Neighbor downhill from myself and I are working on an idea for Halloween, want to build an extremely narrow speaker.
Want to project sound to the intersection 450 feet away at whisper to conversation levels, roughly 40-70 dB so modest goals there. I calculate this to be about 110 dB at source, which seems perfectly achievable
Would like to have very narrow dispersion to keep sound inside the intersection that is 50 feet wide, and to normal height of people, really looking at a design goal of 0.5 degree by 6 degree dispersion. This is so far outside what I have ever seen I don't even know if that is possible.
I see commercial waveguides for line arrays for narrow vertical dispersion of around 15 degrees, but have wide horizontal dispersion.
Do not want to blast people on the sidewalk directly in front of our houses, and will place speaker on my roof which would elevate it about 20 feet higher than sidewalk and street,
Mostly wanting to achieve speech frequency ranges so maybe 300 z to 2k Hz, but wider range would be better. Care about intelligibility but this is not a hifi setup by any means.
I have built tapped horns before, and have understanding of line arrays, but have only designed subwoofer horns before.
Not even sure a horn is the best way to go about this, open to options.
So what is a likely way to achieve these rather odd goals? Or better question, is the angle requirements and intelligibility achievable, and if so how?
Paraline, unity horn, array?
I have multiple amplifier choices and DSP to work with as well.
Want to project sound to the intersection 450 feet away at whisper to conversation levels, roughly 40-70 dB so modest goals there. I calculate this to be about 110 dB at source, which seems perfectly achievable
Would like to have very narrow dispersion to keep sound inside the intersection that is 50 feet wide, and to normal height of people, really looking at a design goal of 0.5 degree by 6 degree dispersion. This is so far outside what I have ever seen I don't even know if that is possible.
I see commercial waveguides for line arrays for narrow vertical dispersion of around 15 degrees, but have wide horizontal dispersion.
Do not want to blast people on the sidewalk directly in front of our houses, and will place speaker on my roof which would elevate it about 20 feet higher than sidewalk and street,
Mostly wanting to achieve speech frequency ranges so maybe 300 z to 2k Hz, but wider range would be better. Care about intelligibility but this is not a hifi setup by any means.
I have built tapped horns before, and have understanding of line arrays, but have only designed subwoofer horns before.
Not even sure a horn is the best way to go about this, open to options.
So what is a likely way to achieve these rather odd goals? Or better question, is the angle requirements and intelligibility achievable, and if so how?
Paraline, unity horn, array?
I have multiple amplifier choices and DSP to work with as well.
If it's for voice frequencies I'd use a parabolic reflector with a small loudspeaker at its focal point.
Super narrow beam with "sound laser"! 🙂 Creating Coherent Sound Beams, Easily | Hackaday. Parabolic reflector would be 10x easier to implement though.
still the size of the reflector determines the nearfield/farfield transition.
one option is to place many small speakers on a circle of >10m diameter.
one option is to place many small speakers on a circle of >10m diameter.
It doesn't have to be that large if low frequency reproduction is not required. 2M should be plenty for voice frequencies.
Mouth width = 10^6/[wall angle*F], so with a 2" exit driver to get to 300 Hz will make for an interesting flare to terminate to a ~1.1" x 1.0", 0.5 x 6 deg mouth.
GM
GM
It's possible but you haven't seen it because it's not practical.
Basic wave physics sets the ratio of source width , wave size and beamwidth.
0.5° is about 1 unit wide per 100 units distance, or 0.01 radians if you prefer.
That means the source needs to be about 100 times the wave size to have that beamwidth.
For ~340 Hz the wave is about a metre, so a 100 m source size.
There are tricks to do better but that's the approximate order.
For 1 kHz and 6° it's only 10 feet/ 3 m.
So that's doable but still pretty substantial.
I would use a parabolic reflector, as William recommended.
My old university used to use one, combined with a JBL PA compression driver for atmosphere sonar, measured wind shear and density or similar.
Used to hit it with a compressed chirp, about a 1 kW IIRC.
Would have made a mean Halloween scare if that's what you have in mind.
Best wishes
David
Last edited:
The hackaday project could be risky. The carrier wave is an ultrasonic signal directed at the listener. Just because no one hears it does not mean that it doesnt matter.
Specify the beam angle and low frequency cut off, then make a CBT using full range 3" drivers to meet those requirement (ie beam angle and length).
I had initially researched the ultrasonic setups, it seems the easy way but there seems to be indications that high amplitude ultrasonic is not ear safe so wanted to avoid those.
I didn't think of a parabolic reflector, that would be a fairly easy build as well.
That may be the first thing to try. I don't have time for trial and error, but that seems highly likely to work right the first time versus lots of tweaking.
If I understood correctly the ultrasonic setup thing could be pointed to an object like a car or pavement or a tree, which would then appear to be source of the sound.
Parabolic dishes have the advantage of being readily available. Having experimented with both parabolic and hyperbolic reflectors, my preference is the hyperbolic.
Check this out:
HyperboLine ™ Projector
Art
An old C-band satellite dish comes to mind. A metal mesh version would have to be covered in something to make it acoustically reflective, but full metal ones were once common. You might be able to get one for just picking it up and ridding the owner of it (so the county will quit b****ing at them). Modern Ku-band dish is too small and not as much directivity. Dish Network dishes can hardly be called parabolic anymore - they just provide some gain. In any case, you would still want to feed it with a compression driver/horn. 2” exit phenolic is perfect - it will get you 300 Hz on a large format 40x60 horn.
In order to get that kind of laser focus with a front loaded horn you would probably need a flare rate in the subwoofer frequency range, going from about 2 inches to about 2 feet at the exit. It would be long and hard to fabricate, and with such a low cutoff frequency distortion in the midrange would be pretty darn high. Lots and lots of high order modes. Re-entrant PA horns suffer enough from this - think “nasal sound” on steroids. And they’re directive, but not THAT directive or you would just use one.
In order to get that kind of laser focus with a front loaded horn you would probably need a flare rate in the subwoofer frequency range, going from about 2 inches to about 2 feet at the exit. It would be long and hard to fabricate, and with such a low cutoff frequency distortion in the midrange would be pretty darn high. Lots and lots of high order modes. Re-entrant PA horns suffer enough from this - think “nasal sound” on steroids. And they’re directive, but not THAT directive or you would just use one.
Maybe something like a cheap old CB radio with a speaker output / speaker hidden at the intersection with a transmitter at the house might solve the problem without building huge horns ?
Half the goal is to build something crazy, the other half to make sure it isn't locatable without significant effort.
We think half the enjoyment people will get out of it is trying to figure out where it is coming from.
Maybe a focusing array? Get a bunch of identical small voice range speakers, set them around the area playing simultaneously from not too far away, pointed at and at the same distance from the point where you want the maximum sound intensity to form. Ideally round up some variable delay units (such as a handful of MiniDSP 2x4 units) so you can fine tweak the arrival times to build up the intensity where you want without having to too carefully place the speakers.
The idea is to get constructive interference where the wavefronts all sum up, and more and more destructive interference further from that point. Similar to a concave reflector but with multiple generators rather than reflection points on a dish.
This is likely to be too finely defined for what you want, though, from what you asked for. If you put the drivers all toward ones side of the target and closer together rather than all around it, that would widen the peak area though (eventually being equivalent to the concave reflector).
The idea is to get constructive interference where the wavefronts all sum up, and more and more destructive interference further from that point. Similar to a concave reflector but with multiple generators rather than reflection points on a dish.
This is likely to be too finely defined for what you want, though, from what you asked for. If you put the drivers all toward ones side of the target and closer together rather than all around it, that would widen the peak area though (eventually being equivalent to the concave reflector).
Last edited:
That would only work for a very small point in space. WL/2 at 1KHz is only 170mm.
If I wanted to scare the crap out of someone on the sidewalk I'd be tempted to flood the area with a near infrasonic heart beat type signal. This would require some capable subs and create significant spill into adjoining areas.
If I wanted to scare the crap out of someone on the sidewalk I'd be tempted to flood the area with a near infrasonic heart beat type signal. This would require some capable subs and create significant spill into adjoining areas.
Last edited:
