![](/i/fill.gif) |
![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Here's some experiments with meshes deformed in standard MegaPov using a
POV-script to move vertices and change normals.
When looking at the render times below, note that I'm using an old Pentium
150 with 16 MB RAM.
For both image 1 and 2 the parse time was about 10 min. and the peak memory
was about 9 MB.
For image 1 (grey pigment) the trace time was about 7 min. making a total
render time of about 17 min. For image 2 the trace time was about 15 min.
making the total render time about 25 min.
These results can be compared to the results produced by ABX's patch shown
at the end of the thread "beauty of twisted rays". I don't know which
computer he's using though.
Note though that for these images the textures were not deformed together
with the meshes. Deformed textures would with my method most likely increase
the memory use a lot and the parse time too. However, if/when the UV-vectors
in the mesh2 feature in MegaPov are changed to support 3D vectors, texture
deformation can be done without considerable increase in memory and render
time.
Comments?
Rune
--
\ Include files, tutorials, 3D images, raytracing jokes,
/ The POV Desktop Theme, and The POV-Ray Logo Contest can
\ all be found at http://rsj.mobilixnet.dk (updated January 28)
/ Also visit http://www.povrayusers.org
Post a reply to this message
Attachments:
Download 'chevy1.jpg' (26 KB)
Download 'chevy2.jpg' (34 KB)
Preview of image 'chevy1.jpg'
![chevy1.jpg](/povray.binaries.images/attachment/%3C3a7d5a5c%40news.povray.org%3E/chevy1.jpg?preview=1)
Preview of image 'chevy2.jpg'
![chevy2.jpg](/povray.binaries.images/attachment/%3C3a7d5a5c%40news.povray.org%3E/chevy2.jpg?preview=1)
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
On Sun, 4 Feb 2001 14:28:57 +0100, Rune wrote:
>Here's some experiments with meshes deformed in standard MegaPov using a
>POV-script to move vertices and change normals.
Awesome.
Now for next week's homework you will write a macro that can produce
a non linear scale on a mesh.
--
Cheers
Steve email mailto:ste### [at] zeropps uklinux net
%HAV-A-NICEDAY Error not enough coffee 0 pps.
web http://www.zeropps.uklinux.net/
or http://start.at/zero-pps
2:38pm up 2 days, 16:16, 2 users, load average: 1.38, 1.37, 1.23
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Ha! You did it. : ) Cool!
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
"Steve" wrote:
> Awesome.
Thanks!
> Now for next week's homework you will write a macro that
> can produce a non linear scale on a mesh.
Well, how do you think I did these images? :)
However, it'll take a while to finish it, write documentation and so on...
Rune
--
\ Include files, tutorials, 3D images, raytracing jokes,
/ The POV Desktop Theme, and The POV-Ray Logo Contest can
\ all be found at http://rsj.mobilixnet.dk (updated January 28)
/ Also visit http://www.povrayusers.org
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Rune wrote in message <3a7d5a5c@news.povray.org>...
> For both image 1 and 2 the parse time was about 10 min.
my parse time was 7 secs with dos version in windows under NT
I use Pentium II 233 with 128 RAM but I use at least 10 main parallel tasks
including financial database server serving data for clients
povtask had low priority
> and the peak memory
> was about 9 MB.
peak memory was 5 MB in my version
> For image 1 (grey pigment) the trace time was about 7 min. making a total
> render time of about 17 min. For image 2 the trace time was about 15 min.
> making the total render time about 25 min.
my render time was about 1h43m
> These results can be compared to the results produced by ABX's patch shown
> at the end of the thread "beauty of twisted rays". I don't know which
> computer he's using though.
result is the same becouse I think triangles of mesh are small comparing to the
pixel size, I think we should compare something bigger :-)
> Note though that for these images the textures were not deformed together
> with the meshes.
Note that with my method they will be automatically deformed together
> Deformed textures would with my method most likely increase
> the memory use a lot and the parse time too.
And note that in my method deformed textures not increase memory use and not
increase parse time.
ABX
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
"Wlodzimierz ABX Skiba" wrote:
> Rune wrote:
> > For image 1 (grey pigment) the trace time was about 7 min.
> > making a total render time of about 17 min.
>
> my render time was about 1h43m
If we compensate for our different computers my method is perhaps about 10
times faster.
> result is the same becouse I think triangles of mesh are small
> comparing to the pixel size, I think we should compare something
> bigger :-)
The Chevy model is a typical mesh model I think, so it's fair to use it as
an example. I know very well that your method has the obvious advantage that
it works perfectly on all objects. No method is better than the other, it's
just a matter of what you need to do.
> > Note though that for these images the textures were not
> > deformed together with the meshes.
>
> Note that with my method they will be automatically deformed together
With mine too if just you turn the feature on, which I didn't here, because
it would be pointless since all the pigment were plain pigments.
> > Deformed textures would with my method most likely increase
> > the memory use a lot and the parse time too.
>
> And note that in my method deformed textures not increase memory
> use and not increase parse time.
With mine neither as soon as mesh2 UV-mapping supports 3D-vectors.
But hey, at least my Chevy model had real textures! ;)
Rune
--
\ Include files, tutorials, 3D images, raytracing jokes,
/ The POV Desktop Theme, and The POV-Ray Logo Contest can
\ all be found at http://rsj.mobilixnet.dk (updated January 28)
/ Also visit http://www.povrayusers.org
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |