Join 3,494 readers in helping fund MetaFilter (Hide)


A Good Tone is Hard to Find
February 1, 2014 5:07 PM   Subscribe

Why is the source of 1khz tone hard to track down?

I work in television control rooms, where "Bars and Tone" is a common thing to see and hear. Whenever tone is being played from a speaker in the room, it is very hard to find exactly what speaker it is coming from. This seems to be a problem unique to Tone. If a person's voice is playing from a random speaker, it is easy to track it down and turn it off, but if there is Tone playing, people will wander from speaker to speaker, effectively guessing where it is coming from. Why?
posted by soy_renfield to Media & Arts (5 answers total) 6 users marked this as a favorite
 
Tones in general are hard to track down, in my experience, not just 1khz. Part of the issue is that you don't have multiple frequencies which would have different wavelengths and allow more auditory processing magic to happen and locate the source. Part of the issue is that the tone is consistent (effectively forming standing waves), so decay of the sound isn't something that can help with location.
posted by bfranklin at 5:22 PM on February 1 [3 favorites]


The problem is that a single medium-to-high frequency usually sets up a standing wave, which means there are bright spots and dead spots all over the room. You do directional-hearing by comparing the amplitude between your ears. But when the sound is a standing wave, the amplitude in each ear has nothing whatever to do with how far it is from the sound source.

That's not an issue with the kinds of sounds you hear out in the real world because there's a very wide variety of frequencies, so even if one manages to set up a standing wave, the others won't.
posted by Chocolate Pickle at 5:59 PM on February 1 [2 favorites]


You figure out where a sound is coming from by the slight difference in time between when your left and right ears hear the sound. With a constant waveform, your left and right ears may hear the tone at a different time, but the waveforms in the left and right ears may be in phase, so you have no sense of difference. It takes only a slight head movement to have this happen, so as you are searching for the source, you continually encounter a displacement of phase, then a conjunction of phase, and your orientation is confounded, because your sense of time difference is nullified. Normally a synchronous phase happens when you are directly facing the source of the sound, but when the sound has the symmetry of a continuous tone, it can happen at different head angles.

I'll give you a visual comparison: A couple of years ago I was hiking through a forest, and through the trees I spotted a deer fence in the distance. It consisted of wire squares, so it was very symmetrical. As I approached it, the fence suddenly moved about a hundred feet away from me. It was the weirdest illusion. I soon realized that I had superimposed the wrong groups of squares with my left and right eyes, so the fence appeared to be much closer than it actually was. This could only happen with a symmetrical pattern.
posted by weapons-grade pandemonium at 7:28 PM on February 1 [5 favorites]


High and low frequencies are hard to locate in space. Mid range sounds are easier.

Pure tones without harmonics (i.e. sine waves) are much harder than more complex sounds to locate.

It's because of how our auditory system works. The shape of our two ears (and the head in-between casting a kind of acoustic-shadow) means the harmonics of the sound are filtered differently depending on where it is coming from and we can hear this difference in the differing versions our two ears receive. We rely on this difference (which also includes relative volume, timing and phase information) to locate sounds in space.

However, when you filter a sine wave all you get is a quieter sine wave, so we have less difference information to go on and find it hard to locate.

The easiest sounds to locate are in the mid range (the range the ear is the right size and shape to filter) and contain a lot of frequencies, ideally some white noise, so the auditory system has more information to go on. In order to locate low pitched sounds effectively we'd need huge ears, or tiny ones to locate high pitched ones. I suspect that if you tested large and small members of the animal kingdom you would find that this is indeed the case.
posted by w0mbat at 10:13 PM on February 1 [1 favorite]


There are two primary methods you use for sound location. The first is phase delay. If you picture a sine wave travelling to your ears, you can imagine the peak arriving at one ear while the downward slope of the sine wave is at the other ear. The brain can detect the direction of the sound by the delay between the phases of the sound. This only works for low frequency sound below about 800 Hz because above that frequency, the half-length of the wave, from peak to trough, is shorter than the distance between your two ears. Therefore you can't determine if the slope following the peak is from the same cycle or the next cycle of the wave. So 1000 Hz is too short a wavelength to detect a phase difference that would indicate the origin of the sound.

The second way of determining sound location is by group delay. You can think of this is like the "ping" of a submarine which is a group of waves that has an attack and then a decay. Your ear can measure the delay of the ping first reaching one ear and then the second ear and determine the direction. But a steady 1000 Hz tone has no attack and decay so there is no arrival at first at one ear then at the second ear. Both ears just hear a steady tone.

So the two primary location methods are foiled by a 1000 Hz tone -- it is too high a frequency to detect the phase and too steady of a tone to detect the group delay. Voices or music have a mixture of frequencies and lots of attacks and decays that give you more clues as to direction.
posted by JackFlash at 11:09 PM on February 1 [5 favorites]


« Older There's this one girl friend t...   |  I’m a new fan of a XM radio, h... Newer »

You are not logged in, either login or create an account to post comments