On Thu, Jan 19, 2012 at 1:25 AM, Kip Warner <kip@thevertigo.com> wrote:
Hey list,

I have a number of \externalfigure commands to typeset some svg images.
This is often handy to have them vectorized because it allows for the
user to zoom in on diagrams of great detail.

In other situations, I find that for images that are not so important,
they would benefit from ConTeXt rasterizing them when they are imported.
The reason for this is I find some vector images get enormously bloated
when they are typeset. As an example, I have a Logo.svg of only 6.8 KB.
The m_k_i_v_Logo.pdf intermediate ConTeXt generates is nearly a megabyte
in size. Why? I have no idea, as they both appear to be the same image
with the same gausian blur.

Is there a flag I can pass certain instances of \externalfigure that
tells the image processing backend that a simple rasterized version of
the figure is fine? Or is there a better approach? I don't want to
pre-convert all of these images to raster formats.


ConTeXt uses inkscape, iirc. So try to see what inkscape says about your file.
If you are under linux, you can also use rsvg .

--
luigi