Time for parse 5h 43m
Time for trace 10m 53s
111590857 bytes
480000 triangles
the resulting mesh file is 138,5 mB,
it takes 9m 22s to parse it,

and still the mesh is not smooth :(

 Generated by a macro, that turns a set of parametric functions into a 
mesh of smooth_triangles with uv_vectors, and does displacement 
mapping.  

Ingo

-- 
Photography: http://members.home.nl/ingoogni/
Pov-Ray    : http://members.home.nl/seed7/

Post a reply to this message

Attachments:
Download 'spiky_shell.jpg' (23 KB)

Preview of image 'spiky_shell.jpg'

> Time for parse 5h 43m
> Time for trace 10m 53s
> 111590857 bytes
> 480000 triangles
> the resulting mesh file is 138,5 mB,
> it takes 9m 22s to parse it,
> 
> and still the mesh is not smooth :(
> 
>  Generated by a macro, that turns a set of parametric functions into a
> mesh of smooth_triangles with uv_vectors, and does displacement
> mapping.

You might not need this smoothness in the body, but only in the 'some kind of
spikes' and the junction to the body, then if you're able to concentrate
smoothness there, you might need a lesser number of triangles for an overall
extremely smooth mesh...

But I guess this might not be as easy as I say...

Simon

-- 
||  'How do you know I'm mad?' said Alice.
||  'You must be,' said the Cat, 'or you wouldn't have come here.'
--
 Simon Lemieux  (lem### [at] yahoocom)

Post a reply to this message

in news:3B193E67.8E524650@yahoo.com Simon Lemieux wrote:

> You might not need this smoothness in the body, but only in the
> 'some kind of spikes' and the junction to the body, then if you're
> able to concentrate smoothness there, you might need a lesser
> number of triangles for an overall extremely smooth mesh...

 What I've been thinking is to check the angles of the normals on the 
four points of the polygon. If one angle is bigger or smaller that a 
certain given angle, or deviates too much from the average, subdivide 
the polygon. Problem will probably be finding the threshold when to 
start subdividing / taking more samples. Especialy for more complex 
surfaces like sperical harmonics. 
 http://www.lifesmith.com/spharmin.html

Ingo

-- 
Photography: http://members.home.nl/ingoogni/
Pov-Ray    : http://members.home.nl/seed7/

Post a reply to this message

>  What I've been thinking is to check the angles of the normals on the
> four points of the polygon. If one angle is bigger or smaller that a
> certain given angle, or deviates too much from the average, subdivide
> the polygon. Problem will probably be finding the threshold when to
> start subdividing / taking more samples. Especialy for more complex
> surfaces like sperical harmonics.
>  http://www.lifesmith.com/spharmin.html

Yes, it's somewhat the same thing, but your way seems even better since it could
be used for different kind of surfaces and not only this one...

Simon

-- 
||  'How do you know I'm mad?' said Alice.
||  'You must be,' said the Cat, 'or you wouldn't have come here.'
--
 Simon Lemieux  (lem### [at] yahoocom)

Post a reply to this message