Popping sounds from oscillator volume change

[heliocentric]

Hey ya'll. Been using ChucK for a few months, making simple programs which manipulate oscillators. One problem I've come across is when I change the volume of an oscillator, I get a popping sound. Using envelopes can help, but the faster I change the volume, the more difficult it becomes to eradicate the popping. In fact it's always kind of there in the background. What's more, I made my own envelopes to troubleshoot the problem, to get a better sense of what's happening, and I found that, somehow, changing the volume of an oscillator was affecting the frequency!
Basically what I'm doing is switching an oscillator's volume between 0.0 and 1.0, with the frequency staying the same. I feel like there's something here I am not grasping. I've tried changing speakers and different computers, no change. Just so you know, I'm a noob to electronic music. Thanks!

[Antimon]

Hi and

Would it be possible to post a code example that shows the problem here? I don't think it should behave the way you describe it, especially if you use envelopes.

[heliocentric]

Thanks for responding!
Here is some code to show the basic problem:

[varice]

[heliocentric]

Wow thanks Varice! Very informative, exactly the kind of info I am looking for. Whenever I searched the net I could only find info about clipping, as I guess this is a much more common noise problem. I've never heard of a transient before, but yes, it looks like that's what's happening here. It's nice to know that this is a general acoustic phenomenon, and not just something to do with ChucK. Again thanks, this has been bothering for a few weeks, really driving me nuts.

Well, I still have the problem of how to fix this. It's nice that using an envelope reduces the problem, but I still get the popping when the envelopes duration is set to 0.001 seconds or less. Granted, this is pretty fast, but in my mind I think "computers can solve any technical problem." I want to go as fast as possible!

@ Antimon: do you think there is anything I can do to make my volume transitions even faster while still avoiding the popping? Oh and thanks for the warm welcome! First forum I have ever signed up to!

[heliocentric]

One thing I would like to understand better is the 'rate' method for the Envelope unit generator. I tried chucking different values to this method, and I can't tell what it does. I took the class on ChucK on Coursera, and I've looked through the ChucK manual and online, but all I can glean from this is that the method takes a float. Perhaps this this method is the key to my problem?

[blue hell]

This is not something a computer can fix .. a fast volume change just /is/ a plop or click ... in fact no device can fix it ... and it is not a problem either ... its just how nature works ...

[Antimon]

Yeah what Jan said.

In fact, when I look at and listen to your first piece of code, what you're doing is changing the gain of an oscillator between complete silence and full volume with a rate of 20 Hz (once every 0.05 seconds), producing a fluttering sound. It sounds a bit like a helicopter, which is about what I'd expect. Modulating something wildly on/off like this can't be expected to produce any kind of "smooth" sound as I see it.

Might I ask what you're trying to do: what kind of sound are you trying to accomplish?

[Antimon]

[fretless]

I suppose the 1 sample loop is a bit of a problem.


class ZCrossGain extends Chubgraph {
inlet => Gain g => outlet;

static float oldGain;
0.0 => oldGain;
0.0 => g.gain;

fun void setZCrossGain(float newGain) {
float theRange, theDiv;
0.001 => theRange;
theRange * -1.0 => float minusRange;

if (newGain != oldGain) {
(newGain - oldGain)/16.0 => theDiv;
for (0 => int i; i < 16; i++){
while ((outlet.last() < minusRange ) || (outlet.last() > theRange) ) 1::samp => now;
g.gain() + theDiv => g.gain; 1::samp => now;
}
g.gain() => oldGain;
<<<newGain, g.gain()>>>;
}
}
}

SinOsc oscillator => ZCrossGain eee => dac;
1.0 => oscillator.gain;
300.0 => oscillator.freq;

0.05 => float timing_variable;

while(true)
{
Math.random2f(0.1, 0.9) => eee.setZCrossGain;
timing_variable::second => now;
}

---

ChangeGain.ck
Description:		Download
Filename:	ChangeGain.ck
Filesize:	960 Bytes
Downloaded:	535 Time(s)

---

[fretless]

It looks like SqrOsc and SawOsc get stuck in the 1 sample loop, so this seems only useful for SinOsc and TriOsc. I think the STK instruments that take an intensity parameter for noteOn and similar don't have the clicking problem.

[heliocentric]

@ Antimon + fretless: Wow! I'm really in over my head now!

Thanks for your replies! I'm going to examine both sets of code you have given me and attempt to understand them better, but I'm afraid I'm far too ignorant at this point to really grasp what's going on with them.

I took the coursera course on ChucK last october - december, and basically I didn't know almost anything about programming or electronic music prior to this, so I obviously have a lot to learn. Thank you both for taking the time to consider this problem though, I'm really appreciative!

@ Antimon: the reason this problem came up in the first place was because I read that every sound can be understood as a collection of sine waves. So I thought to myself: "Oh, so if I just make programs that manipulate the frequency, amplitude and temporal patterns of sine waves, I should, theoretically, be able to make any sound." This has been my big programming project since the beginning of the year, my first real programming project. Obviously being able to make any sound at all is way too ambitious, but I was just curious to see where the project would take me. I was very surprised when I came across this popping/clicking thing. In my mind I was like: "Hmm, shouldn't I just hear the frequency of the sine wave at different amplitudes, or just becoming silent?" What I wanted was simply to hear a single pure tone changing in amplitude, and then make programs where I could have a number of oscillators all changing the amplitudes of sine waves. Anyway, this probably sounds all a bit naive, but that's what brought me to this problem. I love the idea of just building sounds from the ground up, one sine wave at a time, and thus having a deep understanding of how the sound is structured, especially since I have programmed how the sound plays using just oscillators. Heh.

[varice]

[Antimon]

[heliocentric]

@ varice: sorry, you're right, using an envelope is a really good solution.

@ Antimon: Ahh, I see. Thanks, that also helps me understand this much more. You've been a big help!