I don't want to brag, but I think this is a breakthrough for radiosity:

The attached images show a purely radiosity-lit scene, using 
pathologically small light emitters (three tiny spheres).

Image #1 shows the familiar splotchy output even at the maximum sample 
ray count of 1600 and a nearest_count of 10, after a render time of 1 
minute 45 seconds on my Core i7 machine.

Image #2 shows how this can be improved (not surprisingly) by using a 
count of 6400 instead, after having replaced the built-in sample 
sequence with a 160,000-element halton sequence, of which a subset of 
6400 was picked for each sample in a round-robin fashion. Splotches are 
almost gone, but there's a caveat: Render time increased to 7 minutes 20 
seconds.

Image #3 however shows even better quality with a rendering time of just 
2 minutes 18 seconds. On the same machine. With virtually identical 
settings. So what's the deal here?

The basic idea is pretty simple: Trace fewer radiosity sample rays if 
they only hit non-important objects :-)

Of course we can't know in advance whether we will hit important or 
non-important objects; however, we /can/ decide to perform full 
computations only on a subset of rays hitting non-important objects 
(applying a factor to compensate for the discarded rays), so that at 
least we need to perform fewer texture computations - not to mention 
secondary rays for reflections or shadow computations.

For image #3, 12,800 rays were shot per sample, but I guess only roughly 
2-5% of them (i.e. about 256 - 640) were continued beyond the ray-object 
intersection tests.


Hey, I like this!

Post a reply to this message

Attachments:
Download 'rad_count_test_37a_1600__1m45s.png' (177 KB) Download 'rad_count_test_38plus_6400of160000__7m20s.png' (177 KB) Download 'rad_count_test_38plus_12800of160000_weighted__2m18s.png' (170 KB)

Preview of image 'rad_count_test_37a_1600__1m45s.png'

Preview of image 'rad_count_test_38plus_6400of160000__7m20s.png'

Preview of image 'rad_count_test_38plus_12800of160000_weighted__2m18s.png'

clipka <ano### [at] anonymousorg> wrote:
> I don't want to brag, but I think this is a breakthrough for radiosity:
> ...
> Hey, I like this!

Go on, brag. You've got the right :) This looks like a wonderful breakthrough
indeed.

Sam

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> Image #3 however shows even better quality with a rendering time of just
> 2 minutes 18 seconds. On the same machine. With virtually identical
> settings. So what's the deal here?

Wait a minute, in the third image there is less reflected light on the plane
under the red sphere, and to a lesser degree, under the green sphere. Is it
supposed to be that way?

Post a reply to this message

Am 22.08.2010 02:42, schrieb Samuel Benge:
> clipka<ano### [at] anonymousorg>  wrote:
>> Image #3 however shows even better quality with a rendering time of just
>> 2 minutes 18 seconds. On the same machine. With virtually identical
>> settings. So what's the deal here?
>
> Wait a minute, in the third image there is less reflected light on the plane
> under the red sphere, and to a lesser degree, under the green sphere. Is it
> supposed to be that way?

I'm not quite sure - I might still have one or two flaws in the code. I 
guess I'll check with MCPov for comparison.

Post a reply to this message

"clipka" <ano### [at] anonymousorg> schreef in bericht 
news:4c706dff@news.povray.org...
>I don't want to brag, but I think this is a breakthrough for radiosity:

Good work! Congrats!

Thomas

Post a reply to this message

On 08/21/2010 09:23 PM, clipka wrote:
> I don't want to brag, but I think this is a breakthrough for radiosity:
> 
> The attached images show a purely radiosity-lit scene, using
> pathologically small light emitters (three tiny spheres).

Looks good ... I've not had too much luck with purely radiosity lit
scenes in the past, so bragging rights are indeed yours!

> Image #2 shows how this can be improved (not surprisingly) by using a
> count of 6400 instead, after having replaced the built-in sample
> sequence with a 160,000-element halton sequence, of which a subset of
> 6400 was picked for each sample in a round-robin fashion. Splotches are
> almost gone, but there's a caveat: Render time increased to 7 minutes 20
> seconds.
> 
> Image #3 however shows even better quality with a rendering time of just
> 2 minutes 18 seconds. On the same machine. With virtually identical
> settings. So what's the deal here?
> 
> The basic idea is pretty simple: Trace fewer radiosity sample rays if
> they only hit non-important objects :-)

Be sure to keep us //all// up to date with your progress and I'll make
sure to integrate your findings into the documentation.

> Hey, I like this!

Concur!

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> The basic idea is pretty simple: Trace fewer radiosity sample rays if
> they only hit non-important objects :-)

is that the so called importance sampling? :)

yeah, pretty neat, but I noticed the same as Sam too...

Nice to see radiosity still improving.

Post a reply to this message

On 22-8-2010 2:42, Samuel Benge wrote:
> clipka <ano### [at] anonymousorg> wrote:
>> Image #3 however shows even better quality with a rendering time of just
>> 2 minutes 18 seconds. On the same machine. With virtually identical
>> settings. So what's the deal here?
> 
> Wait a minute, in the third image there is less reflected light on the plane
> under the red sphere, and to a lesser degree, under the green sphere. Is it
> supposed to be that way?

I am not sure it is not an artefact of our brain. In the second image 
there are 'lines' of slight contrast just like in the first one, where 
they are horrible. It might be that we misinterpret those lines as if 
they are reflections of the floor although they are mere artefacts.

Post a reply to this message

> I don't want to brag, but I think this is a breakthrough for radiosity:
>
> The attached images show a purely radiosity-lit scene, using
> pathologically small light emitters (three tiny spheres).
>
> Image #1 shows the familiar splotchy output even at the maximum sample
> ray count of 1600 and a nearest_count of 10, after a render time of 1
> minute 45 seconds on my Core i7 machine.
>
> Image #2 shows how this can be improved (not surprisingly) by using a
> count of 6400 instead, after having replaced the built-in sample
> sequence with a 160,000-element halton sequence, of which a subset of
> 6400 was picked for each sample in a round-robin fashion. Splotches are
> almost gone, but there's a caveat: Render time increased to 7 minutes 20
> seconds.
>
> Image #3 however shows even better quality with a rendering time of just
> 2 minutes 18 seconds. On the same machine. With virtually identical
> settings. So what's the deal here?
>
> The basic idea is pretty simple: Trace fewer radiosity sample rays if
> they only hit non-important objects :-)
>
> Of course we can't know in advance whether we will hit important or
> non-important objects; however, we /can/ decide to perform full
> computations only on a subset of rays hitting non-important objects
> (applying a factor to compensate for the discarded rays), so that at
> least we need to perform fewer texture computations - not to mention
> secondary rays for reflections or shadow computations.
>
> For image #3, 12,800 rays were shot per sample, but I guess only roughly
> 2-5% of them (i.e. about 256 - 640) were continued beyond the ray-object
> intersection tests.
>
>
> Hey, I like this!

Very good! Congratulation, you are welcome to brag about your work :)

The 3rd image seems the best to me. In the others, the area under the 
large spheres, especialy the red one, look to bright.


Alain

Post a reply to this message

clipka wrote:

> The basic idea is pretty simple: Trace fewer radiosity sample rays if 
> they only hit non-important objects :-)

Sounds great!

How do you decide if an object is non-important?

Would this also help for scenes where light falls
through a narrow opening (door, blinds, ...)?

Post a reply to this message