Why does a solo sound sound different when inverted?

[mhardy6647]

The polarity thing seems, empirically/anecdotally, to be similar to perfect pitch.
Some folks have it, some don't.
The quasi-logical explanation makes intuitive sense. When a drum is struck, the initial movement of the drumhead is in one direction, the polarity of the sound wave is... well... what it is, and if it is inverted, it will (could) sound different.
Of course, there are all sorts of loopy ideas on the internet (and elsewhere) that make intuitive sense if one doesn't think too hard about them. I mean, phlogiston made sense at one point, right?
The big challenge is that not every program source is rendered with correct polarity, so it is a playground (or minefield) for those with sufficient OCD to make a career of swapping cable connections or flipping switches. Additionally, who is to say that every mic/amp and every channel on a modern "stereo" (multiple-track mono mixed to two - or however many - channels) recording is captured with the correct polarity.

 

[anmpr1]

Yep, absolute polarity inversion is audible.

The question is not so much whether the effect is audible--that has been consistently determined under controlled conditions. Rather, how audible is it and under what conditions is it audible? And does polarity mismatch significantly (or even minimally) degrade the listening experience?

Other than the late Clark Johnson (and a few of his friends/followers) who claimed that it is a huge problem, it's never really been made out to be much of an issue by anyone. I don't even know the tweakiest of tweaks saying, "You know, the absolute phase in this recording is off, and that's screwing up the pace, timing, front to back depth, sound-stage and plankton." Even those neurotic types don't bring up the issue.

Consider: in a typical 'modern' recording how many electrical circuit stages are present until final mastering? How can anyone be certain of the polarity of them all? In a very simple set up it is possible. At RCA and Deutsche Grammophon?

 

[anmpr1]

FWIW, anyone interested in exploring this topic may want to download issue 18 of Audio Critic (below). There is a several page (rather rambling, if I can say so) letter from Clark Johnson, along with Peter Aczel's in-line comments and a final response (pp 3-6).

https://www.biline.ca/audio_critic/audio_critic_down.htm

 

[andreasmaaan]

I've been leaving gaps between, I have 2 tracks that alternate back and forth, one has the phase flipped
The difference is so slight that it probably isn't even worth talking about, but I did notice a difference.

Is there a general preference to having the peak of the waveform pushing the speaker out as apposed to being pulled in?
My logic is that if the peak pushes the speaker out, then there might be more punch, but maybe it makes no difference...

My first point would be that the idea that the polarity of the recording itself is "correct" is a pretty big starting assumption. And even if it (or some component in it) hasn't been specifically inverted, there are plenty of reasons its waveform may not actually resemble that of the acoustic event that was recorded.

As to the specific question, my memory is a bit hazy regarding the research on this, but IIRC, a polarity inversion may be perceivable as different, but for music signals (not including asymmetrical signals that have been specifically generated for the purpose of illustrating polarity inversion differences), neither the correct nor the inverted signal is likely to be preferred.

 

[LTig]

Regarding the drum hit example: When I hit my drums I always hear the drum sound in "wrong" polarity (since I move the drum head away from me) while listeners in the public hear "correct" polarity. Of course a drum sounds different to me and the public but I can't complain about bad sound ...

 

[tvrgeek]

A couple of points. You have no idea what is "right" through the entire recording process unless you do it your self. Modern studios try to keep it "correct" as in original, but older recordings are hit or miss. With all the pitch boxes, delays and splicing, who knows?

Many claim to hear a difference, many suggest it is placebo. It may, and I say again, may be detectable by your ears and that highly unpredictable signal processor between them.

Drivers do produce slightly different distortion depending on which direction they are driven. It was a fad a few years to use two woofers, in push-pull to cancel ( or more likely average) this distortion as the cone will deform differently if pushed or pulled. Drivers have improved and I do not see it done much any more. I suspect the real reduction was in doubling the surface and hence halving the displacement which can make a very large difference.

A difference in a recording? OK, which side of a drum skin is the mic on? So is the initial wave front positive or negative pressure. Rabbit hole.

 

[anmpr1]

Regarding the drum hit example: When I hit my drums I always hear the drum sound in "wrong" polarity (since I move the drum head away from me) while listeners in the public hear "correct" polarity. Of course a drum sounds different to me and the public but I can't complain about bad sound ...

Spoken like a drummer. Drummers always want everything in order. You should have been a guitar player. Dial in the distortion pedal, add some over the top overdrive, tap in the reverb, switch the humbuckers out of phase. Believe me, the last thing a guitar player is ever concerned about is polarity!

 

[Jdunk54nl]

What are you listening to this on? It could possibly be a speaker design issue and the out movement isn’t the same as the in, which would cause an audible difference.

A speaker perfectly symmetrical in its design shouldn’t have any differences as the same amount of air should be moved.

Side not...it bugs me when people use phase and polarity as the same thing...they aren’t....you are talking switching polarity.

 

[pozz]

The phase control is there to help mixing. By itself for single tracks it has little value.

The only reason I can think of is that inverted phase produces a slight change to the leading edge of the sound, which is important perceptually at low frequencies. But that's conjecture. I don't remember any studies on this.

 

[pozz]

However, the process of inverting polarity itself is far more audible than the effects of the inversion.

+1