Why isn't there a sound equivalent of the one-way mirror?
July 30, 2007 1:04 PM   Subscribe

I know that one-way mirrors aren't really one-way, but why isn't there an equivalent to this general concept when dealing with sound?
Why is it that you can't have a box which lets sound come out, but blocks it from going in?
I await your elucidation.
posted by dunstanorchard to Science & Nature (18 answers total) 2 users marked this as a favorite
 
Probably because light acts as both a wave and a particle but sound only acts a wave. A reflective surface will force photons to bounce off (like a particle). Sound is phenomenon that is a wave and cannot be deflected by such a simple method.

You can make a box that works like thus by using a microphone and a speaker between sound proof glass.
posted by damn dirty ape at 1:11 PM on July 30, 2007 [1 favorite]


Apparently, you can.

here's a google search for one way sound...

First link is to a type of insulation that attempts to achieve this.
posted by twiggy at 1:13 PM on July 30, 2007


I suppose most music halls are like this - the audience can hear the performers, even unamplified, but performers can rarely/barely hear the audience.
posted by GuyZero at 1:19 PM on July 30, 2007


Sound is the result of vibration of some sort -- vibration of materials (strings, drum skins, etc) turning into vibrating waves of air molecules. This energy is transmitted from material to material (hit something sitting on a table, and the entire table will vibrate).

In other words, anything you use as an encasement around a sound-creating device will vibrate ... and sound emitting from outside the encasement will vibrate the encasement, too.

You can dampen vibrations (with foam, for example), or shape the acoustics with material structures. But you cannot prevent it from happening, and you cannot prevent it from transmitting to all materials in all directions. There is no ability to "go one way and not the other" in terms of dampening vibration.
posted by Cool Papa Bell at 1:29 PM on July 30, 2007


@twiggy Those search results don't really say anything, the first one is rather mislabeled, and the others aren't relevant. But thank you for the thought.
posted by dunstanorchard at 1:34 PM on July 30, 2007


That one-way sound Google link above has lots of results for dampening materials and structures. That's useful for dampening sound from one location generating sound to another -- say, preventing the guy located next to the recording studio from being bothered by the racket. But it will also dampen sound being generated in the opposite direction, as well, but less effectively.

Even the very best recording studios have occasional trouble with trucks driving past the building, planes flying overhead, etc.
posted by Cool Papa Bell at 1:36 PM on July 30, 2007


imagine a box with something very noisy in it, like a big speaker playing loud music. the box is a good but not perfect sound dampening box. so, if you're outside the box you can still hear the racket coming from inside. on the other hand, inside the box its very loud, and you will have a hard time hearing quiet sounds coming from the room outside the box. voila, one-way sound mirror.

the answer doesnt really have to do with the wave-particle duality, or whether the waves are longitudinal or transverse or whatever.

i'm pretty sure it's got more to do with the way human sensory organs perceive things. our eyes and ears can perceive sound and light changes over many orders of magnitude, and the mental signal-processing machinery dynamically adjusts our notion of what is loud and quiet, or bright and dark, depending on what the ambient or average level is.

the criminal suspect in a brightly-lit room has pupils that adjust to the ambient light levels; this means that small changes in light intensity, such as those coming from behind the one-way mirror, are negligible and therefore difficult to observe. on the other hand it's dark in the observation room so the relatively large variations in the light that makes it through the one-way mirror are easily detectable.

from a physics point of view, i can't think of *any* way of making an actual one-way mirror, even using fancy polarization rotation tricks. (maybe a better optician will come in here and correct me... maybe you could do it with a brewsters-angle thing, but even then it would only be one-way under very speific circumstances.)
posted by sergeant sandwich at 2:10 PM on July 30, 2007


There are a couple reasons for this, touched on in the above answers.

Part of it has to do with how large the range is in terms of things you can perceive. The "dynamic range," if you will, of vision is extremely high. Your healthy human retina, properly dark-adapted, can reliably detect a single photon. However, your brain doesn't tell you how many photons you're seeing. The single photon emerging out of darkness looks, to you, like a brief flash; a firework detonating in the sky giving off bazillions of photons also looks like a brief flash. In other words, your visual apparatus is not set up to judge the absolute intensity of a visual stimulus.

So when you're sitting in the dark room behind the half-silvered mirror, receiving half the amount of light you normally would, your brain compensates. You don't realize that you're only seeing half the number of photons as you would if the mirror were not there.

In contrast, if half the sound energy of an audible stimulus is attenuated, you'll notice. Your ear contains only a very rudimentary way of attenuating loud inputs, resulting in only a slight decrease in perceived loudness (when the tensor tympani and stapedius contract.) So even if you were to construct a material that reflected half the sound energy that struck it, and transmitted half of it unchanged, a person sitting behind that window would notice that the sound levels were much lower, possibly inaudible.

Finally, plain physics make it hard to construct such a sound-mirror that only works one-way. Sound is waves in a particulate medium, unlike light; that means that sound propagation occurs by means of elastic collisions of molecules. How are you going to tell the molecules that when they strike side A of the sonic mirror, that their energies should be absorbed and dampened, or re-reflected out of phase; but when they strike side B, the mirror should vibrate in such a way as to pass those energies through, vibrating the air on side A with fidelity in terms of amplitude, frequency and phase?

What kind of material possesses that property? None, as far as I know.

Now since you realize that one-way mirrors aren't really one way, you understand that whichever side of the one-way mirror is darker is the side where you can sit and see through to the other side.

Consider two identical rooms that share a thickly insulated wall. Suppose on one side sits a person, Joe, whispering to himself. On the other side the Who are playing a concert at 120 dB. Joe can hear Pete Townshend, but Pete Townshend cannot hear Joe.

Now have Joe and the rock band switch rooms and keep doing what they were doing. Joe can still hear Pete Townshend, but Pete Townshend still can't hear Joe. This is the analogous situation to the one-way mirror, and in fact it actually does work.
posted by ikkyu2 at 2:12 PM on July 30, 2007 [3 favorites]


...from a physics point of view, i can't think of *any* way of making an actual one-way mirror...

Closed-circuit television. That is, you convert the light to something else and then back. In the limiting case, its a wall-size LCD screen with tiny cameras mounted behind it.
posted by vacapinta at 2:19 PM on July 30, 2007


Also, the one-way sound wall was described by damn dirty ape above: Soundproof wall with mics on one side and speakers on the other.
posted by vacapinta at 2:21 PM on July 30, 2007


You could construct a one-way sound barrier, if you are willing to define "one-way" in the same way as it is used in "one-way mirror". Imagine you get yourself something that transmits about 1/2 of normal speech and reflects as much as possible of the remaining half (wood would probably be a good choice). On the "suspect" side, everyone talks twice as loud as normal and thus are heard on the police side at a normal volume. On the police side, everyone talks just above the threshold of audibility, and thus are inaudible on the suspect side. Additionally, some of the sound on the suspect is reflected by the wood back at the speaker, thus covering the sound coming from the police side. Not quite the same as a one-way mirror because light reflects and passes through a one-way mirror in a way quite different from how sound reflects and passes through a piece of wood. Also, a fair bit of sound would be absorbed in the wood, which is not the case when light passes through glass.

An analogous situation is when you can hear your neighbour's loud music, but they can't hear your yelling at them to turn it off, because their music is much louder than you can yell and there is a wall between you.

Cool Papa Bell makes a good point: There is no ability to "go one way and not the other" in terms of dampening vibration. Similarly, a one-way mirror does not let light pass in one direction and not the other.

@damn dirty ape: The wave-particle duality of light isn't really germane in this case. Sounds is a wave (of vibration) and it reflects just fine (e.g. echoes). The reflection of light can be explained in terms of a wave or of a particle.
posted by ssg at 2:26 PM on July 30, 2007


I think the closest equivalent to a one-way mirror for sound would be an anechoic chamber with a small opening at one end, leading into a room that was very reflective to sound (hard walls).

Someone making noise in the room with hard walls would be very audible to a person in the center of the anechoic chamber, but if the person in the anechoic chamber made noise, it wouldn't be nearly as pronouned to the person in the very "loud" room. Particularly if you piped in some white noise to the loud room to raise the noise floor and make the person's ears there less sensitive to sounds. (Analogous to the bright light on one side of a one-way mirror.)

"One way mirrors" are just semi-silvered glass anyway. Maybe the closest approximation to one-way glass for sound would be some sort of material that reflected most sound waves but let some through (most acoustic tile would work) placed over an opening, with the 'observer' hiding behind it and not producing any sound themselves.
posted by Kadin2048 at 2:34 PM on July 30, 2007


...from a physics point of view, i can't think of *any* way of making an actual one-way mirror...

Closed-circuit television. That is, you convert the light to something else and then back. In the limiting case, its a wall-size LCD screen with tiny cameras mounted behind it.


Closed circuit television would certainly seem to do the trick-- yet if it does, it implies the truth of a surprising and controversial claim; namely that closed circuit television must operate according to physical laws which are not symmetrical in time.

Otherwise, you could simply 'play the film backwards' and light would fall on the TV screen and various physical processes would take place which would result in light emerging from the lens. None of the physical laws we use to describe the process in the forward direction appear to forbid reversal in time, and the irreversibility we see is usually attributed to probabilities (i. e., entropy).


I am not too happy with the probability argument, but if you accept it, then even the closed circuit television set-up can and will operate in reverse, even if only relatively very very rarely.
posted by jamjam at 3:19 PM on July 30, 2007


They've tried, but it never works out.
posted by gally99 at 5:43 PM on July 30, 2007


seargant sandwich and ikkyu2 have it; it's got hardly a thing to do with the wave/particle duality, although that was a fun thought experiment :) Hopefully dunstanorchard, will come back and reconsider the best-answer designations. Alas, the best answers here are longer and more tedious to read, not as short and pat as the marked answer ... The Who example illustrates it well, nice job.

Now where's that flashlight-in-a-mirrored-sphere thread ...
posted by intermod at 8:07 PM on July 30, 2007


Another vote here for seargant sandwich and ikkyu2 having described the situation correctly. The other thing acting in an (optical) one-way-mirror's favor is that the reflected image will provide lots of visual clutter which will make it hard to discern anything on the other side, even if enough light comes through that you could see something otherwise. The hidden room is made visually "quiet" (nonreflective surface, dark walls, and so on). This is in addition to the dynamic range business.

sergeant sandwich: I think I've seen polarization-rotation tricks used for this in instruments. Take a polarizer, then something that rotates polarization right 45°, then a polarizer at 45° from the first. Going one way, all the light that gets through the first polarizer will be rotated to line up with the second. Going the other way, all the light that gets through the first will be rotated to be at 90° from the next polarizer and will be blocked. (At first I thought you could only get a sqrt(2) ratio this way but it looks to me like you can get 100% isolation.) Yeah, the nonreciprocity of this bothers me too.
posted by hattifattener at 9:43 PM on July 30, 2007


hattifattener, you just described a homework problem i had once. the problem with a plain ol polarization rotator is that the handedness of the rotation is reversed when going the opposite direction.

however digging into it, it looks like some clever use of the faraday effect and birefringence gets you most of the way there.
posted by sergeant sandwich at 10:19 PM on July 30, 2007


Hm, I thought the reversal of rotation was why it would work at all. I was describing exactly the device on the page you linked to, except using something like a plain ol' dextrose solution instead of the Faraday rotator. As far as I can tell, it should work the same. Am I wrong?
posted by hattifattener at 11:39 PM on July 31, 2007


« Older They can't ALL be lying...   |   Life in Tokyo Newer »
This thread is closed to new comments.