Heya,
On 4/01/2008, at 9:02 AM, Andy Jones wrote:
Are the textures stored locally or somewhere remotely on the network?
Local. One figures this optimization out pretty darn quick as soon
as you start using .map files.
The MR docs talk about setting the textures as "local", it's in the
memory mapped textures section.
I think what you're talking about has more to do with ray hosting
(off-loading buckets to other machines), which I'm not doing.
Hmmm..., the docs say that it only caches the map textures when they
are set to 'local' which leads me to believe it's for all renders and
not just ray hosting. I'd say it'd be worth a try but I can't seem to
find anywhere in XSI that lets you set this parameter so you'll have
to edit the mi file.
I was using scanline with raytracing enabled. I suppose I should
have given it a shot with the rasterizer, in case it does texture
lookups in a different order, but since the sampling pattern is
different anyway, I think it would be hard to compare unless one or
the other is WAY faster. But I will say that if the rasterizer is
way faster than scanline for texture lookups, something is rather
"broken."
Well, not really - as it all depends on how the texture is being
accessed.
In this case, I've been testing on very small little render regions,
not full frames. So, I think the slowness would still be a problem,
even if I tiled the render. I think the main reason for doing this
has to do with the output buffers occupying too much memory. I
think the problem here is different.
It may help, we've found that it's more than just the output buffer
size forcing us to move to sliced renders. But if the problem is disk
thrashing ( which it is like you said ) then it might just make it
worse.
cheers, colin.
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
