish does not. Yes, that is indeed the case, see Cleve’s Corner: Cleve Moler on Mathematics and Computing

The description of the FEX contribution Relativistic rocket by Mikhail contains the paragraph: "See published html help for more information. The html help uses MathJax javascript to display formulas if you are connected to Internet. Otherwise, it falls back to pre-generated PNG images."

And see How can I use MathJax to render the LaTeX math in the published output in MATLAB 7.10 (R2010a)? by the MathWorks Support Team.

And see the discussion in the comments of LaTeX Formula Publishing   and Google finds more for "Matlab MathJax".

&nbsp

I made a small experiment (R2016a) with an example of the documentation in combination with a sentence of Cleve's blog. It confirms your observations.

&nbsp

&nbsp

The rendering of Cleve's blog looks better.

&nbsp

The quality of the rendering of the example in the documentation is similar to that of my experiment.

[Naor Movshovitz]

@per,

Thanks for your experiment and for the tip about MathJax! Good idea to post the question directly to Cleve - I'll do that and link his answer back here.

[per isakson]

I added a couple of links on "Matlab MathJax" to the answer.

[Naor Movshovitz]

thanks for the great links!