Can you focus a loud sound like a train horn?

Factual Questions
1

Preamble - what brought this to mind:
Our local train operator is changing their horns louder and softer.
They first made them louder because of a string of recent “accidents”. (In quotes because some were clearly suicides and in the others the victims knew the train was coming and tried to cross anyway)
Then they reduced them because the neighboring cities complained. Then louder again because they had gotten below government standards.

Now here’s the question:
Can sounds like train horns or police sirens be focused to warn only those directly in front, and not be heard by those far behind or far to the side?

I know in the reverse, listening microphones can have a parabolic dish, so people can clearly tape birds in the trees, etc.
Would a parabolic horn be that focused?

2

The parabolic reflector works in both directions but not sure what would happen on a moving object. I would guess that it would certainly make it louder in front and lessen the volume to the sides.

3

Yes, you can focus a sound. There are experimenting with wayside hornsat train crossings to eliminate use of train horns in communities. It involves the use of claxons that focus the sound toward the driver. They’re doing it in a city near me but I haven’t seen them yet. As it was described to me, the process includes full gates at crossings instead of one gate on each side.

4

If you want to make a directional wave of any kind (light, sound, radio, etc), the device needs to be larger than the wavelength. That’s the bare minimum needed to have any amount of directionality. The larger the device, the narrower the beam.

The really deep train horn is somewhere below 200 Hz, so you need a parabolic reflector about 6 ft across at the very least. It’d be much easier for a high-pitched siren.

5

Picture of a wayside horn.

6

You can use parabolic reflectors. You can even make a drum with balloon-rubber heads and put light pressure or vacuum on it to make the surfaces spherical, and make lenses. A heavy gas like CO2 or tungsten hexafluoride makes a convex lens collimate a beam, and with a light gas like helium makes a concave lens collimate a beam. Do not distract yourself with trifling concerns like tungsten hexafluoride gas; you wouldn’t need a horn at all if the front of a train weren’t already dangerous.

You don’t have to make the device bigger than the lowest frequency component wavelength to have useful effect. For deep horns, the higher frequency harmonics are a big part of the noise, and they focus with smaller devices. Also, in this context a six foot “device” can be several smaller devices in an array that spans six feet. Just check out the way radio telescope arrays work to see what I mean. Six feet isn’t necessarily a solid six feet.

7

I’m sorry, but - what? Could you clarify these statements a little bit, because I have no idea what you’re talking about. Using air pressure to make spherical lenses I understand (although how you then intend to focus sound with it is anybody’s guess,) but unless you’re whooshing with latter half of your post, I’m quite confused.

Heavy gasses to focus air pressure waves? Wha?

Here’s Cecil’s somewhat unhelpful take on it.

8

Years ago the some science museums had a room with a big weather balloon filled with CO2 in the center. You’d sit on one side, a friend on the opposite and then you could whisper to each other thru the balloon. Neato. I’ve done this but I can’t remember whether it was the SF Exploratorium or the Ontario Science Centre.

I’ve also been in buildings that have special built oval rooms where you could stand close to one wall and whisper to a person on the opposite side.

Wavelength doesn’t apply to parabolic reflectors. Many loud horns in fact have a parabolic horn (and many regular speakers especially PA systems).

9

Wouldn’t this be more a result of CO2’s increased density holding the compression wave together more effectively (and I’d imagine also significantly increasing the speed of transmission compared with normal air) rather than any weird collimating effect?

Genuinely curious!

10

No, the heavy gas causes a lower speed of sound. Just like glass slows the speed of light. So If you contain the heavy gas in a lens shaped container with accustically transparent walls (say thin rubber or plastic sheet) , then you can create an accustic lens.

Sulfer hexafloride is minimally (if at all) toxic. Filling your lungs with it will give you a Conway Twitty voice instead of the Mickey Mouse you get with Helium. Being heavy, SF6 tends to stay in your lungs, so the effect last longer than helium, but also carries greater risk of asphyxiation.

11

Sorry, but this is plain wrong. If the aperture is not a large fraction (better a multiple) of a wavelength, then diffraction effects will dominate.

12

Relevant comic

13

No whooshing, and others are filling in the light spots too, but sound waves refract at interfaces between gasses having different speeds of sound the same way light waves do between glasses of different index of refraction, which can be considered the ratio by which they slow the speed of light (actually in a way it isn’t light while it’s going though the glass, but that’s another story).

In the latter half of the post I was making two points. First, sound usually includes a range of frequencies and therefore wavelengths. There is a diffraction problem when you direct waves with an object whose size is not way bigger than the waves. But if you control the shorter waves, the higher frequency parts of a sound, you control a very noticeable portion of the sound. Irritating railroad horns would certainly be improved if you reduced just the high frequencies. And, second, you can use objects with complicated and disconnected coverage of a little bit of area, as long as it is distributed over a distance larger than the wavelengths, to manipulate waves. Having a bigger-than-wavelengh device doesn’t necessarily mean a solid object that big in all its dimensions. You can make a radio telescope with an aperture thousands of miles in diameter by properly linking two radio telescopes that are thousands of miles apart, at least in the sense of resolution and wavelength. But this isn’t necessarily good enough, because it isn’t very complete at redirecting the energy of the waves, it’s leaky. The exciting part about it was that it partly solves the wavelength dimension diffraction problem.

14

So all those parabolic horns out there are intended to make the sound come out all distorted???

15

They need crossing gates and horns? Is sounding a train’s horn at a crossing with gates standard procedure in the US?

16

I’m not sure what a “parabolic horn” is, but the horn or trumpet bell shaped curvaceous funnel coming out of a little box or block isn’t for direction, it’s a transformer for impedance matching and usually a resonator to help the horn oscillate.

The impedance matching function is needed because you can’t push all that hard on air, whereas the box at the small end of the horn can push very hard indeed but only on a small area of the air. Gradually tapering the horn means there is a gradual transition to a large area of air and a smaller pressure, which does a better job of coupling energy into the air. If on the other hand you try setting off a firecracker at one end of an open pipe, and then listening to that end, you will hear an echo from the far end of the pipe, because sound travels down the pipe and then reflects off the transition from pipe to air at the far end, where the impedance changes suddenly. The same thing happens to light as it goes between transparent substances of different index of refraction (the reason you can see glass), or to radio frequency electrical power if it goes from one cable to another and they have different impedances.

The resonator function is similar to what the body of wind or brass instruments does, and is the reason shorter horns and shorter organ pipes have higher pitch. Many horns in general work on the principle that there is some kind of bistable valve, such as a thin sheetmetal plate loosely held close to the end of a pipe it can reach out to plug, or such as lips that a trumpeter is essentially trying to keep barely closed. A column of air in the length of the horn sloshes back and forth, or a standing wave of air pressure rings in the pipe, and causes the valve to alternate between open and closed. The bursts of additional air from the valve cause the column of air to slosh back and forth. This makes an oscillator.

17

Yes, pretty much, because dumb people do dumb shit like this and still get themselves killed.

18

It seems like that crossing only had a gate on the right half half of the road. (Which is typical of level crossings in the US, but IIRC very uncommon in Europe and Japan.) Is the use of a horn still required if a gate goes across the whole road?

19

What they did may have been “dumb shit”, but the news article doesn’t say enough to know that it was. There are good reasons to drive around a closed gate with its lights going. For example, there is a neighborhood a couple of miles from here that includes a restaurant, but it is accessed by only one road, which has a railroad crossing. One evening we left the restaurant and the gates came down as we approached the crossing, and a train came toward the crossing but slowed and stopped about 60 feet away. Two engineers climbed down out of the still-running locomotive and walked along the tracks to the intersection, then headed out of the neighborhood, turned right at the next street, and walked into a bar, leaving the gates down. We waited around a while, and I got out and looked around, and gave up, as I couldn’t see any other trains and there was no sign of the engineers, and drove around the gates, with their lights flashing. I actually had to do some back and forth to maneuver between them because it was tight quarters.

Going through a closed gate can look safer than it is, too. A parked train can trigger the gates to close, but there can be more tracks behind it, and a fast train can appear from behind the parked one much faster than you can respond to. This sometimes nails even pedestrians, who have much better sight and sound clues than drivers.

Sometimes gates are closed and flashing for no apparent reason. A single track crossing near my workplace has about a mile of straight track visible on both sides, and sometimes it’s down and flashing with no train visible or audible, and sometimes I’ll wait for a minute but not for long.

A friend, when he finished college, drove his stuff and that of several other students to another city in a rented truck. He got stopped, first in line, at a crossing gate, and the train went by. The gate went up. He pulled forward, and the gate suddenly came down again, between the cab and the body of the truck, and another train appeared coming rapidly around the bend, not very far away. So, he kept going, breaking the gate and lodging it between the cab and the body, and for some reason was able to go far enough to get clear of the intersection without breaking the cable for the lights, which were still all flashing away as if on a Christmas tree. Just to complete the scene, there was a cop that had been on the other side of the intersection, also waiting to cross. But fortunately the cop had seen every single detail and gave him no trouble at all, and helped get everything unstuck and kept him from getting in any trouble.

Anyway, I think there are many things that can go wrong with the whole scheme of closing gates to keep people safe.

20

I still don’t see the additional safety afforded by a train sounding its horn to some fool crossing against the gates. For example if a freight train travels at 60 mph and sounds its horn from a tenth of a mile (just using some arbitrary round numbers - surely you won’t hear the horn from much farther in a car with windows closed, people talking, radio or CD player on), then the train will hit you in six seconds. A passenger train of 120 mph, in three seconds.

We have ~40-70 fatalities/year on level crossings here in Germany - mostly people who drive on bypassing the gates (half-gates on the right sides rather than full ones are mandatory for safety reasons at crossings that are not visually monitored by a human every time the gates are lowered). I doubt that approaching trains sounding their horn would lower this toll any because if you are on the crossing and hear the horn you are already as good as dead (see above). (trains only sound their horn for some low-use crossings without gates on low-speed lines.)