Physical Modeling in Kontakt! (UPDATED audio example)

[bioLogic]

Wow, very cool. Like others said, it will be very interesting to see where this goes. Especially considering what you stated about achieving near-perfect realism with SE... I also agree that the most important thing (in the final analysis) is whether the sound is realistic--even if certain elements of the model don't correspond to real world parameters.

Physical modeling meets sonic modeling.

[Gregjazz]

There are some elements that are going to be difficult to model. For example, getting just the right realistic release to the notes. Modeling the pure sustained tones are one thing, but there are a lot of complex elements happening to get some of the extraneous sounds.

One thing that occurred to me is that it wouldn't be too hard to sample a guitar or bass with a special kind of pickup and then be able to dynamically change the pickup(s) position and type, like the Line6 Variax can do.

[dynamitec]

Hi Greg,

Check this out,
http://www.till.com/articles/PickupMixing/index.html

especially the link to the Roland VG8 patents. It describes mathematically what you wrote here.

Best,
Benjamin

[runagate]

Very impressive once again, Gregjazz!

Since it starts with an initial impulse, that means that one could use a variety of attack transients in your model.

Such as a bass guitar emulation with pick, finger and slap.
Since the attack transients aren't pitched and have little to do with the rest of the envelope of an excited string anyways.

[Gregjazz]

Check this out,
http://www.till.com/articles/PickupMixing/index.html

especially the link to the Roland VG8 patents. It describes mathematically what you wrote here.

Hi Benjamin,

Interesting link, thanks for sending it. I'll have to read this in more detail when I get a chance.

Thanks,

Greg

[EvilDragon]

So wait, THE ONLY sample that was used was impulse? Everything else was done with Kontakt's FX and some scripting? I'm interested HOW.

[Gregjazz]

Well, the impulse is essentially the shape of the initial wave on the string (well, two waves, as they go apart from the location the string was plucked). Each time a wave hits either end of the string, it gets slowly muted as well as inverted and reversed. Since the bridge has a different amount of muting than the fret (or finger, if it's fretless), that causes cancellation in the frequencies throughout the decay of the tone--which is what's really happening when you hear individual harmonics fading in and out during the decay of the string.

[EvilDragon]

And THAT was reproduced in Kontakt... HOW? Filters? AET?

[Gregjazz]

This is compatible with K2, so no AET (though that certainly offers more possibilities), but instead a simple LP filter and lots of scripting.

[EvilDragon]

Cool, very cool