Driver/speaker box optimization

I was hoping to get a little insight into making a speaker cabinet for a project I’m considering. My thoughts are, when looking at a ‘high end Radio Shack or Bose’ speaker or a Wilson Audio/Focal/Magico speaker, when it comes down to it all you really have is a driver mounted in one side of a 6 sided rectangular solid box. One ends up with a stack of boxes with drivers on one side of their respective boxes. With all due respect to those willing to share ideas and opinions, it’s a given that a lot more than what I said has gone on to get that driver mounted in the box. I’m not interested in construction details of the box or anything regarding crossovers. I’m curious about what one does to optimize the box to the particular driver.

I am considering a MTMWW array, all sealed enclosures, using Scan-Speak drivers. D3004 tweeter, 12MU upper mids, 18WU lower mids, and a powered sub, probably a Rythmik servo sub, tbd. A Marchand 4 channel active tube crossover will be used feeding a stack of amps, again, tbd. We don’t need to talk about driver selection or crossovers or amplifiers. The upper mids and lower mids will either be in separate enclosures or each driver type/size will have both drivers in one box. THE QUESTION: with a driver(s) in a box, with minimal baffle area around the driver, attempting to not have parallel sides of the box, with a mic 1 meter directly in front of the driver, what does one test for when optimizing box volume? On their data sheets, Scan-Speak recommends a box volume for sealed enclosures. Madisound also recommends a box volume for sealed enclosures. I’ve built the Klang&Ton NADA kit which uses the 18WU driver and am happy with how it sounds. Its volume does not come close to either Scan-Speak or Madisound’s recommendations. These numbers are far enough apart to not quite know what’s ‘right’. Does one build box volume so the driver and box plays as loud as possible? Does one build the box volume to increase frequency response? When building a speaker cabinet, not considering materials or bracing or resonance, I think you have only two variables to consider – shape and volume of the cabinet. In this case, the shape is more or less given. The box will be a rectangular or trapezoidal solid shape. I think I’m left with volume to evaluate. Each driver will have its own discrete amp channel driving it, and each amp will have a restricted frequency range to limit sending too low of a frequency to a driver that might be damaged by too low/too much power/bottoming out concerns. I’d like to know what to test for when optimizing a driver to box volume.

I look forward to any help you guys can offer. Thank you.
Cabinet internal volume for a particular woofer/midwoofer makes it possible to arrive at a certain cut-off frequency (F3/F6/F10).
No manufacturer can tell which is a more appropriate value to any user because people have different priorities.
Cabinet volume is no longer a concern if you are to cross the drive unit significantly higher in frequency for any reason.

NADA kit cabinet size is simply a decision made by its designers. In that respect they have written a couple of clarifying sentences in magazine issue K+T 4/2011.

KlangTon 4-2011 page 19.png

What's telling us is that one would be disapointed in Nada kit if one were focused on how deep the bass response measured. Instead, one should take a listen to realize what the loudspeaker can really sound like. (Meaning way better than assumed!)

jeah, the box volume is important for the low knee of particular driver, and rest of the particular driver bandwidth is affected less. So, basically box size could be almost anything for all drivers except for the low driver that is responsible for the lowest bass in the system.

Simplified, lowest bass is dominated by room modes, so even for the low bass the volume is not too critical, could easily be +/-10% or even more, just move your head 20cm and bass changes more. Basically one could choose any reasonable volume and it could sound quite the same in room.

So, for example my speaker boxes don't consider the inside volume at all, that's about the least important parameter for big speaker system where there is plenty of bass regardless, but big boxes and drivers start to bring problems on the midrange, thus minimizing outside dimensions makes sense.

On the other hand, if you have a smaller system with small drivers, you probably need to use bit bigger box to get all the bass out from the system as there is rarely enough.

So, for box volume, use your favorite simulation program and take what it suggests and it's likely fine. If you later on find out something else is more important, and want to readjust the box, then do it. Best way to good sound is to make prototypes.
Forgot to mention that on some application one might want to optimize for good thermal exchange, or minimize excursion in a way that lowest lows are sacrificed for bit more output above and so on. You'd likely come across to Hoffman's iron law, which is basically thay one can choose any two of size/max putput/extension, while the third one is the trade-off.
I go for max volume that does power handling(manufacture specified) without exceeding xmax for woofers. Mid woofer volume does not matter mostly as long as I don't make it boomy. May be tweak a little here and there to better cross over to other bands.
Edit: Also I personally prefer qtc of 0.8 to 0.9 for subs/woofers. They just feel fast and more dynamic for movies or music.
Last edited:
  • Like
Reactions: 1 user
Qtc of around .707 to .900
is somewhat normal or desired goal for volume.
With mids on crossover filter Qtc can be higher
if space is concern.
I just keep things large as possible and stuff
the heck out of them with absorption.

Tiny tiny magical baffles, just make the baffle step
at higher frequency. And "edge diffraction"
just happen sooner.
Im over it, most aren't
Wide baffle, with driver offset mounted in
ideal location
will give flattest response.