FSP-LA

Deadbeat,

is there any reason we seem to be having a discussion with a third person with you acting as a medium? It is not that difficult to register on this forum.

I'm not sure of any reason, Tom seems to be content with emailing me his replies.

He can be found at:
danleylabs at the website which is comcast dot net
(sorry for my over-zealous anti-spam tactics which probably don't affect anything)

If you continue the discussion through email, be sure to post some of it so we can all benefit. I myself am quite interested in LA theory, though I'd never build one because I cannot stand the sound of most of them (I think it is the comp waveguides....)


Edited by: Deadbeat

This would be the most easily achieveable waveguide for a DIY LA in my opinion:

Hi,

Tom has registered on the forum, I am posting this before the confirmation email arrives; I'll be observing though! Nice view from here.

Tom:


Hi Jsg, Steve, all



In thinking about it, I realized that I haven’t given you guys much of
an explanation for what probably appears to contradict the marketed
popular belief.



The inverse square law results from a spherical expansion of sound, it
spreads out as pressure is want to do according to the local pressure
gradient.

SPL at a distance from the origin is found by;



Point source SPL = (10*log (acousticpowerWatts / 10^-12) – 20*log(distance in meters) – 11dB

Thus, one finds one acoustic Watt radiated spherically produces 109 dB at a one meter radius.



A point source is a common thing and is what results if one has a
source too small to produce its own directivity, like a woofer.

A conical horn driven by an acoustically small source can also produces a “point source” radiation but over a limited angle.



The issue of the line source is much more involved, while there is a
formula like the one for a point source, it only applies when the line
is acoustically infinitely long, where it produces a cylindrical
wavefront.

Here the pressure is confined to spreading in one plane only. This is
the condition one has with a very very long line source or a ribbon
speaker running from floor to ceiling (here the acoustic mirror extends
the length to acoustic infinity)



Anyway, for such a source;



Line source SPL = (10*log(acousticpowerWatts/ line length in meters / 10^-12) – log(distance) –8dB



Real line sources are generally not long enough to reach this point
(acoustically speaking, infinity long) and as one moves far away from a
finite line source, you also move away from this condition. Thus, one
finds that most of the time, a line array behaves somewhere short of
theory.



In the line array math one also finds that the total “power” is
distributed over its length instead of concentrated at one origin.

So, to use the math one is forced to assign a length yet if that is not
long enough, the math over estimates the “line source” advantage.



So, lets pretend one has a point source and line source, both radiating one Watt.

The line source we will make very long so it is closer to a cylindrical source.



For this example, lets say the line is 10 meters long and is assumed to
be large enough to radiate as an omni source in the Horizontal plane.



For the point source, it radiates in all directions equally.



So when you do the math, one finds that at 5 meters the sound levels are equal, the line and point source produce the same SPL.

As you move closer, the point source becomes louder, farther away say at 20 meters, the line source is 6dB louder.



A real line source has a confined Horizontal pattern so lets add the
+6dB one would accrue by limiting the Horizontal angle to say about 90
degrees.



Similarly, A horn based point source also has a confined radiation
pattern, for example an SH-50 has a measured axial Q of about 24, that
is it has a 24dB gain over an omni directional source so at 20 meters,
the point source based on that would be about +12 louder than the line
source. In this case, one only gives a 50 degree wide pattern and has
more angle than one might need Vertically.



In other words, the directivity that is automatically part of a conical
horn point source gives it a large advantage over the omni directional
case of a point source. In this case, at about 80 meters the systems
are equal, anywhere closer and the point source is louder, farther away
the line is louder (if it is still long enough).

In actual use the audience is rarely more than 80 meters away from the
speakers and then one normally finds delay rings are used for this or
even closer distances.



I should explain how this came to me.

We sell a lot of subwoofers to people that have line arrays and It dawned on me that the numbers didn’t add up.

It took a much larger number of drivers and Watts to keep up with a
given number of our subs when deployed as a “line array” compared to
“if’ they were used in a point source system like we also sell.

I had mentioned this observation to a friend who is a driver engineer and he said “yeah the line array craze saved our tails”



Ever notice how much smaller a function one array for a given size audience is compared to line arrays?


Best,

Tom

Tom makes a very valid point. If the point source is radiating in all directions, its SPL drops 6dB with every doubling of distance (inverse square law). However, horns can add directionality, meaning that the area covered by them does not increase in a 'squared area related to distance' fashion.

There i was trying to figure out all these analogies and hypothetical situations. Both point and line sources have their uses, but at the moment, the 'point and shoot' point source concept appeals to me.

What would happen if point source boxes were vertically arrayed? (ie to keep a particular coverage pattern, but increase overall SPL) You would have interference in the vertical plane, but would this be noticeable??
I would like to think, that a couple of point sources stacked on top of each other (as long as you weren't stacking, 50 or so high), would not exhibit too much destructive interference except at HF where silly things like head shadowing would become an 'issue' if you were to look at it in that much detail.

Has anybody had a look at the new Community LA (ENTASYS)? It's like no other PA-centric line array speaker I've seen.

Since it's christmas time, I thought I'd post my crazy LA concept! paying only slight attention to driver size, rigging, and other concerns, I put hand to paintbrush...



My paint skills are horrible, but I hope it gets the point across. I drew this one after wondering what happens when I have a few BMS 4510 in a row, put them on a horizontal OS waveguide with appropriate radiusing at the ends, then whap two drivers on it as on the Unity horn, and have the LF section as two tapped horns. Alternatively, you could ignore the Unity bit and make a special driver as on the Martin Omniline.

It looks pretty strange, but could be yet another (very rough) starting point for anybody looking to design a weirdo LA.

Sorry if this link has already been posted, but I found the thread very interesting - not only because of what if pooh poohs, but more about the strange behaviours Tom Danley discovered in his research into a line array replacement (Genesis GH60).

www.diyaudio.com/forums/showthread.php?postid=1670179

NB The thread is quite long - I ended up cutting and pasting and editing TDs posts into a word documentto get a readable doc.

Regards,
Jazomir

Edit Here is a link to an abbreviated version with fewer posts but all of TD's comments.
www.diyaudio.com/forums/showthread.php?s...page=25&pagenumber=1