Theres likely as many problems as they want to deal with.
In making a physics model of acoustic instruments you can be as detailed as you want to be, as much as a cpu will let you.
Reaching the best compromise between detail and cpu power is i bet very challenging.
To tell you the truth i was a little shocked when i heard arturia was taking on this one, and a bit relieved when i heard IRCAM was onboard because this is the kind of stuff you need those big research facilities for.
Yamaha previously worked with stanford/cal tech sondius in developing the VL series of physical acoustic modeler hardware synths. Those have panels and panels and panels of parameters that youve probably never heard of before, like throat and other weird things, and every single note can be individually tuned and you literally build an instrument in physics that not only can be played well but also, like a real instrument, can even be played badly.
thats called personality.
In their online demos so far theyve shown their model has a lot of personality and their trumpet is already more trumpety than the sondius model can get, and their trombone has a very nice quality to it.
I'll be totally impressed if in the end it turns out to be a vl70-m light, I'll be in shock and awe if its anything better.
even if it sucks BRAVO to arturia and ircam and yamaha for researching in this complex but lovable synthesis method and i hope history rightly shows this as the big foray into detailed acoustic physics modeling on the pc vsti genesis.
