insane physmod unit 'Pianoid'

[magronbass]

has this been brought up here before? Sorry if it has.

 

An amazing hardware unit that does physmod like no other. They do physmod through a bunch of differential equations.. and don't even bother asking me what that means. lol

 

http://www.pianoid.com/teamblog/?p=29#comments

 

Check out the newest demo. Very amazing if you ask me. It only uses 400 times more calculations that possible on a normal cpu! I really hope these guys find an investor....

 

Their room sim is even driven by these differential equations... I wonder if differential equations can solve relationship problems and also bring peace to the world. (and lower gas prices)

[mate stubb]

These are those guys from Russia that showed up on that other forum, right?

 

They will bear keeping an eye on.

[Mark Schmieder]

They sure don't say much. I guess there are audio demos we're supposed to listen to and that's the whole story. Maybe I'll remember to revisit the site later tonight at home. But I gather they're in very early stages, and only focusing on piano vs. other instruments. They must feel they have something to offer that Pianoteq doesn't, or why bother (they need to make money).

[magronbass]

well, their modeling theory is a different one (says the pianoid ppl).. actually the physmod piano research at purdue and stanford lead me to them. They use a different approach which is computationally extremely intensive, but more detailed.. Their special processors allows them to get the calculations done in real time..

 

I did find their "product" site.. which also doesn't say much.. and has bad sounding demos.. I guess they're updating the sim engine very rapidly..

 

http://www.pianoid.com/

 

But this thing actually sounds right, it sounds very, very good. Their latest demo only, though.. all their other demos sounds like they were just testing out a specific element of the sim.. or in other words, not so good.

 

I hope they atleast get to some sort of mass production phase..

[Mark Schmieder]

I see what you mean -- only the mid-range sounds like a piano. The low end sounds like... I don't know, maybe a distorted bass guitar overlaid with a DX7? The high end sounds a bit like a Glockenspiel overlayed with a Celesta.

 

Actually, it sounds pretty cool... just not like a piano :-). It sounds like a trio, with the piano covering the mid-range.

[magronbass]

ya, the older demos sounds like a broken sound blaster.. you did check out the newest one from their blog? That's the newest update I'm guessing. That one I think is pretty darn good.

[Mark Schmieder]

Is that the one at the top of the page using a Rachmaninov MIDI file? That was the only one remotely listenable.

[Tusker]

It's coming closer. The old demos remind me vaguely of the RMI pianos from the 70s.

 

I have to admit, I am not hearing a whole of lot of sonic detail in the models, including the new demo. The character of the sounds appears to be more derived from excitation than resonation. I am hoping these can be traded off in the model they eventually market.

 

Wishing them every success.

 

Jerry

[marino]

The latest demo sounds good to me, considering it's mono (from what I can hear) and completely dry. The previous ones are laughable.

[magronbass]

ya, their older demos are.. I don't even know what it is..

[bloodyMary]

More physical modelling products, That's great. I hope they can get it right!

 

By the way, I guess they use differential equasions to process string movement, and maybe diff. equasion stuff to process wave movement in the air. Add impulse response-driven body resonance (think convolution reverb) and there's great potential. It's possible to simulate strings affecting each other, too (when one string vibrates, it causes neighbour strings to vibrate a little, too). Something impossible with sampling!

 

By the way, I've tried pianoteq yesterday, and it sounded better than any sampled one to me. It had life in it.