How to go about the new SSS option?

[N-drju]

I'm sure glad I could provide some food for thoughts. Use it well.

Due to mainly economical and maintenance reasons, I'm afraid I won't be using the SSS option often. Like you said, faster option that can bring similar or "close enough" results is worth the effort too.

One more idea if I may?

I find Hannes' solution to SSS better, because it gives a lot of control over how the fake subsurface scattering (let's name it FSSS for short) behaves.

The transparency is one thing. But the final FSSS effect can be dramatically different just by how the two objects are located in relation to one another! This is what you see when you compare the middle candle and the rightmost candle.

Which brings me to another point. In the future, you may consider adding some sort of control that will allow a user to apply a 3D SSS map. This way, a user could, for example, control how high or deep the transparent burnout rim of the candle reaches out. I'm not sure if this is already possible. If it is, I'm sorry if I missed the option.

I really appreciate that you try to bring such miracles to this software. However, I also find "TG archeology" to, time and again, achieve results that new technologies occasionally struggle with.

[WAS]

Which brings me to another point. In the future, you may consider adding some sort of control that will allow a user to apply a 3D SSS map. This way, a user could, for example, control how high or deep the transparent burnout rim of the candle reaches out. I'm not sure if this is already possible. If it is, I'm sorry if I missed the option.

I *think*  that can be achieved using a grayscale SSS map as the SSS density function, utilizing the UV interpretive mode.

[WAS]

Coming back to this with an idea. You could also do another method of fake SSS and utilize two materials. Unlit, and lit, and use a merge shader with a "transmission map" that's mapped to UV. Where you want it to be lit basically, and then transition between the two shaders, one being lit of course, with translucency. Similar to transmission maps for SSS in general (that I just posted about).