About the filter

[timothee-haudebourg]

​Hi,

I dived into the source code, and I'm struggling to understand how the low/high pass filter works. I suppose it has something to do with the low, high and band variables in jammer.js, but there are no comments here. Could you explain to me what kind of filter you used, and how it is implemented?

Thanks a lot.

[timothee-haudebourg]

Ok so I've been able to recreate your filter on my own.

These are the filter parameters:

a = cutoff/samplerate
f = 1.5 * sin(PI * a)
q = 1.0 - resonance

And if a name x[i] the input samples, and y[i] the outputs, we have:

low[i] = f * band[i-1]
high[i] = q * (x[i] - band[i-1]) - low[i]
band[i] = f * high[i]

y[i] = low[i]

So it looks like an IIR filter, but in a weird way. Could you explain the purpose of each variable? What is the resonance parameter?

Why is there this 1.5 factor on f? I've plotted the filter's gain for different cutoff values, and it seems that it's a little bit off:

However if you remove this 1.5 and set f = sin(2.0 * PI * a) which makes a bit more sense to me (even tho I don't fully understand whats going on), then you hit the cutoff perfectly: I've tried for several values. Is it intended? The concerned line is the line 255 of jammer.js:

f = 1.5 * Math.sin(f);

I'm trying to understand why this filter works so well. Is that a well known filter, or did you make it up?

Thanks again!

[timothee-haudebourg]

Found it. It's a digital state variable filter.