[ODE] Contact sound generation

Tyler Streeter tylerstreeter at gmail.com
Thu Nov 4 09:25:08 MST 2004


On a somewhat-related note, this paper talks about generating sound
effects automatically from a rigid-body simulation.  He uses something
called modal analysis (which requires tesselating each object into a
mesh of tetrahedra and calculating some special matrices per object
beforehand) to synthesize sounds on the fly (i.e. no pre-recorded
sound files at all).  I saw this talk at the GDC 2004 which was mainly
about deformation and fracture using the same technique.  It was
pretty impressive.

http://www.gdconf.com/archives/2004/obrien_james_04.pdf

Tyler

On Thu, 4 Nov 2004 12:04:59 -0000, Bob Dowland <bob.dowland at blue52.co.uk> wrote:
> might there be some way to tie in surface/structural props..?
> 
> ie friction and restitution coefficients - I've not had a chance to try this yet but it occurred to me that one might be able to set up some correspondence between restitution and "loudness" on the basis of some very simple model like "the energy that is disipated should become sound", maybe friction coeff could also be making a contribution to frequency say... one could perhaps increase the realism (in the direction of a decent approx) by having some representation of the bodies material makeup, I'm thinking along the lines of things like natural frequency / resonance etc..
> 
> what makes a sound other than frequency and loudness..? could a sound signature be concocted on the fly like this?
> 
> 
> 
> 
> > -----Original Message-----
> > From: Jon Watte [mailto:hplus-ode at mindcontrol.org]
> > Sent: 04 November 2004 01:20
> > To: Megan Fox
> > Cc: Will Thimbleby; ode at q12.org
> > Subject: RE: [ODE] Contact sound generation
> >
> >
> >
> > But two boxes sliding against each other would still make two sounds
> > for the one contact, whereas a box sliding on gravel and on metal at
> > the same time would only make one sound.
> >
> > Thus, I prefer keeping a hash table of <object,object> tuples and
> > keep object pair sound data in that table. When objects stop making
> > noise, they go away from this table so that repeated bouncing can
> > still make repeated sounds.
> >
> > Anyway, both ways will work :-)
> >
> > Cheers,
> >
> >                       / h+
> >
> >
> > -----Original Message-----
> > From: Megan Fox [mailto:shalinor at gmail.com]
> > Sent: Wednesday, November 03, 2004 3:22 PM
> > To: Jon Watte
> > Cc: Will Thimbleby; ode at q12.org
> > Subject: Re: [ODE] Contact sound generation
> >
> >
> > To get around the multiple sound problem (6 different contacts per
> > object per scene, or whatever), as well as the stacking sounds problem
> > (one at full volume, then the next frame, one at slightly less volume,
> > etc - spirals out quickly), you might consider sound states rather
> > than just generating sounds.
> >
> >
> > _______________________________________________
> > ODE mailing list
> > ODE at q12.org
> > http://q12.org/mailman/listinfo/ode
> >
> 
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>


More information about the ODE mailing list