OK, I just repeated your tests, using the Chesky disc:

In-phase/out-of-phase, results were correct both in DD and in 6 CH BYPASS (without changing anything in my system).

I then shut all my monoblock amps off to the sats, set all speakers to 'large' and sub to 'on'.

Using the in phase-out of phase test #23, 60 Hz, in DD mode (BTW, the readings were identical for both in phase and out of phase, as you expected, with only the sub playing).

I got basically the same results as your test. This does NOT mean that summed bass is getting to your sub when all speakers are set to 'large' and sub 'on'. It means that the 950's x-over is working properly.

A Linkwitz/Riley 24 dB/octave crossover, which is what the BM consists of in the 950, causes the signal of both the HP filter and the LP filter to be attenuated 6 dB at crossover so that when the 2 signals sum, they are at unity.

When you play a 60 Hz tone and set the crossover to 40 Hz, the signal from only the LP should be down approx. 16 dB.

When you set the x-over at 60 Hz, it should be down 6 dB, as explained above.

When you set the x-over at 80 Hz, it should be down approx 3 dB.

When you set the x-over at 100 Hz, or higher, it no longer attenuates a 60 Hz tone, so it should play at full volume, which, in your case, is 6 dB higher than the reading you get with the x-over set at 60 Hz, playing a 60 Hz tone.

Look at your results and you'll see that the crossover is pretty deadly accurate, accounting for room interaction at 60 Hz at your LP.

What this test DOES prove, is that the so-called 120 Hz global LP filter on the LFE signal is governed by the crossover, and, as such, is NOT global.

It's moot to me, as my LFE is discrete and free of the 950's influence anyway, but Scott assured me in an e-mail that I was mistaken and offered to bench test my 950 to prove it.

If the LFE filter was a global 120 Hz, ALL the readings would be the same in this test.

Sorry for the side issue, and leaving it aside, I hope the explanation above helps you to understand your test results.

As far as the analog BM switch goes, it does NOT read 'in/out' (as in 'bypass'), it does NOT read 'on/on'...it reads 'on/off'. That refers to the sats as 'large' or 'small' and nothing more. The specific function of this switch is more than clearly explained in the manual on page 9.

You can disagree with the nomenclature of the switch for whatever reason, but that does not make the 'on/off' labeling by Outlaw a mistake.

As I said above and many times before...experiment with the switch vs say, a uni player's DVD-A/SACD BM and see if you don't get a better curve with the 950. This is a whole other story for another time, which I will be more than happy to aid and assist anyone who would like to understand Outlaw's engineering of this aspect of the 950 with.