Thanks, misstl,

These are the sort of comments I was hoping to evoke from members. Clear the flexible, resonant cabinet is a completely different approach from the very rigid with lots of internal stuffing. Are Harbeth speakers of the "thin walled" type? They certainly have a great reputation.

What I wonder is which speaker design is likely to produce the more predictable, less problematic result for the DIY speaker designer? I suspect that rigid cabinet design will be freer from anomalies -- that is, more likely to work the first time without a lot of testing or experimentation.

Quote Originally Posted by mlsstl
Resonant boxes are one solution to the problem of back-waves from the drivers. A number of well regarded speakers use this method. There is also a degree of misunderstanding about the theory of the design. (Full disclosure, I have a set of Spendor SP1/2Es which use this cabinet design. The newer Spendor "S" series don't.))

As the cone oscillates back and forth, the driver creates a back wave of sound that is just as strong as the front wave launched into the room. The question becomes what happens to this energy?

Bass frequencies can be sent through a transmission line or port and actually reinforce bass output. (Or used in an acoustic suspension design to stiffen the compliance of the driver at bass frequencies.) Higher frequencies can be absorbed by the batting or stuffing that is used inside the box. However, there is a middle and lower range of frequencies that aren't well absorbed by the speaker stuffing.

If not handled in some fashion, they can re-radiate though the driver cone. However, they'll be out of time with the front launched wave and this can smear the sound. The ear is particularly sensitive in this frequency range.

The goal of a cabinet that flexes a bit is rather like crash crumple zones in car design. Rather than transmit 100% of the impact into the passenger compartment, the goal is to use up a lot of that energy by bending metal.

By having a speaker wall that flexes, the internal pressure wave is partially dissipated in flexing the cabinet walls. Part two of the design is to have a heavy coating on the interior cabinet wall to dampen the panel at the frequencies where the ear is most sensitive. This helps the exterior of the panel reduce what it radiates into the listening room.

Of course, there are other approaches for how to dissipate the rear wave, but it is an issue that all designers have to face. Simply having a cabinet that is so stiff that it sounds dead when you rap it does nothing by itself to get rid of that midrange energy. And you have to be careful in deadening the driver cone to prevent the back wave from passing through as this can adversely affect the driver's other performance factors.

Like many engineering problems, there are multiple ways to address it. Each has their advantages and disadvantages. However, there are some marvelous sounding speakers that use the thin-wall BBC type design, and they are all highly regarded for their clarity in the reproduction of the human voice. It works for me.