Polarity mystery: Can you help me solve it?

There was a very interesting article published years ago about absolute/relative phase. "The Wood Effect" by Clark Johnson.

I believe that all you did was change the sound for those disks recorded in one polarity (e.g., inverted, non-inverted). So while those disks may sound better, the ones recorded in the opposite polarity should sound worse. However, since the speaker drivers are apparently not polarity-coherent to begin with, I'm surprised you could hear any difference.

One sure way to check the drivers, crossover, and internal wiring of the speaker is to place a
1 - 1/2 volt D cell battery across the external speaker box terminals. Disconnect the speaker cables from the speaker box terminals. Connect two test lead wires to the speaker box terminals long enough so you can view the drivers from the front. Tag the test lead that is connected to the positive speaker terminal. Use your fingers to hold the neg test wire on the neg post of the battery. Momentary touch the positive test wire to the positive post of the battery. The woofer cone should push forward, not suck back.
Not sure if you will see any movement in the other speaker driver.

The fact that you think it sounds better doesn't mean it is performing better. It could be a problem with how the speaker mates with your room. Here are the Stereophile measurements.

Hi Bryon,

Gee, that's a tough one, assuming as you indicate that the effect is repeatable across a wide range of recordings, presumably ruling out the polarity of the recording as being a factor.

The one possibility that occurs to me is that the output of your power amp has some amount of dc offset present, due either to itself or to the preamp or processor that is feeding it, or the source component for that matter (if the entire signal path is dc coupled). Does your friend, who observed the same phenomenon, have similar electronics?

A dc offset would cause your woofers to have a rest position that is either slightly forward or slightly backward relative to their normal zero-signal rest position, the direction depending on the polarity of the offset (and therefore depending on the polarity with which the speakers are connected to the amplifier).

That would tend to bias the contribution of the woofers such that acoustic compressions or rarefactions in the speakers' outputs are slightly emphasized or deemphasized relative to one another, depending on the polarity of your connections but independent of the polarity of the recording. I think :)

That's just a wild guess, of course, but it's the only theory I can think of that seems to fit all the facts.

Best regards,
-- Al