a question for the physical modellers – struck plates

Feb 22, 2007 at 12:27am

a question for the physical modellers – struck plates

Hi all

this might not be relevant to any here but i thought it worth a shot…

Basically I’m wondering of anyone has seen any papers which have looked at the spectra/energy of the first moments of a struck object such as a plate.

Not the ongoing complex decay information really, just a amplitude/frequency plot of that first reaction, such as something which has mapped out the various modes going on and their relative levels in relation to the fundamental (not so much the “bang” of the strike/impulse).

Make sense? i hope so…

im finding two types of informaiton out there:

high level calcualtions for modal analysis (for the acousticians), and lots of theory on physical modelling… but what im hoping for is just a bit of analysis from an actual (or virutal) struck 2d object

its the basis for a bit of additive synthesis…

directions, papers, examples much appreciated.

thanks all

#30400
Feb 22, 2007 at 12:44am

Hey

I think this will be in the high level physical modelling category, but you might want to look at some of Stefan Bilbao’s work.

http://www.music.ed.ac.uk/Contacts/DrStefanBilbao_publications.htm

Cheers

Jules

#97084
Feb 22, 2007 at 12:50am

heh

yep been mailing him already!

he’s not sure if his work/outputs will be that useful for me unfortunately

amazing work though!

thanks

#97085
Feb 23, 2007 at 7:20am

Eliot Palmer wrote:
> Basically I’m wondering of anyone has seen any papers which have
> looked at the spectra/energy of the first moments of a struck object
> such as a plate.
….
> its the basis for a bit of additive synthesis…

You can feed analysis from Spear directly into oscilator banks, there
was recently some discussion about it. A working program might be more
interesting than a paper about the subject… ;-)

Stefan


Stefan Tiedje————x——-
–_____———–|————–
–(_|_ —-|—–|—–()——-
– _|_)—-|—–()————–
———-()——–www.ccmix.com

#97086
Feb 23, 2007 at 9:51am

Sorry what is spear? sounds interesting…

yeah, im not interested in producing a paper – the aim is some more complex synthesis but im hoping for a spectral (or preferably modal) breakdown from which to start from…

cheers

#97087
Feb 23, 2007 at 1:28pm

i googled the subject and found a ton of information.. i think the keyword that you need to use is “membrane”. try google searches for:

membrane vibration
vibration mode of circular membrane
vibration mode of rectangular membrane

#97088
Feb 23, 2007 at 8:44pm

thanks for the tip

been googling – and getting a ton of (not intended) information, but yet to try with “membrane”. will look into that

#97089
Feb 23, 2007 at 9:57pm

I use the RMI in the IRCAM lib alot, amazing stuff, they have a plate model. I also use the CNMAT resonators~

Spear is amazing.

Not sure if that helps at all. Good luck!

Email me, I think I might have something I could send you.

#97090
Feb 24, 2007 at 1:26pm

Eliot Palmer wrote:
> Sorry what is spear? sounds interesting…

Googles second leads to the home page:

http://www.klingbeil.com/spear/

There are severa threads on the list discussing Spear and the cnmat sdif
externals as well as FTM which both are supposed to read the analysis
Spear spits out…

Can’t add to much of my own experience here, but if you gather some,
please share….

Stefan


Stefan Tiedje————x——-
–_____———–|————–
–(_|_ —-|—–|—–()——-
– _|_)—-|—–()————–
———-()——–www.ccmix.com

#97091
Feb 25, 2007 at 11:11am

yep ive been looking at spear – those CNMAT tutorials are VERY impressive!

like the functionality of resonators, especially if i do manage to get datasets of frequency, amplitude decay triplets from modal analysis.

now this might be a question maxers arent able to help with, but if i am concerned with the phase interactions between a number of overlapping “events” of a membrane stike, wouldnt a process such as that with SPEAR produce quite different results to a modal plot from matlab/calcs and the like? (i am interested in how the energies of this plate, model etc are effected by repeated strikes, whlist the sound of the previous event is still decaying)

the analysis of a composite sound (recorded, analysed, plotted in spear) will be the result of interation between sound from all the modes, so there would be a certain amount of loss of useful data from phase interferences, subtractions and the like?

whereas a modal analysis done from a physics/matlab/calculation point of view would hopefully yield the building blocks from which more complex, long term and varied phase/frequency/energy interactions could be interpolated? (broader than that yielded as the result of the particualr event being analysed)

Does anyone know if this difference is purely acedemic (or finctional)? (or unintelligible?)

thanks again all

mayber there is a more appropriate forum/site out there where such things would be best discussed?

#97092
Feb 25, 2007 at 12:57pm

I think you should really look at the work done by Julius Smith,
Perry Cook, Gary Scavone and Scot Van Duyne from CCRMA. Probably
the best place to start is with the Synthesis ToolKit (STK):

http://ccrma.stanford.edu/software/stk/

It’s implented in [rtcmix~]:

http://rtcmix.org/rtcmix~

and also is featured prominently in the PerCoLAte objects:

http://music.columbia.edu/PeRColate/

Perry’s book “Real Sound Synthesis for Interactive Applications” is
a great intro to all this stuff if you need it.

STK has a mailing list, I think the links to join it are from the
stk page above.

brad

http://music.columbia.edu/~brad

Quoting Eliot Palmer :

>
> yep ive been looking at spear – those CNMAT tutorials are VERY
> impressive!
>
> like the functionality of resonators, especially if i do manage
> to get datasets of frequency, amplitude decay triplets from modal
> analysis.
>
> now this might be a question maxers arent able to help with, but
> if i am concerned with the phase interactions between a number of
> overlapping “events” of a membrane stike, wouldnt a process such
> as that with SPEAR produce quite different results to a modal
> plot from matlab/calcs and the like? (i am interested in how the
> energies of this plate, model etc are effected by repeated
> strikes, whlist the sound of the previous event is still
> decaying)
>
> the analysis of a composite sound (recorded, analysed, plotted in
> spear) will be the result of interation between sound from all
> the modes, so there would be a certain amount of loss of useful
> data from phase interferences, subtractions and the like?
>
> whereas a modal analysis done from a physics/matlab/calculation
> point of view would hopefully yield the building blocks from
> which more complex, long term and varied phase/frequency/energy
> interactions could be interpolated? (broader than that yielded as
> the result of the particualr event being analysed)
>
> Does anyone know if this difference is purely acedemic (or
> finctional)? (or unintelligible?)
>
> thanks again all
>
> mayber there is a more appropriate forum/site out there where
> such things would be best discussed?
>

#97093
Feb 26, 2007 at 10:44am

Thanks brad

i have indeed tried out percolate before and wasn’t able to get the sort of behaviour i was after given my application. But ive changed my approach recently so i might have more joy this time around…

havent checked the book though – thanks for that!

#97094

You must be logged in to reply to this topic.