Crowd Size and Physics of Pi(e):Yet another sizing post

[mtglass]

Bruce, would you kindly let me know which compression drivers you're running? I searched the forums, and went 20+ pages deep, as I thought there was a discussion on which performed "best."

Live music, just a cover band. 60-70s mostly. But, as you probably guessed, I'm the bass player and on songs like Purple Rain I want a millennium of sustain on a 34.65Hz C#. I know people say that the second harmonic is dominant on a bass, but I don't see that on a spectrum analyzer. So I settled on T48's knowing that I'm going to get a steep fall off at 40Hz, and they beat the pants off everything else I've tried: it's still an audible decline in SPL as I walk down the scale.

2 per side is a start. Bill says I won't need more than 3 per side, so I certainly want to keep the option open to go to three. Then I think about some of the establishments we play in, and I think it might be helpful to have a couple more per side, strictly for more sensitivity. So I'm planning on 4 maximum per side total, 2 to start. Still make sense to just stack non-array OTops? We're completely out of my experience here. I've only ever used a single main per side.

Also, after adding a 12dB hump centered at 60Hz, I'm struggling with subwoofer math. Using Bill's T48 - 8 stack SPL chart, it looks like a struggle to get to 112dB at 75': 48 x T48's? That seems like I'm missing something still. Sorry. I don't really need an answer since I'll just keep adding T48's until I get where I want.

[Seth]

I believe Bruce's cabs have NSD2005 compression drivers

[Marflinger]

Hey, i did some simulations before about the max spl of T48 vs T30.



This would be 4 T48 dual 12" loaded 30" wide, stacked with a V-plate. Not sure if this is correct in terms of the simulation, but i think not that far off.
The driver choice really does change the limits; here is the recommended LaVoce 12" used.

To have 112 db flat at 75' (22,x meters?), at 1m you would need to have the capability of ~144db (what a stack of 4 already does above the 35Hz probably)
The Graph is the hardware limits of the driver used, so not the pressure to safely produce. But already 4 should do that.

Please correct if wrong, but i think it should work like that.

[mtglass]

Thank you Marflinger. I'm using 24" wide T48's loaded with 1 x 15" 15SW2000-4. I assume it's more or less the same. Thank you again. I've been using an online calculator, so I guess I'll have to get a better tool to do the calculations.

EDIT: I missed your simulation has 2 x 12"... Math..

[Marflinger]

I've learned that spl results from area moved (cone area sd) and how far it is moved (xmax)
Single cab of dual 12" (same assumptions as above) would be this in max spl (grey is single, dark is stack of 4)



Maybe you can compare that to yours and scale the differences to the max spl...
Rest of the thread is here: viewtopic.php?p=299135#p299135

[mtglass]

Uhm, HornResp is going to have a steep learning curve for me, mind sharing the input parameters? I completely understand if you don't want to give that away. Best!

[mtglass]

2 x LaVoce 12" = 2 * 531 cm^2 * 1.15 cm = 1,221 cm^3
1 x PRV 15" = 1 * 908 cm^2 * 1.25 cm = 1,135 cm^3

2 x 12 has 86 more cm^3 volume, about 7%. I don't think that's logarithmic, so I can de-rate your graph by 7% and end up with results for the PRV? That yields 136.71dB, so just a little shy. Which means even with a 12dB hump, Bills original answer of 6 max is correct, and I'm way down a rabbit hole.

EDIT: Oh, plus I'm a 24" wide, so more de-rating...

[Marflinger]

I wouldn't mind at all, just not sure if this would brake anything which is not intended.
Did change the segments to 24" width and imported your driver

This would be your driver, single cab, 1w/1m sensitivity on a 24" wide

Screenshot 2026-02-12 162439.png (9 KiB) Viewed 40 times

Max SPL single Cab

Screenshot 2026-02-12 162551.png (9.54 KiB) Viewed 40 times

Max SPL 4 vs single:

Screenshot 2026-02-12 162705.png (11.04 KiB) Viewed 40 times

Max SPL 6

Screenshot 2026-02-12 162807.png (9.91 KiB) Viewed 40 times

Max SPL 8

Screenshot 2026-02-12 162854.png (9.94 KiB) Viewed 40 times

The dent at 44 Hz is limited by diaphragm displacement.
Coupling with a V-Plate not included.

I hope that helps; not sure if my simulating is absolutely correct; see the other thread.

[mtglass]

Thank you kindly! I'm going to have to take a bit to digest.

[Seth]

Gotta say, guys...it's nice to see some more mathematics oriented, number crunchers chatting it up around here. You guys are speaking my language and I like it!

[mtglass]

Proposal, for your suggestions:

I think the thing to do is get my cabinets out, and measure the pair with my 50V / 625W limiter in place. I'll take several points 30, 40, 50, 60, 80, 100, or something. That will re-affirm my memory of their performance, and inform the end target when compared to the hornresp model. Also, I built them with reducers.

I'm thinking to do it without a V-Plate since Bill says V-Plates aren't as useful with a > 4 speaker stack, so the V-Plate would throw off the scaling.

I'm going to leave the 48dB/octave 40Hz high pass filter on, and remove the low pass filter.

I use an American Recorder SPL meter, for reference. It's been a while since calibration: I think about 18 months. Certainly less than 2 years.

Might take a while, don't hold your breath.

EDIT: I use the 50V / 625W limiter for two reasons: (1) driver protection, and (2) I'm trying to keep total system AC power demand under a single phase 220v/50A feed.

Thoughts?

[Bruce Weldy]

I believe Bruce's cabs have NSD2005 compression drivers

You are correct, sir!

Those are no longer available. Bill has another recommended driver in the plans. I think it's the F151, but don't quote me.

[Marflinger]

Thoughts?

Probably the only way to really get reasonable data.
The real data beyond simulation.

Room eq wizard would be an option to measure it in all points.
Voltage limiting no error source i think. (Peak limiting, rms probably also not)

Keep in mind, hornresp max spl just show the limits of a combination, not the useable situation (in theory useable when exactly reproduced the curve, would not recommend)
And simulation has a tolerance of +-3db usually against real life data.

[mtglass]

REW, great idea, I don't know why I didn't think of that.