I notice that different types of sand sound different - the timbre of the sand falling onto the table and shuffling around is noticeably different between the various types. I was curious if the differences in the sand were being physically modeled to synthesize sound.

It’s a physics-based generation but the simulation model is somewhat limited; essentially a few parameters that vary between the sand types, some randomization, and responsiveness to flow rate. Still, it’s highly effective; something like the fine, silty desert sand is noticeably brighter and sounds more like a hiss as it pours and flows than the rocky volcanic sand.

Very cool effect, this is joyful to play with.