What is an "Integrator"?
I'm reading the following article on waveguides and so on:
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?
-Ron.
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?
Thanks Joe
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.
This is my current inspiration:
That's a real hammered dulcimer, not a synth, but it sounds like it could be done with enhanced Karplus-Strong.
-Ron.
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.
In electronics terms
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.
-Ron.
Yay! my electronics have been helpful!
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"):
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
Leakiest Integrator
http://en.wikipedia.org/wiki/Water_integrator