Not directly related, but do you know of recent achievements with a wow factor in the physical modelling synthesis field [1]? Just as some offline rendered 3d scenes are amazing, It'd be nice to have a pure virtual instrument with its string and resonance chamber model, inter-string effects, plucking mechanism, and so on.
Physical modelling peaked in 1993 with the Yamaha VL1 - which was an amazing synth that crashed right into the uncanny valley between sampling and real instruments.
PM is a complete nightmare to program - even harder than FM. Which was why the VL1 was a preset-only machine. The full programmimng system was never made public.
There are some modern remakes in the box, especially from TAS, but IMO they don't sound as impressive.
The appeal of analog isn't just the sound, it's the conceptual simplicity. Ditto for sampling. Almost anyone with an average-ish IQ can understand the basics.
There are plenty of alternative synthesis techniques, but they're not popular because most people find they're too much effort for too little reward.
PM, modal, spectral, and straight additive can all make some very wild sounds, but - ironically - they don't sound like most people's idea of a synthesizer, so they're used much less often, and usually only by people who are hardcore sound programming nerds.
> The appeal of analog isn't just the sound, it's the conceptual simplicity. Ditto for sampling. Almost anyone with an average-ish IQ can understand the basics.
But software can still the same, right? Albeit with a lot of distractions.
While I agree with you in the context of making synth sounds, I just wanted to note that physical modelling is amazing for real instruments if done right. For example, take MODO bass - none of the sample based solutions come close.
Check out Geoshred[1], developed as a collaboration between Jordan Rudess and researchers at Stanford University.
Its using a physical model for the guitar as well as audio FX (which tend to be easier to get good sounding than acoustic instruments).
In the domain of pure research, Kurt Werner's work in physical models of 808 circuits is also pretty cool [2].
I've used these too (well, Trombone and not Trumpet), and like them a lot. I've mostly played them live with a breath controller, and they're expressive in a very realistic way.
[1] https://en.wikipedia.org/wiki/Physical_modelling_synthesis