[Thomas An.]

Last one looks very good ...

About the noise ... what benchmark do you get ?
Usually with images like that, having SSS and dielectric noise, you should expect level 21 or higher (I let them go to sl23 on a fast machine).

Well, my machine couldn't be said Fast...it's...uhmm 8.8. Maybe because i am running MW under mac OSX, since i need to use the computer with other apps during the rendering process.

Yeah, that benchmark is low (but SSS is known to be demanding) ... if there is nothing you can do to simplify the surroundings of the scene (make the ground plane as small as possible, remove any unnecessary reflector planes.. etc), then you are stuck ... and just have to wait for SL21 - SL23 .

[Mattia Sullini]

I'm exhausted...and my mac too. So here it is the final version. SL 20.10 (48h more to reach SL 21.00......). Noise reduction and little color correction in photoshop. (i guess nobody cares about rendering times, since those 90h depend by my low specs)
Anyway Thomas, i guess that even if i had waited until SL 23 the noise coming from SSS would have not disappear, since those parts were still very noisy.

[Mihai]

Are you using additive mode in your sss material? That can take a really long time to clear up, if you have 2 bsdfs for example, material is set to additive, one of the bsdfs has an sss component, and the weighting is set to something like 100/100, or 100/50. If you do use additive mode with sss, try to keep the weighting of the bsdf without the sss component fairly small, like 15.

This is anyway what I have observed, when using sss and normal blending, the material seems to clear up faster.

[Mattia Sullini]

Really useful tip Mihai! You are right, i took the very bad habit to use additive mode in layer's blending. I will try to follow your suggestion, even if i can't set the weight of the first BSDF (that controls transmittance) to such a low value since i need a fully saturated base color for what seen trough transparency. Otherwise i get a result very similar to the first pic. Anyway, since i am a novice for what SSS is concerned, do you think i could get a similar result letting SSS become dominant?

[Mihai]

For the material you're trying to make, I would use one sss lambert bsdf, and one glass type bsdf, mix the two, perhaps even making the glass one dominant. Perhaps this is what you've done already.

It looks like you're trying to make a gumy bear like material?

[Mattia Sullini]

Right again 1: this is what i did!
Right again 2: Yess, even if here in Italy you can find under crocodile form as well...i love to sgnack off their little heads....