gregjohn wrote:

> Mike Raiford <mra### [at] hotmailcom> wrote:
> 
>>Wow.. Is this a mesh rather than blobs?
>>
>>Very nice ... I'm still trying to make out
>>what he is saying. But, very
>>nice. :)
> 
> 
> I cannot remember whether the lip-sync file was for (the first few seconds
> of)  "The quick brown fox jumps over the lazy dog" or "povray rocks baby."
> 
> Thanks.  The chin is a blobbed stack of about 37 spheres, whose radius and
> position are determined by a natural_spline. Half of them are stationary,
> half of them are rotated by an angle that goes from zero to the angle by
> which the lower jaw is rotated.  One of the latest improvements was simply
> clueing in to the idea that a natural spline would work better than a
> linear one.   If you look at it carefully, you'll see that the tip of the
> chin actually looks like it flabs down a bit more than it should, almost as
> if it were made of rubber.  I thought this was funny so I posted it.
> 
> Also the chin length in profile was totally wrong for the old version of
> MIME Man, even though I though't I'd based it off an anatomically correct
> skeleton.  These are based on the poser guy, as evidenced by the background
> image.
> 
> Now, I'm thinking about how to use this stack-of-spheres to give more
> realistic calf & shoulder muscles, plus lips!

This comes across as a buttload of work to get the job done.  POV-Ray's CSG
is not a very good tool for doing organic shapes; organic shapes aren't
impossible, but they take a heap of work.  One thing I was amazed with
when working on the mesh modeler--especially when combined with a good
surface subdivision algorithm--was the speed by which I could get a model
done.  Things take hours instead of days and come out looking much better.

When the object being modeled is made from cylinders, spheres, cones, etc,
CSG produces much better results for a given effort level, but for organic
shapes Beziers, splines, and subdivision surfaces deliver much better
results for the same investment of time.

Regards,
John

Post a reply to this message