[bjorn.syse]

I made a test with SSS volumes enclosed in PET-plastic, which took ages to clear, and never did. It was discussed in another thread that enclosing the SSS volume in a caustic transparent material (ND other than 1) was a bad idea, since all the light SSS has to work with is caustic light. I was recommended to use the AGS solution for my PET-plastic instead (no caustics, just reflections). These are the tests compared.



The avocado soap (second from the end) is a bit greener/lighter, and obviously the plastic has lost it's attenuation. The AGS version still has 1,5 hour to reach the same SL, but even when it does, I'm nore sure the noise levels are much lower. Do you?

btw, feel free to fill in with more SSS tests in this thread.

[tom]

Could you try replacing emitters with IBL illumination?

[bjorn.syse]

This scene uses both, should I remove the emitters on the right?

[tom]

Yes and you could try a studio-like HDR image.

[bjorn.syse]

I did use a studio like HDR, but it was called a LDR. I'm not sure if it makes a difference. It was the BackTopLeft-one from Hyltoms collection. However, in this new test to the right, I removed the emitter totally and changed that Hyltom LDR to another one, and it seems even clearer than the first one only after 25 minutes rendering.

What is going on here?

[bjorn.syse]

Ok, here's an update. I made a small region render to quickly see what would happen around SL 18, which is a HUGE improvement over the other two. And all this because I did two things.

1. I switched Hyltoms LDR (this one)


...to another one looking like this:


2. I removed a plane with an emitter material and consequently let the IBL be my only lighting.



So, what's the lesson here?

[tom]

I think there's something strange with the material settings. I'll take a look and hopefully cure if you can send it to me.

-Edit: Oh, cool. You made it work, then. The lesson is, yet SSS is not efficient with emitters and sun but it's still possible having good results with IBL. For emitter like illumination, you will need those studio-type HDRIs.

[bjorn.syse]

Allright, so maybe I could even go back to my original caustic plastic If I stick to just the IBL and steer away from emitters?

Howcome emitters work so bad? or can I just make them larger, big softboxes, and get the same results as using IBL?

[bjorn.syse]

I have a question. What if I create a plane, make an emitter material but map an MXI or HDR-texture to this emitter. Would that clear as fast as an enviroment HDR or would I encounter the same behavior as using a plain emitter?

[Hervé]

any emitter with an mxi or hdr mapped to it is considered as an emitter... so yes ibl is the one to use..
what also work are very large emitters close to the object...

... and yes, if you use ibl only, you can go back to normal pet mat...