Yes, I just improvised on the noise, bp filtering and strike velocity, as well as on other factors. I certainly agree that more than 12 frequencies should be present for realism; I believe I read that a real bass drum can have up to 150 identifiable partials, so I doubt the validity of the small number of partials the article cites. Your suggestion on frequency to amplitude proportionality sounds like it is worth a try. I think I will revisit the instrument from a fresh perspective and start again from scratch. Probably the only really usable info there is the freq = .766/diameter*sqrt(tension/density) equation, if I can only figure out the right way to use it. Thank you for your response!


On Sun, Jul 4, 2010 at 4:22 PM, Mark Van Peteghem <> wrote:
The article doesn't mention how the envelopes of each frequency varies
over time, or what the ratios of the amplitudes for different
frequencies are, so it is not easy to get it right. The article doesn't
mention noise or bandpass filtering, neither strike velocity. To get it
better I would stick to simple sines, not the other waveforms, have more
frequencies (the article mentions up to 12) and have amplitudes for each
frequency. Or maybe simply making the amplitude inversely proportional
to the frequency (or the square of the frequency) is enough, that often
makes it sound more realistic.

Brian Wong wrote:
> I just uploaded a drum synth to blueshare called CMDrum-01. It is
> roughly based on the physics of circular membranes which I found on
> this website :
> I am not a physicist or mathematician so my implementation is rather
> flaky, and I would much appreciate any improvements or suggestions
> others might have.
> Brian Wong

This email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit --
Bluemusic-users mailing list