Maxwell Render | Official public forum

Hi!

I did this test of a glass with wine in it... Tryed ThomasAn. technique... and so on...

2 things I discovered make me a little nervous when thinking of the final version... since I did not hear any statement whether these issues will be fixed in final version.

1. When I look at the apillarity of the wine... (I knwo it's not modeled too well) why is it still complete red? It sould fade to colorless...
ThomasAn. describes a nice workaround, but I thought M~R aims to become absolutaly physically based render.
Are my settings wrong?? More absorbance, less color!?!?

2. Maxwell in some cases does not really seem to converge to a noisefree solution... in cases of using dielectrics or SSS.

This is what I'm talking about:
Level 23.64 reached within 34h

reference:


Level 20.24 reached

reference:


Lightsource:
(everything except the black material is assigned as emitter):



The only thing I want to know is, whether NL is aware of this problem. ( I guess they are)
And the more interessting question: Will they get this under control?

Just want to add: I absolutally take my hat off to NL-Team for developing this software!!!! It just makes me nervous a little especially because I was thinking about reselling M~R.

I'd be so grateful if NL would just drop a line here

You have to optimize your lightsource, more polys in lightsource=more noise.

maybe the cause of all that noise is the geometry of your light source, when rendering with maxwell it's safe to use very simple/low poly count objects as emitter to reduce the noise.

edit:
whoops, mihai again. he always types faster than me.

Because I also use my feet when typing.

well, your spelling is really good even if feet typed.

very nice work

you can give it for all freely? for testing thanks

ciao
torolf

cool tests.... by default Maxwell samples "wish" is set at 25... is it where all noise disappear for sure... mmmmm ...

@Mihai Iliuta, @rivoli

So you are saying M~R Is not really capable of rendering a scene with a 'complex' emitter liek this... !?

The "Breaking barriers" tells a different story. N lights 25 emitters 4'045'400 faces

And I'm not really happy with a software that is 'overstrained' with this scene / emitter.

@Hervé: That is what I thought... considering this as example I'd say no

I mean if it is the case... (impossible to have a no noise render), then I guess NL is concentrating on this issue with number ONE priority... speed is one thing, but noise is in my humble opinion much more important.... and with or without dielectrics...

sam7 wrote:
So you are saying M~R Is not really capable of rendering a scene with a 'complex' emitter liek this... !?

Well, its not that its not capable, its just that for the beta the more faces you have on an emitter the more noise it generates, and the render times increase exponentially, meaning, sooner or later (days months yrs) it will converge to a highly accurate noise free solution....

For the time being (until late october one hopes), in order to have less noise and faster render times its better to use a single plane for a light source...

Hi I rendered this Scene with RC5 now WITH havy use of indirect light.
So it's not comparable to the images above.

sl18.33
13h46m


As far as I remember I can say that RC5 seems to handle indirect Light much better than in beta!

But as you can see it's still pretty much grain.
And problems of caustics (from indirect light) through glass.

Another Problem: I set transmittance to white and attenuation distance to 999m. But the glass still seems a little to dark.

can someone confirm this?

Sam7, good beta tests and hope you will repeat them with RC5. It's great to see the photos with the M~R.

sam7, it depends also how reflective you've made the glass, if it's heavily reflective then much light will not pass through but instead reflect back.

If you want to post the scene I can look at your materials.

Mihai Iliuta wrote:sam7, it depends also how reflective you've made the glass, if it's heavily reflective then much light will not pass through but instead reflect back.

If you want to post the scene I can look at your materials.

Please post the scene so that this problem might be solved by the community.

ludenhud wrote:

Mihai Iliuta wrote:sam7, it depends also how reflective you've made the glass, if it's heavily reflective then much light will not pass through but instead reflect back.

If you want to post the scene I can look at your materials.

Please post the scene so that this problem might be solved by the community.

OK

No Problem:
Scene