Why I hear shrieks that have the same pitch as sample rate when the frequency of the signal is very small?


Usually, human can't hear sound with a frequency lower than 20 Hz, but what's confused is, some very slight shrieks generate after I sampled a 1 Hz sine wave with a high sample rate.

To be precise, the shrieks are around the position of peaks and valleys of sine wave, and the frequency of the slight shrieks seems to be equal to the sample rate i.e. a 1 Hz sine wave with a sample rate of 2000 will generate 2000 Hz shrieks. (Remember to regulate volume since the shriek is really slight. )

enter image description here

This question was first posted in mathematica.SE but hasn't got a satisfactory enough answer yet, since this seems to be a "multidisciplinary" issue, I'd like to post it here(not migrate it!) , too. The original question has used some mathematica code, so I've tried to re-express it in a more general way. (I think the original is also understandable even for guys who don't know mathematica, click here to read the original post and answer it if you want!)


Posted 2013-11-09T09:35:29.100

Reputation: 113



Be nice if you can post a sample or way we can hear this - can I run the Mathematica code easily?. 2000hz is really low for smooth audio, are you sure it's not just extra frequencies created by aliasing through the playback rate. In fact, I bet it's that.

edit. I managed to recreate that in Max/msp using an object which reduces the sample rate (degrade~), as you lower the sample rate, so the pitch of the harmonics you're hearing decreases, but all with the same 1hz rate you describe. Here's a video for those curious:

Mark Durham

Posted 2013-11-09T09:35:29.100

Reputation: 4 445

1Just reading the other posts and that seems to be the consensus. It's like you're not playing back a smooth sine wave because of the sample rate, you're actually playing back a stepped sine wave, and it's these steps which create the extra harmonics - because they're like square or sawtooth waves depending on the shape and distance. I think recording the output (at 48khz sample rate) and looking at it in a waveform editor would explain what you are actually hearing. – Mark Durham – 2013-11-09T11:00:01.903

"Can I run the Mathematica code easily?" Yes, if you don't mind to download and install it. (It's free for 30 days, and the Play function is available since version 2 and hasn't change since version 6. )…… Oh, you've managed to recreate it? Then it's not necessary :) , and it'll surely help if you post the video! – xzczd – 2013-11-09T11:44:42.463

Finally I managed to climb over the GFW to watch the video 囧. That's a good demonstration. Then, how to explain the position of the shrieks? Why are they obvious only at peaks and valleys of sine wave? – xzczd – 2013-11-09T13:42:44.300

I don't think I'm hearing it like that, I believe it's on the up and down sections of the wave - where there is more change. It is a bit difficult to see in the video as the sync is slightly off - good old YouTube. – Mark Durham – 2013-11-09T14:29:43.077

Oh, and now I know what GFW stands for - must be so frustrating – Mark Durham – 2013-11-09T14:47:01.227

After watching your video, I came up with an idea to check the position of the shrieks with some mathematica code(I've added it to the post in mathematica.se), and I should admit that I've misheard the position of the shrieks 囧. Thanks for your video!

– xzczd – 2013-11-11T09:16:44.590

Pleased it helped. And thanks for the cross post - we should all do more of that now we're part of Stack Exchange. It's interesting to see different approaches to the same problem from varying fields. – Mark Durham – 2013-11-11T10:39:00.733


I suspect that if the 'shrieks' are harmonics, you'll only hear them when the wave is in a non-zero position, which is why it's audible only when it's positioned up or down. If it's directly in the center, there is no signal to distort; i.e., silence.

Kyle Hughes

Posted 2013-11-09T09:35:29.100

Reputation: 41