I believe I understand that we determine the direction of a sound source by virtue of some automatic calculations based on the difference in strength and timing of the sound waves entering each ear. But how do we have the sense that a sound source is near or far? Is that related to surrounding accoustics? If I lie in bed next to my radio, it sounds near to me. But if I’m across a room from it, I can tell that, too. Is my hearing system also processing ambient acoustical signals and concluding that the sound is coming from farther away? Do I automatically include faint echoes and reverberation in that calculation? How else could I detect the distance?
Good question, but are you sure you can tell the difference between a radio next to your bed and a (much louder) radio across the room, with your eyes closed and no previous knowledge of the location of the radio?
Sound waves at lower frequencies are attenuated less over distance than higher frequency sound waves, so if you hear someone’s voice from far away, it’ll sound different-- as thought the lower frequencies had been amplified relative to the higher ones. Thus, more distant sounds are going to sound distorted, if they contain a range of frequencies. I’d be surprised if you could tell how far away a single frequency source is, though.
It also depends on whether you’ve heard the sound before (and thus have an idea of how loud it’s supposed to be. I’m sure we’ve all heard those weird squeals or high pitched noises where you’re not sure whether it’s inside your car/home or outside, because you have no reference point.
That’s a good point. Awareness of the situation and points of reference are certainly two ways that we determine distance. But what about the sense that a sound is being made right near us, vs the sense that a sound came from some distance away. How do we do that?
Interestingly we often can’t. That’s most true with monotonic sounds. You can prove this for youself by going outside on a warm night and trying guess show far away a cricket is. It’s damn near impossible, even with practice. You can narrow it down to 5’ vs 50’ but trying to distinuish 5’ from 10’ is nearly impossible even with practice. All you can do is move closer and closer until it stops, and they are usually much further way than you estimate.
It’s been along time since I covered this, but basically the ear can detect the distance of sounds in two ways.
The first is by simple triangulation. You have two ears, one on either side of your head. By comparing when sound arrives at each ear the brain can work out not just direction but distance, just as the eyes can. That is assisted by moving the head side to side, an unconscious reaction when trying to pinpoint sound. For humans that will give us gross readings on the distance to a sound source.
The other way that we can pinpoint sound is by the shape of the echoes within the external ear itself. All those little ridges and curves in the ear allow us to pinpoint the direction of sound with astounding accuracy. What that means is that if we move the ear we can also pinpoint the distance to the sound, since obviously we can keep track of the movement of the ear and thus any change in apparent direction must related the distance.
You can probably try this experiment yourself right now. Cover one ear and concentrate on the sound of your computer fan or other nearby noise. Now move your head side to side and back and fowards. You wll notice that the pitch and volume of the noise change noticably. The brain is able to convert those changes into distance meaurements. As a result even people with one working ear can judge distance to sound sources.
Why monotonic sounds confuse us is the lack of reference points. As John pointed out, different frequencies attenuate at different rates. With somehting as complex as a human voice we can judge distance by comparing these frequencies. If the highest frequencies aren’t vanishing any faster than the lowest as we move our heads then the object must be failrly close. If they are become inaudible with just a slight turn of the head then the source must be along way away. With something producing just a single frequency we lack these distance cues and so we have to rely purely on triangulation to try to judge distance, and that’s not terribly accurate.
Does moving your head “side to side” mean actually shifting the location of your head, and not just turning it about the axis of your neck? If it does, then yes, that will help. Just rotating your head, though, will only confirm the direction, not the distance. You’d get a better reading by actually walking around, which is what I think we do in cases where we’re confused about the distance.
There’s a lot of brain processing going on as well. We’re best at judging the distance and direction of familiar sounds, like voices, music or traffic, and the way that distance filters out higher frequencies gives the impression of distance when subconciously referred to a normal, closer sound of the same type. Filtering effects allow you to locate a sound source from behind you too.
Didn’t know about the ear bumps thing; cheers Blake, that’s quite fascinating. Good point about the directional vagueness of monotonic sounds, and IIRC there’s one particular frequency (3kHz?) that is notoriously difficult to pinpoint due to it lying on the crossover region between two different parts of the cochlea. That’s about the same as the resonant frequency of the auditory canal as well. 3kHz sirens are really annoying.
People will actually crane their necks when they are trying to locate a sound, as will most animals. Not just rotating but actually sticking it it and up as well. And as you say, we walk around, shift from foot to foot and so forth. Anything to increase the effective base betwen the ears.
I liked the rest of your post, but I don’t believe this particular sentence. From my armchair understanding of the workings of the ear and brain, it’s just not wired to measure the timing of events between the two ears. Your brain could do it in software, but its response time is much longer than the fraction of a millisecond difference in the time your ears hear the same sound. No, the direction-finding is by processing the subtle resonances of the outer ear, especially when you turn and shift your head.
As I said, it;s been a while, but that was the accepted wisdom when I studied physiology 13 or so years ago. Uf you have more up to date knowledge then I won’t argue the point.
I don’t follow this at all. Triangulaton doesn’t rely on the time differences of the input, it relies on the difference of angles. The brain performs triangulation calculations for visual input constantly, and the time difference for light arriving at the two eyes is cnsiderably less than the millisecond range. It’s simply not an issue because it’s the angle that you need to triangulate a signal, not a time delay.
On rereading it appears the confusion was due to my sloppy writing.
Would make more sense if it said
Gack. My bad.
It’s the difference in amplitudes not the difference in timing that determines the distance from the sound source. For example if someone sitting immediately to my left was speaking, there would be a noticable difference in how loud they sounded from my left ear and my right ear. Then if you recorded their voice and played it back to me via a loud speaker from a distance of a hundred feet away, the voice might be the exact same amplitude when it reached my left ear. But at that distance the amplitude would also be almost the exact same when it reached my right ear. Subconsciously, my brain would notice that there was no apparent difference in amplitudes between my two ears and would figure out this means the source of the noise is distant enough for the distance between my ears to be irrelevant.
As an example of this, consider how a mosquito buzzing an inch from one ear can sound very loud in that ear. But it will be almost silent to the other ear. Your brain notices this and realizes the mosquito is actually only an inch away from one ear.
I have a lot of hearing loss from flying for so many years and my left ear is worse than my right. Sudden sounds that are not repeated are very hard for me to determine their location.
I need more than one hit to get the location down at all unless there is some component that makes me have a good idea where the sound is.
A car door slam is not likely to be in the house.
A road noise that is not in front of the house can only be from a quarter mile in back or from across the canyon 1½ miles away. So that helps to determine where the sound is because my brain knows the likely place for it to originate.
I think this is the answer I was seeking. As for identifying direction, I would guess that the differential between the ears that is calculated by the processing centers would take in both amplitude and timing. And, maybe, if the sound is being made close enough, there will also be some involvement of the acoustics, including those of the external ear. This is very informative.
If this were true, a person would be unable to determine the distance of a sound if it was anywhere on the plane midway bewteen the ears and perpendicular to the axis between the ears (ie, directly in front of you, directly overheard, etc). Is that the case?
On one of the Royal Institution Christmas Lectures, I saw an experiment that suggested that sound timing differences are indeed used to determine direction.
A C-shaped loop of plastic hose about a yard in length was capped with ear pieces, and these were placed into the ears of a blindfolded kid from the audience. The hose was tapped to one side of its midpoint or the other, and the kid raised his hand according to which side he thought the sound was coming from. He could reliably distinguish taps only an inch to one side of the midpoint.
Page 9-8 of this pdf states that human ears can register timing differences of 10/1000000 seconds, which I find quite staggering.
http://www.med.uwo.ca/physiology/courses/sensesweb/L9Auditory/L9Auditory.pdf
On one of the Royal Institution Christmas Lectures, I saw an experiment that suggested that sound timing differences are indeed used to determine direction.
A C-shaped loop of plastic hose about a yard in length was capped with ear pieces, and these were placed into the ears of a blindfolded kid from the audience. The hose was tapped to one side of its midpoint or the other, and the kid raised his hand according to which side he thought the sound was coming from. He could reliably distinguish taps only an inch to one side of the midpoint.
Page 9-8 of this pdf states that human ears can register timing differences of 10/1000000 seconds, which I find quite staggering.
http://www.med.uwo.ca/physiology/courses/sensesweb/L9Auditory/L9Auditory.pdf
And another, somewhat more rigourous cite:
http://www.aip.org/pt/nov99/locsound.html
Hm. Posted twice before I’d finished composing the thing properly. Anyway, from my second cite:
“The idea that this sensitivity is obtained from an ITD (interaural time difference)initially seems rather outrageous. A 1° difference in azimuth corresponds to an ITD of only 13 ms. It hardly seems possible that a neural system, with synaptic delays on the order of a millisecond, could successfully encode such small time differences. However, the auditory system, unaware of such mathematical niceties, goes ahead and does it anyway. This ability can be proved in headphone experiments, in which the ITD can be presented independently of the ILD (interaural level difference). The key to the brain’s success in this case is parallel processing. The binaural system apparently beats the unfavorable timing dilemma by transmitting timing information through many neurons. Estimates of the number of neurons required, based on statistical decision theory, have ranged from 6 to 40 for each one-third-octave frequency band.”
Now I’m done, I promise!
I knew were were good but I didn’t know we were that good…
Makes dogs damn good.
…of course, dogs can turn their external ears to get a better fix on the source of the sound, too. With higher sensitivity to a wider range of frequencies, they may just be better built all around for sound detection.