[Bubbaloo]

I have done an interior scene with multilight to test sunlight in interior renderings. All glass material was set up using the wizard AGS. All geometry in the scene has a basic 1-layer diffuse gray with 99 roughness. Here I will show the rendering results:

No Sunlight:


At S.L. 10 at 1280 px, the image is relatively clean.

With Sunlight:


Also S.L. 10 at 1280 px, the image is very grainy due to the specks of light from the sun's bounces.

Would sunlight portals reduce this effect?
Would the render clear faster?

My next test will be to replace the exterior windows with emitters to see if that is an option.

I would like to hear the thoughts of you guys who have more experience lighting interior scenes. And the thoughts of Next Limit developers on light portals and if they can fit in with Maxwell's philosophy of physically correct lighting.

[Fernando Tella]

An emitter simulating sunlight would perform better.

[lebbeus]

try taking out the glass (hide/delete the geometry) entirely if you're going to render with the sun.

AGS/clipmapped/void materials are doing something funky with caustics even though they're not supposed to be creating caustics in the first place (remember my double-sphere test with the clipmap outer sphere and the 'butter' inner sphere??)

[Bubbaloo]

Ok, I'll try it.

[JTB]

try taking out the glass (hide/delete the geometry) entirely if you're going to render with the sun.

This is the obvious solution, however, the reflection shown on windows is lost and so is the reality and the quality of our renderings.

[oz42]

Bubbaloo - I'd like to see sunlight portals too but apparently they don't make too much difference. A certain other unbaised renderer used to have them but have since taken them out as they didn't make a significant speed increase.

However I think your test still has great value and I'd like to see various other setting if you have the time e.g. no sunlight - with and without glass, sunlight - with and without glass, HDRi - with and without glass and finally, emitter - with and without glass. It would be interesting to see both what they all look like at a certain sl and also how long it took to get there.

Keep up the good work!

[Bubbaloo]

An emitter simulating sunlight would perform better.

Hey Fernando,
Are you talking about emitters in place of the glass or a bright emitter placed outside the scene to simulate sunlight?

[Bubbaloo]

Thanks, OZ.
I'll try to do more tests next week.

[lebbeus]

try taking out the glass (hide/delete the geometry) entirely if you're going to render with the sun.

This is the obvious solution, however, the reflection shown on windows is lost and so is the reality and the quality of our renderings.

true, but in this case not all of the "glass" would have to be removed, nor will it be missed due to the shot and exposure--you won't see reflections on the panes in the back of the scene even if the glass was there.

you could always run two renders; one without glass and another with highly reflective black plastic in place of the planes and then composite in Photoshop

[Fernando Tella]

An emitter simulating sunlight would perform better.

Hey Fernando,
Are you talking about emitters in place of the glass or a bright emitter placed outside the scene to simulate sunlight?

I was thinking about the second with the emitter not too far away so the light doesn't spread too much over the outside part of the scene but not too close to keep it credible. I haven't tested it but that's an idea that has been going on in my head for some time for this kind of scene.

[kami]

I tested it they way you described. It's a lot faster and you don't lose too much realism!

[Brett Morgan]

I had the same problem very recently, rendered the image (with sunlight only) with and without AGS glass and even at an SL of 20 there were some very bad grainy spots on the ceiling, and to be honest removing the glass from the window panels made little difference, is this a bug or something we maxwell users have to live with?

Brett

[JTB]

It is not a bug, it is a problem of the current version of rendering engine.... Dielectrics.... the difficult part of rendering.

[tom]

AGS/clipmapped/void materials are doing something funky with caustics even though they're not supposed to be creating caustics in the first place (remember my double-sphere test with the clipmap outer sphere and the 'butter' inner sphere??)

I can't agree because your example is only valid when there is SSS. Clipmaps and ghost layer should render in same efficiency like there-is-no-object without introducing anything funky.

Bubbaloo, can you try doing the following?
1) Remove windows and see if it makes a difference in terms of noise and I guess not.
2) Turn off/Hide emitters when rendering with sun and see if it renders much faster.

[Bubbaloo]

With AGS to S.L. 10:


Without Glass Panels to S.L. 10:


Not much difference in noise amount. But the scene brightness is different.

All other emitters were turned off.