Maybe this is more appropriate for meta, but I'll try it here first (please move if inappropriate).
I have noticed for a while now that @egreg has some very decent screenshots of examples; for instance, on Positioning includegraphics vertically on a line?, for the MWE:
\documentclass{article}
\usepackage{graphicx}
\begin{document}
Just testing something here... Lorem ipsum dolor sit amet,...
\noindent%
Attempt at signature:
\makebox[0pt][l]{\hspace*{2cm}\raisebox{-3ex}{\includegraphics{testing.png}}}%
\hrulefill
That's all, folks ...
\end{document}
... I, using ImageMagick's convert
from the command line:
convert -density 300 -resample 150 test.pdf test.png
..., obtain something like this:
...; while @egreg obtains something like this:
... where the fonts look a lot better.
Maybe it has to do with @egreg's original image being 1990x452 pixels, while mine was 776x183 pixels - but in this post I've scaled his image down (cubic) to approximately the same resolution as mine, and it still looks better.
I'm assuming we're both just using pdflatex test.tex
to compile the MWE, so it should be the same PDF; how is this nice(r) rasterization then obtained? Is it just rasterizing at greater resolutions at first - or is it down to the rasterizing tool(chain)?
standalone
class might be better suited ;-)