Hi everyone.
I have been thinking about creating a pd patch that is a physical model of a singing bowl or a crystal sound bath type bowl for awhile.
I found a paper on the internet about the physical modeling of a singing bowl.

I uploaded a patch trying to use some of the info in the paper. To use the patch you need the else library to use the compress~ and resonator2~ objects.

The paper says the bowl resonated at 186, 551 1007 1627 and 2337 Hz. I tried to scale the amplitude from the paper to something that would work. I looked at Figure 2 in the paper for info on how the different frequencies decayed.

I added a frequency close to each frequency to get some beating. The paper says that due to the imperfections of a singing bowl each frequency has some ringing to achieve that.

The patch I created kind of sounds like a singing bowl.
Does anyone have ideas to improve on to this? I was also thinking to have each frequency go into a bandpass filter so I can control it's bandwidth, but that seems like a lot of extra work and objects.

Thank you for any input.

Nick

OK I looked at some spectral analysis images of singing bowls. I think this one might be better. If anyone has suggestions please share them.

I think one thing that could improve this would be if I could figure out how to emulate a softer attack sound, like with a soft mallet or something similar.

singingBowl2_D.pd

PD-Pi
imp.pd

I think that's a good suggestion. I should work on a different impulse that is more gentle to simulate a mallet.

What does the [1 5, 0 5 20 (. message going into vline mean? Does it go from 0 to 1 in 5 milliseconds, then to 0 from 5 to 20 milliseconds?

Thanks.

PD-Pi

after 5 ms go to 0 in 20ms

... I'm pretty sure it's "after 20 ms go to 0 in 5 ms" in this case though? can't test on the phone here right now, but imho, a message consists of: target value, ramp time, delay time (which is also the inlet order).