rpieket

I'm reading the following article on waveguides and so on:

http://lib.tkk.fi/Diss/2000/isbn9512251965/article1.pdf

I can translate the diagrams fairly easily into Max/MSP patchers, except for one detail. Just before the output stage in diagrams 4 and 5, there is a block labeled "Integrator". What could that be? What is the equivalent of an "Integrator" in Max/MSP?

what-is-an-integrator

I'm not sure of the object, but this is a vector sum, no? IE calculus integration. It should be sum~ or something - sorry, my msp is not as strong as my max.

Could you post the patchers do you think?

I know what integration means in calculus. But I'm not sure what it means in sound processing. Thinking out loud: if function f(x) is the waveform of the string simulation, then x means time, and the integral would be the sum of the signal over a certain time interval - like the sum of all samples in a short delay line. Sounds like a low pass filter.

I will post the entire instrument when it is in a more complete state. I'm still very new at Max/MSP, and don't have a ton of time to dedicate to my hobby, so I'm not moving very fast.

http://www.youtube.com/watch?v=XCV2xNumB7Q

That's a real hammered dulcimer, not a synth, but it sounds like it could be done with enhanced Karplus-Strong.

I don't think that's right. AFAIK X is not time - X is distance from origin, ie instantaneous power. Vector integral (in this case) is vector sum instantaneous distance from origin for both signals (phasors anyone?) over given time period.

http://www.electronics-tutorials.ws/opamp/opamp_6.html

Sorry if I am unclear/erroneous - my math and electronics ain't where they gots ta be.

Thanks! My math skillz are poor, and I did electronics... oh gosh 35+ years ago. But I recognized the opamp circuit immediately! Major flash-back. The number 741 comes up for some reason... :)

I believe this circuit is also called a "slew limiter". And I think a onepole~ could stand in for that.

it says in the pdf that the integrator can be approximated by a first-order recursive low pass filter whose cut-off frequency is less than the lowest fundamental to be synthesized.

Maybe they really mean a leaky integrator, which is in fact a differentiator... to remove the dc buildup...

*shrug* IMHO audio ppl who know formal signal processing are rarer and more precious than rubies.

I'm trying to be the exception to the rule... still working on it...

here's a summary of the maths behind it (I googled "leaky integrator"):

http://www.claysturner.com/dsp/Leaky%20Integrator.pdf

its got a formula which would be easy to implement in max btw

EDIT: it's also evident from the formula that in my previous post saying a leaky integrator is really a differentiator is wrong

http://en.wikipedia.org/wiki/Water_integrator

Leakiest Integrator
http://en.wikipedia.org/wiki/Water_integrator


What is an "Integrator"?