Agreed - my SPL meter is supposedly rated to 32Hz, but I don't find it accurate even below 50. I've only used my eyes in the measurements with respect to the gain on the receiver. I used Test Tone Generator software, which just plays a continuos tone of a given frequency. Here are the results:

Using a -6db receiver volume setting (which is quite loud), I simply looked at the woofer come movement.

10Hz, near 0 cone movement
12Hz, 20% of maximum cone excursion
14hz, 40% of max
16hz, 60% of max
18hz, 90% of max
20hz, 100%, -13db before max excursion
22hz, 100%, -16db before max
24hz, 100%, -16db before max
26Hz, 100%, -16db before max

Granted, these so-called tests I'm doing are not even close to scientific, but it's pretty clear that near zero come movement will not be producing much in the way of sound. How I determined cone excursion levels is to play a tone and turn up the gain until I either heard distortion or the voice coil bottom out slightly. I only played for a second before dropping down the levels. Actual real excursion may well be less than that, but it's a good absolute max value. I could also see the cone move nearly a full inch peak to peak at those times. So when I estimate 20% cone excursion at 12Hz, the cone is really barely moving. One other side note: the passive radiators exhibited the same behavior as the woofer cone.

Hey murphy2112 - thanks for the explanation, now it all makes sense. As you said, not scientific and nothing to do with what you can hear, but informative nonetheless.

As promised, I inquired with engineering about the filter, which I presumed was in place at very low frequencies (single digits). Turns out it is a bit higher than that and confirms your results (sort of). The explanation from engineering is very informative if you are interested in such things and comfortable with this level of technological discussion. I expect most forum members are not, and all that follows will be hugely confusing. To everyone not an engineer and not into this stuff for the technology (but rather the sound and the music/movies), stop reading now. For those who care to see some of the design thinking that results in these amazing loudspeaker systems, read on...

From our world-class engineering team and Bob Johnston, our head of engineering in our Arnprior, Canada facility:

The ANSWERS IS YES. It is below 18Hz and varies between models. A simple explanation would be "we do it because it would be stupid not to on our system", but for the long answer, following is a bit more engineering design detail.

As we have complete DSP control over the whole bass system the DSP only allows the bass section to be driven by signals that will result in well controlled and clean sounding output from the sub-woofer. If it will not sound good or will not result in useful acoustic output the DSP does not allow it to pass on the the power section and drive the sub-woofers. There are several reasons why we chose to do this:

1) There is no point in the sub-woofer driver cone moving large distances if there will be no usefull acoustic output so the DSP controls unproductive sub-sonic driver movement. Unproductive driver movement simply adds IM, Doppler and amplitude distortion to the reproducible bass degrading sound quality significantly.

2) If the DSP knows you will not hear or feel it, system resources are not wasted on futile attempts at reproduction. Driving the sub-woofer system with large sub-sonic signals that the DSP knows can not be reproduced wastes a lot of power that can be used reproducing the program material. The DSP uses its resources wisely in ways that take best advantage of the available acoustic system's capabilities to allow maximum sound pressure level of the program materiel with the least distortion.

3) Heat in any high power system is the enemy. It is a key limiting factor in how loud the system can sustain the program material over the long haul. The DSP wastes no power (heat) on driver movement that will not add to the listening experience. This allows the system to play the program materiel cleaner, louder, longer.

4) Protection. Excessive sub-sonic driver movement can easily cause nasty noises like driver bottoming or even in some cases driver damage. The DSP monitors for this and controls the drivers to prevent excessive excursion that can sound bad or damage the system.

5) As all signal processing is done in the software domain by the DSP so the audible artifacts from additional traditional analogue filters is avoided. There are no "extra" sub-sonic filters to degrade the signal path as would be encountered in traditional analogue filters. Sub-sonic noise removal is simply one task of the DSP processor.

(One note is our sub-sonic noise removal was not designed to deal completely with warped records. It will help but it is not intended for that purpose.)

Moderator wrote: From our world-class engineering team and Bob Johnston, our head of engineering in our Arnprior, Canada facility...

That was a world-class answer. Much appreciated. It seems that the customer service on this forum could equally be called world-class. Regarding points 1 through 5: At very low listening volumes, I could see a case where trying to reproduce some of the content in the 16-20Hz range might be a good thing, but I more appreciate that the TTs can play loud in the range that they are supposed to play in.

That said, I am going to be looking for some replacement subs that can do the ultrasonic duty.