 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi,
I have a silly problem.
This is a dummy background for my next (hopefully) IRTC entry.
Problem is memory consumption. This little scene with 6 houses (from Gilles
are placed and randomly rotated within a while loop) needs whooping 830 MB
peak memory.
One single tree consumes only about 21 MB.
A 3.5 bug?
Does anybody know more ?
Norbert
Post a reply to this message
Attachments:
Download 'dummy.jpg' (51 KB)
Preview of image 'dummy.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> A 3.5 bug?
I don't think it's a 3.5 bug. I don't know why there's so much memory
consumption, but the best way to reduce it is to use image_map on very thin
boxes. Have a look to this image and it's description :
http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.jpg
http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.txt
Sylvain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
If youre using radiosity (I can't say that from this notebook) you might
expect such requirements (though I'm really in doubt...)
Without radiosity you would expect something like 20*8=160 MB (since IIRC
only the foliage is a mesh), but with high radiosity settings this can
quickly raise.
If you're not using radiosity forget it.
--
Jonathan
"Norbert Kern" <nor### [at] t-online de> ha scritto nel messaggio
news:3bab8e86@news.povray.org...
> Hi,
>
> I have a silly problem.
> This is a dummy background for my next (hopefully) IRTC entry.
> Problem is memory consumption. This little scene with 6 houses (from
Gilles
(they
> are placed and randomly rotated within a while loop) needs whooping 830 MB
> peak memory.
> One single tree consumes only about 21 MB.
> A 3.5 bug?
> Does anybody know more ?
>
> Norbert
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Norbert Kern" <nor### [at] t-onlinede> wrote in message
news:3bab8e86@news.povray.org...
> Hi,
>
> I have a silly problem.
> This is a dummy background for my next (hopefully) IRTC entry.
> Problem is memory consumption. This little scene with 6 houses (from Gilles
> are placed and randomly rotated within a while loop) needs whooping 830 MB
> peak memory.
> One single tree consumes only about 21 MB.
> A 3.5 bug?
> Does anybody know more ?
>
> Norbert
>
Are you using the file output option when generating the tree and including
the file 8 times, or are you calling the macro directly 8 times with identical
parameters? I would expect the latter to consume more memory than the former,
as well as being slower to parse. The arrays and the recursion would eat up
a considerable amount of memory above and beyond the actual objects used in the
trees.
--
Batronyx ^"^
bat### [at] cadronhsacom
http://www.batronyx.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
For image-mapped objects in a scene to reduce memory requirements of the
final scene, see my site at:
http://www.angelfire.com/md2/castlewrks
"Sylvain Boutry" <bab### [at] yahoofr> wrote in message
news:3bab9a64@news.povray.org...
> > A 3.5 bug?
> I don't think it's a 3.5 bug. I don't know why there's so much memory
> consumption, but the best way to reduce it is to use image_map on very
thin
> boxes. Have a look to this image and it's description :
> http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.jpg
> http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.txt
> Sylvain
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Yes, I used the file output option and memory saving was only minimal.
Norbert
>
> Are you using the file output option when generating the tree and
including
> the file 8 times, or are you calling the macro directly 8 times with
identical
> parameters? I would expect the latter to consume more memory than the
former,
> as well as being slower to parse. The arrays and the recursion would eat
up
> a considerable amount of memory above and beyond the actual objects used
in the
> trees.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I know your technique since your fine entry in the Garden round.
All 1 GB RAM of my machine were in full use with about 17 mio polypons in my
warm_up entry.
Norbert
"Robert J Becraft" <cas### [at] aolcom> schrieb im Newsbeitrag
news:3babb8bc$1@news.povray.org...
> For image-mapped objects in a scene to reduce memory requirements of the
> final scene, see my site at:
>
> http://www.angelfire.com/md2/castlewrks
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
all meshes can be downloaded from his homepage IIRC). Its definetively fast
parsing and memory saving.
Nowadays I have a new 1.4 GHz Athlon C with 1 GB Ram and thougt to make
"real" 3d scenes.
But perhaps I have to substitute the full shadowed trees. thanks for the
tip.
Norbert
"Sylvain Boutry" <bab### [at] yahoofr> schrieb im Newsbeitrag
news:3bab9a64@news.povray.org...
> > A 3.5 bug?
> I don't think it's a 3.5 bug. I don't know why there's so much memory
> consumption, but the best way to reduce it is to use image_map on very
thin
> boxes. Have a look to this image and it's description :
> http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.jpg
> http://www.irtc.org/ftp/pub/stills/2000-06-30/cbwild.txt
> Sylvain
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi JRG,
yes, I use radiosity. But the rca file is <100 MB for the full scene with
good quality settings.
Its a very tricky problem with no solution so far.
Norbert
"JRG" <ant### [at] liberoit> schrieb im Newsbeitrag
news:3bab9bf0@news.povray.org...
> If youre using radiosity (I can't say that from this notebook) you might
> expect such requirements (though I'm really in doubt...)
> Without radiosity you would expect something like 20*8=160 MB (since IIRC
> only the foliage is a mesh), but with high radiosity settings this can
> quickly raise.
> If you're not using radiosity forget it.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Norbert Kern wrote:
>
> Hi,
>
> I have a silly problem.
> This is a dummy background for my next (hopefully) IRTC entry.
> Problem is memory consumption. This little scene with 6 houses (from Gilles
> are placed and randomly rotated within a while loop) needs whooping 830 MB
> peak memory.
> One single tree consumes only about 21 MB.
> A 3.5 bug?
> Does anybody know more ?
>
> Norbert
How many light_sources have you got?
In several scenes featuring tens of light sources, I found that
switching light buffers off dramatically decreased memory consumption.
Concerning rendering time, the impact was not that terrible, but I might
have been lucky.
--
__ __ __ __ _
| | / \ / / |_ / |/
\/\/ \__/ /_ /_ |__ \_ |\
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
