interpolated render probleme

[jeromedeverite]

Hello... sorry for my so bad English
Then, here is the problem , the interpolated render of 9.3 ,9.5 is not really accurate when we are in the presence of chalk-lining in the shade(shadow), (in fact a groove of some centimetres in a volume).

When it is directly enlightened there is no problem (see exemple, for one is interpolated, the other is a correct render but with Monte Carlo)
In fact it is necessary to push the regulation really far kind , like mps=0 mes=0 and whith high A-Aliasing to have something which does not drool (almost), but we lose all the saving of time of an interpolated render.

Any solution???
thanks

[3DGFXStudios]

babelfish huh?

[Mitja]

What are your radiosity settings?

[jeromedeverite]

I join you one scene 9.31 , juste one light, no materials, it's just for example

the settings are hight ( i think)
interpolated radiosity - monte carlo
-directional ray ( on)
rpe 180
MES 5mm
MPs 0

camera settings - perspective
AA =8
adaptive sampling 0.05

[Mitja]

I'm checking the scene right now, but you want to render with 9.3 or 9.5?

[Mitja]

Some things I came across:
Delete all 2-point polys and merge points, in modeler.
2-point polys are evil!
In layout: change the light to something different than a distant light (area perhaps?) and set the falloff to inverse square (real light falloff) and set to 0 the ambient light.
This won't affect the "splotches" you are having, but it'll help with realism.
Those artifacts are due to radiosity settings, but since I still don't know whether you'll be using 9.3 or 9.5 I can't help for now.
Maybe 8 AA passes are to high, 4 should be more than enough if you lower the AS to 0.03-0.04. The MPs od 0 isn't so good too.
I'll help more when I'll know if 9.3 or 9.5!
Cheers!

[Mitja]

And yeah, I was forgetting... nice model!

[jeromedeverite]

thanks Mitja, in fact i've test to render it with 9.3 -9.5- and 9.51 , and in none of them i've good render, always ' splotches".
with 9.3 it will be ok

[MooseDog]

hope the attached picture may help.

the settings are getting close, but there still is some blotchiness. i did not deal with the non-planar polys though, so that may be a factor.

tested on an old p4 :). i figure that if i can get close on that box, it should rip on the newer one. render time was under 8mins.

bien fait le model, bien fait :thumbsup:

[jeromedeverite]

thank to spend some time for help me :D
As you can see, it is not so easy to have an appropriate render in that case , but in the city where I live arras (the North France) it is often this style of architecture (but i I like, lol).
With 9.5 by pushing settings
intensity =100
IB = 1
RPE=180
angular tolerance= 45
MPE= 0.5
max PE= 100
Multiplier= 200
I have not too much " splotche ", but render times grow longer :2guns:

[MooseDog]

http://www.except.nl/lightwave/RadiosityGuide95/index.htm

very helpful for understanding these settings. maybe you've seen it though.

in my expereince, when the mpe drops below 2, render times grow.

also, try the multiplier at 50%. still works very well!

bonne chance et merde :hey: