is over iso vers. of plant (15 kbu) (Message 1 to 3 of 3)
From: Wlodzimierz ABX Skiba
Subject: is over iso vers. of plant (15 kbu)
Date: 8 Nov 2000 03:22:54
Message: <3a090d5e@news.povray.org>
ough, after parse 0 sec and trace 26:43:23 on P200
I have a small iso plant
it will grow and move in animation if I resolve problem
with redundand calculations
(see post "complicated functions" in p.u.p)
btw: this is my first isosurface (belive or not)
and I think it has possibility to make
even simple interesting trees
ABX
Wlodzimierz ABX Skiba wrote:
> > ough, after parse 0 sec and trace 26:43:23 on P200> I have a small iso plant> it will grow and move in animation if I resolve problem> with redundand calculations> (see post "complicated functions" in p.u.p)> btw: this is my first isosurface (belive or not)> and I think it has possibility to make> even simple interesting trees>
That looks good, I would really like to see the isosurface function used for it.
The rendering time seems quite high for such a relatively simple shape.
I remember having similar rendering times for my 'isosurface landscape' scenes,
but they used complicated pigment functions and radiosity leading to several
billions of Noise/DNoise calls.
Christoph
--
Christoph Hormann <chr### [at] gmxde>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/
From: Wlodzimierz ABX Skiba
Subject: Re: is over iso vers. of plant (15 kbu)
Date: 8 Nov 2000 05:33:15
Message: <3a092beb@news.povray.org>
Christoph Hormann wrote in message
<3A0925CF.2A5EBA24@schunter.etc.tu-bs.de>...
> That looks good, I would really like to see the isosurface function
used for it.
This first try
I want work a little with it before public
perhaps with my own patch for speed it up
>The rendering time seems quite high for such a relatively simple shape.
shape is simple but function not
the main thing is not shape
but modification of shape
this is calculation of two definite integral of sqrt of derivative
of equation parametrized with coordinates
real shape is 20 standard "Witch_of_Agnesi" placed randomly
simple function R is called about 400 times in some subfunctions
and this is no way to simplify - better way is write patch witch allow
local vars or more parameters or calling last results - more at p.u.p or
p.p
>I remember having similar rendering times for my 'isosurface landscape'
scenes,
>but they used complicated pigment functions and radiosity leading to
several
>billions of Noise/DNoise calls.
perhaps it could be also resolved with my planned patch, but I'm not
sure if I can do it
ABX
de>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/