 |
|
|
|
|
|
| |
| |
|
|
From: Jörg 'Yadgar' Bleimann
Subject: Crackle Media - something strange!
Date: 14 Nov 2009 10:13:51
Message: <4afec92f@news.povray.org>
|
|
|
| |
| |
|
|
High!
Another experiment with emitting media, now trying out the crackle
pattern... but the scene looks like the media covers the faces of the
object (a box extending from <-5, -5, 0> to <5, 5, 20>, with the camera
looking into it from <0, 0, -2>) like a texture rather than filling it!
Is this a special feature of the crackle pattern? When I tried the same
pattern with a sphere (see second picture), it appeared distorted like
it was projected onto the sphere's surface!
See you in www.khyberspace.de!
Yadgar
Now playing: Walk Like An Egyptian (Bangles)
Post a reply to this message
Attachments:
Download 'media0000.png' (281 KB)
Download '2009-11-13 media test, take 18 - crackle pattern.jpg' (101 KB)
Preview of image 'media0000.png'
Preview of image '2009-11-13 media test, take 18 - crackle pattern.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> High!
>
> Another experiment with emitting media, now trying out the crackle
> pattern... but the scene looks like the media covers the faces of the
> object (a box extending from <-5, -5, 0> to <5, 5, 20>, with the camera
> looking into it from <0, 0, -2>) like a texture rather than filling it!
> Is this a special feature of the crackle pattern? When I tried the same
> pattern with a sphere (see second picture), it appeared distorted like
> it was projected onto the sphere's surface!
>
> See you in www.khyberspace.de!
>
> Yadgar
>
You will need insanely high samples for such an image to work. What
happens is that when POV samples the inside of the object it falls
most of the time on places where the density is 0, so it misses the
inside strands. I'd say you need at least 100 samples to get
something with your box, but be prepared to wait a looong time for
the render...
Jerome
--
mailto:jeb### [at] free fr
http://jeberger.free.fr
Jabber: jeb### [at] jabberfr
Post a reply to this message
|
|
| |
| |
|
|
From: Jörg 'Yadgar' Bleimann
Subject: Re: Crackle Media - something strange!
Date: 14 Nov 2009 13:08:13
Message: <4afef20d@news.povray.org>
|
|
|
| |
| |
|
|
High!
> You will need insanely high samples for such an image to work. What
> happens is that when POV samples the inside of the object it falls most
> of the time on places where the density is 0, so it misses the inside
> strands. I'd say you need at least 100 samples to get something with
> your box, but be prepared to wait a looong time for the render...
What "samples"? Could I achieve this by upping max_trace_level? Or do
you mean radiosity samples?
See you in Khyberspace!
Yadgar
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> What "samples"? Could I achieve this by upping max_trace_level? Or do
> you mean radiosity samples?
Maybe try to read the documentation about media?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> What "samples"? Could I achieve this by upping max_trace_level? Or do
> you mean radiosity samples?
Meadia samples of course.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> High!
>
> Another experiment with emitting media, now trying out the crackle
> pattern... but the scene looks like the media covers the faces of the
> object (a box extending from <-5, -5, 0> to <5, 5, 20>, with the camera
> looking into it from <0, 0, -2>) like a texture rather than filling it!
> Is this a special feature of the crackle pattern? When I tried the same
> pattern with a sphere (see second picture), it appeared distorted like
> it was projected onto the sphere's surface!
>
> See you in www.khyberspace.de!
>
> Yadgar
>
> Now playing: Walk Like An Egyptian (Bangles)
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
What are your media settings?
If you only have the default of 3 samples, you'll miss almost all parts
that should be visible. If you look closely, there are some floating
points as well as several bright spots in places where you don't have a
front and back lines crossing each others. Those are places where the
sampling did hit an inside strand.
In the case of the sphere, most of what you see are the internal
details, but you need more samples that what you are using. The depth of
the sphere is obviously smaller than that of the box. Each straight line
is part of the internal structure.
Try adding to your media something like:
samples 10
do a test render. You should get some internal structure.
Next, increase samples by +10 until you get a reasonable result.
There will be a point where adding more samples will no longer reveal
more details. When you reatch that point, stop adding more samples.
I did something similar in the past, and I got good results with samples
30. As your container is deeper than in my experiments by a factos of
about 3 to 4, I suggest that you use something like "samples 90" to
"samples 120"
insane... Insane can be something like 1000 or more.
As you seems to use emissive media, samples 150 will not cause excessive
rendering time.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
