Radiosity test and a question

[ThirdEye]

I think you have to increase the max resolution (please post here a pic with stochastic samples and your radiosity settings, I'll be more precise)

[amapimaster]

oh bollocks- didnt save it,

settings ( i think )

strength 150
diffuse 4
stoc 200
min 20
max 100

Also on radiosity in general, the radiosity tutorial in the manual uses a sphere with a light source outside of it. Im trying different methods of lighting scenes, this one is a large luminous plane above the scene at a slight angle with no other lights.
Do any of you have a specific setup you use?

[ThirdEye]

Every scene has different own settings. You have to start with low values (i always start with 20 stochastic samples, 1 min res and 10 max res) and remember that min res must be at least 1/10 of the max res (for example 3-30). I keep always high the accuracy value (between 85% and 95%), then I increase min and max resolution due to the distribution of stochastic samples and their density. When I have reached a good distribution I increase the number of stochastic samples et voilà. (this is not my personal interpretation, this is the explaining of finalRender official video tutorial)

[derwolpertinger]

ok, you gotta increase your max resolution. but also increase the min res. otherwise the rendertime would increase dramatically cause c4d gotta interpolate between those two values. and the greater the the distance between these values the longer the rendertimes. i also think that it would be fine to decrease your stoc samples a bit. for me i always use less samples and have very fast rendertimes. could you tell us your accuracy value plz and your rendertime for this pic, too. hope i could help you out in some way. :D

[heretic]

Why this 1 to 10 ratio? I'm not sure I understand....

Thanks!

[ThirdEye]

min and max res have to be not too distant between them to keep fast the calculation. A 1/10 ratio is optimal to keep high the quality and fast the calculation

[heretic]

This doesn't make sense to me since a min of 10 would require 100 max samples using your ratio...... This is quite a difference!

Am I missing something?

Thanks again!

[kiwi]

Hang on that abnormality looks suspiciously like a polygon behind a polygon which messes up your mesh in a hyper nurbs......damn thats a terrible explanation :annoyed: ...I will try again if you dont get what I mean.


Try selecting those front polygons and see if there is a line of ploys immediately behind the first,if there is and they are not part of the back of the sofa then thats why you have that nasty abnormaility and gi wont fix it.


Also make sure all the normals are alingned as well.



Stu.

[AdamT]

The first thing I would do is use a point or area light in place of the luminous plane. Lighting with luminous objects requires much higher stochastic samples than using lights.

[ThirdEye]

@heretic: when you analize a stochastic sample distribution over a scene you have first to optimize max res in critical zones. Then you can adjust min res using the 1/10 ratio. I've never used 100 max res, if you keep high the accuracy (85-95%) you'll see that a value like 35 would be high ;)

[LucentDreams]

Originally posted by AdamT
The first thing I would do is use a point or area light in place of the luminous plane. Lighting with luminous objects requires much higher stochastic samples than using lights.

Just to go with what adam said, I use Stochastic oinly mode if I am using a luminous object as a main source fo lights, mainly for the same resons as why you use if for animation rather than the basic radiosity mode, it is more even and predictable, though typically slower and requiring higher sample to get rid of noise.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.