 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I have noticed that when I have atmospheric media in a scene, and a light
and an object I want to interact with the media, I have to use very high
numbers of samples and intervals to get nice "rays". (Imagine a sphere with
holes, and a light inside and media around, and the nice rays of light that
pop through the holes.) How come 3dsmax does this sort of thing quickly and
well? How can this be improved? There must be some trick or shortcut that
will provide the same look with less computational cost... right?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Um, maybe extinction 0?
"Tony[B]" <ben### [at] catholic org> wrote in message
news:3bfc5002@news.povray.org...
> I have noticed that when I have atmospheric media in a scene, and a light
> and an object I want to interact with the media, I have to use very high
> numbers of samples and intervals to get nice "rays". (Imagine a sphere
with
> holes, and a light inside and media around, and the nice rays of light
that
> pop through the holes.) How come 3dsmax does this sort of thing quickly
and
> well? How can this be improved? There must be some trick or shortcut that
> will provide the same look with less computational cost... right?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Um, maybe extinction 0?
How exactly does that speed things up, Bob?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3bfd19fe@news.povray.org>, ben### [at] catholicorg says...
> How exactly does that speed things up, Bob?
I'm not sure extinction 0 increases speed, but adding no_shadow to the
container object does. Unfornutately, athmospheric media doesn't have a
container object...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 21 Nov 2001 20:13:26 -0500, "Tony[B]" <ben### [at] catholicorg>
wrote:
>How come 3dsmax does this sort of thing quickly and
>well? How can this be improved? There must be some trick or shortcut that
>will provide the same look with less computational cost... right?
What you see in a pixel is the integral of the global illumination
function along a ray through that pixel. POV does this integration
explicitly, i.e. by sampling. It has to. MAX can do it directly. It
all comes to simple (compared to noise3d isos :) ) calculus and linear
algebra.
Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG e-mail : pet### [at] tagpovrayorg
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> What you see in a pixel is the integral of the global illumination
> function along a ray through that pixel. POV does this integration
> explicitly, i.e. by sampling. It has to. MAX can do it directly. It
> all comes to simple (compared to noise3d isos :) ) calculus and linear
> algebra.
Well, hopefully PovRay 4 will use more calculus and lineair algebra.... :-)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Lutz-Peter Hooge" <lpv### [at] gmxde> wrote in message
news:MPG.16673d50e53a86059896bd@news.povray.org...
> In article <3bfd19fe@news.povray.org>, ben### [at] catholicorg says...
>
> > How exactly does that speed things up, Bob?
> I'm not sure extinction 0 increases speed, but adding no_shadow to the
> container object does. Unfornutately, athmospheric media doesn't have a
> container object...
My intent by suggesting that was you wouldn't need as much scattering or
density perhaps, thus reducing the sampling of media "particles" or pixels.
Thing is, I seem to recall there being a speed-up if the media is denser and
less able to be sampled deeper. I could be wrong.
Maybe you can get by with using a set 'intervals' value to prevent that from
going higher? Not sure if that goes for v3.5 beta.
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <fptqvtsneu0nkriphm42lcghujilodf1ur@4ax.com>, pet### [at] vipbg
says...
> What you see in a pixel is the integral of the global illumination
> function along a ray through that pixel. POV does this integration
> explicitly, i.e. by sampling. It has to. MAX can do it directly. It
> all comes to simple (compared to noise3d isos :) ) calculus and linear
> algebra.
Why is this not possible with POV? Ok, I know that it wouldn't work with
things like noise3d, granite etc, but for an athmospheric media with only
a spherical or gradient density....?
Lutz-Peter
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <MPG.16682a183e0047499896bf@news.povray.org> , Lutz-Peter Hooge
<lpv### [at] gmxde> wrote:
> Why is this not possible with POV? Ok, I know that it wouldn't work with
> things like noise3d, granite etc, but for an athmospheric media with only
> a spherical or gradient density....?
Lets just say "all patterns are created equal" *. Anything else would
introduce an enormous amount of overhead in code.
Thorsten
* Not entirely true, but very close.
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Fri, 23 Nov 2001 09:41:58 +0100, Lutz-Peter Hooge <lpv### [at] gmxde>
wrote:
>Why is this not possible with POV? Ok, I know that it wouldn't work with
>things like noise3d, granite etc, but for an athmospheric media with only
>a spherical or gradient density....?
Even with a constant density, object cast shadows in it, and if these
objects are complex... well, you get the picture.
With meshes only, it's much easier. It all comes down to projections.
Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG e-mail : pet### [at] tagpovrayorg
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
