Outdoor testing of a sound source

[Qacer]

If you have a sound source (e.g. megaphone), how would you test it properly in an outdoor environment?

What comes to mind is a subjective speech intelligibility test versus distance from the point source. But when trying to do quantitative tests using sound pressure level meters, all the sound propagation factors are a bit baffling (no pun intended). Ground reflection/absorption is crucial, so measuring the SPL at different areas is subjected to all sorts of attenuation factors.

.... So how would you do it properly? I'm thinking of just basic setup of SPL meter and GPS device, but then I read that wind, temperature, etc. has significant effects. Then again, if I use a amenometer and thermoter, I would not know what to do with their readings afterwards in terms of relating it to the sound.

:/

 

[Born2bwire]

I have not done this kind of stuff with acoustics, but my limited knowledge of the subject leads me to believe that there is a direct correlation with other wave problems. You could probably create a computational system of your source over a lossy half-space. I'm sure you can find out easily how to model the earth (like as a rigid boundary or such). The easiest way to do this would be to decompose your source into plane waves, where the solution due to the direct source and reflection would be simple to solve. This is probably already done in literature for a point source and you could probably easily adapt it for a baffled point source but I don't know about a megaphone (maybe an array of point sources that would approximate it's beam and directionality...).

The whole purpose of this would be that you could run simulations to predict what your results are and to see what factors would be important to you. I doubt temperature differences would be a significant factor in that it would probably be more or less uniform over your test area.

Or see if there is a semi-anechoic chamber for you to do measurements and lay down a reflective matting that would approximate your lossy earth. You'd be limited by the size of the chamber but would be able to remove the variables of wind.

 

[Cattlegod]

First you have to define what you want to test. What are you looking for in your results? What kind of analysis do you want to do?

 

[Qacer]

Actually, one the test that I wanted involves the speaker's directivity. But my main concern is the fact that certain frequencies have interference zones at a certain distance from the megaphone. I don't know too much about this topic. I wanted to test the megaphone on a hard surface, but I don't know what the right distance (radius of circular perimeter) from the sound source. I'll mostly be dealing with speech, which is considered to be from 500 to 2500 Hz.

 

[Cattlegod]

Originally posted by: Qacer
Actually, one the test that I wanted involves the speaker's directivity. But my main concern is the fact that certain frequencies have interference zones at a certain distance from the megaphone. I don't know too much about this topic. I wanted to test the megaphone on a hard surface, but I don't know what the right distance (radius of circular perimeter) from the sound source. I'll mostly be dealing with speech, which is considered to be from 500 to 2500 Hz.

what is your end goal? what are you trying to figure out and why?

 

[Qacer]

Originally posted by: Cattlegod
what is your end goal? what are you trying to figure out and why?

The goal would be to find an optimal speaker placement that has the optimal coverage area. I can probably get away with testing a each suitable location, but it would not be efficient. I actually found a paper titled something like "Nor2000..." and it has a sound propagation model. I'm going to try to read it. It says it uses ray tracing theory, but I don't know about it.

 

[Born2bwire]

Originally posted by: Qacer

Originally posted by: Cattlegod
what is your end goal? what are you trying to figure out and why?

The goal would be to find an optimal speaker placement that has the optimal coverage area. I can probably get away with testing a each suitable location, but it would not be efficient. I actually found a paper titled something like "Nor2000..." and it has a sound propagation model. I'm going to try to read it. It says it uses ray tracing theory, but I don't know about it.

Is this for a specific speaker? Because otherwise you could probably estimate it as a baffeled source and use wave-theory to figure out the radiation pattern. Better yet, what I would do is think of them as being an in-phase array of two elements. This way you can evaluate the far-field radiation patterns based on speaker placement. Though with acoustics I cannot recall what the far-field condition is. Ray-tracing is a little more crude than doing the wave theory, and I it probably relies on a general model in the same way that using a baffeled source does. My intuiution is that if these speakers are highly directional as megaphones, then you will probably need to do something similar to an anechoic chamber test. That is going out and measuring the patterns across various planes of the speaker (unless the speaker is axially symmetrical then you could probably get away with a single plane).