So, those that have been keeping themselves up-to-date about my current
activities, here's an image from the ongoing scripting process.

There are 10.000 hairs in this image, each made of 50 smooth, uv-mapped
triangles (though the uv-mapping isn't used). But I only use 17 MB memory.
Why? Because the 10.000 hairs are selected from an initial bunch of 254
mesh2-hairs and are then rotated to point at the correct direction. The
pointing-algorithm isn't working 100% as planned yet, so I'm just rotating
around the y-axis needed to align the surface-normal to a certain degree.
There's some angular-rotation missing, which I will try to solve tomorrow.
This effect is most noticable along the silhouette, the hairs seem to create
layers instead of one random bunch.

Generally, this approach is meant to take advantage that mesh2-objects will
be instanced instead of using the entire memory for the same object again,
as is the case with CSG (two spheres will stay two spheres, two identical
meshes will be one mesh and one point in memory).

Anyways, the macros aren't finished yet, this was a quick
"hack-em-together"-job to get some first results. Although fairly convincing
already, I want to do some additional tests.

Later on, these macros will be made simple to set and can be used to either
visualize your own data, or use data passed over from my Surcoat-Macros. In
this manner, putting hair/fur on objects shouldn't be that difficult
anymore. :-)

Note that the hair will receive a better 3d-model later on (2d-blankets
don't look to good in a silhouette), when I've got the basic stuff scripted.

Once I got some more interesting stuff done (e.g. using arbitrary geometry
or such), I'll post another demo-pic. :-)

Regards,
Tim

-- 
"Tim Nikias v2.0"
Homepage: <http://www.nolights.de>

Post a reply to this message

Attachments:
Download 'hairtest.jpg' (118 KB)

Preview of image 'hairtest.jpg'