sound point-source models - theory

[Dave Jones]

Various software packages use models based on imaginary point sources. I do not know if that's the right nomenclature. The phrase "pole-zero pair" may resonate (so to speak) with experts on the subject. I refer to software like edge.exe (The Edge). I have also seen charts from M.J. King that make me think its possible to implement such models in mathcad.

QUESTIONS: Are there resources on the net where I can learn how those things work? And if so, where? What are some good terms to google for?

UNRELATED BONUS QUESTION: I recently found some animated gifs on the net that illustrate how passive radiators work, using a pendulum attached to a pendulum as a metaphor. I lost them, andhave not been able to google them back. Anyone know where those are?

[Dave Jones]

Bump for the early morning crew in Europe.

[godfrey]

Maybe google: acoustics "point source" calculation
or look for an acoustics tutorial or textbook.
I don't think "pole-zero pair" has anything to do with what you're looking for.
fwiw I think The Edge is seriously flawed although many folks seem to believe in it.

Happy hunting!

[Dave Jones]

Originally Posted by godfrey

Maybe google: acoustics "point source" calculation
or look for an acoustics tutorial or textbook.
I don't think "pole-zero pair" has anything to do with what you're looking for.
fwiw I think The Edge is seriously flawed although many folks seem to believe in it.

Happy hunting!

Thanks for the tips. Can you recommend an acoustics book? They are all so very expensive, I am reluctant to buy one on spec.

What's your beef with edge.exe? The fact that it assumes the baffle step is always 6dB? Or is there more?

[JohnPM]

Search for "image source method".

[godfrey]

Originally Posted by Dave Jones

What's your beef with edge.exe? The fact that it assumes the baffle step is always 6dB?

No, not that.
The problem I found was with open baffles. If you put a tiny driver in the middle of a big baffle, the result looks reasonable - big wobbles in the frequency response with the lowest peak where the baffle width is about a wavelength, and a steep roll-off below that.

However, if you move the driver right to the edge of the baffle (or into a corner), the output at very low frequencies is reduced, BUT not nearly enough. e.g. It says output at 30 Hz is only 3dB down with a point source 1mm away from the edge.

IMO that's not right - you don't lose half the output, you lose almost all of it since there's an almost perfect acoustic short-circuit from front to back.

I don't know what the problem is, how bad the errors are with more reasonable driver placement, whether it also affects calculations for sealed boxes, or how to fix it, but it doesn't look right to me so I don't trust it.

OTOH maybe I'm wrong. That happens too.

[godfrey]

Originally Posted by Dave Jones

Can you recommend an acoustics book? They are all so very expensive, I am reluctant to buy one on spec.

Yup, scary expensive.
Some years ago, I used to visit the local University library quite often. As a visitor, I couldn't take books out, but was allowed to read all I wanted, and make photostats. I found that kind of funny - that you could get all the education you want for free. It only costs serious money if you want the piece of paper with your name on it. Anyway, nowadays I find most of what I want on the internet.