Can you hear inverted phase

wklie

wklie

Well-known member
Joined
Jul 29, 2014
Messages
3,163
Location
Hong Kong
Some months ago we did some exhaustive testing before shipping Lumin X1, and that included phase inversion feature testing. A product manager thought he could not hear phase inversion. Yet when I played some bass-heavy music (e.g. Billie Jean from Thriller SACD, if I remember correctly), he was surprised that he could actually hear it, and the drum sounded so different in our Vivid Giya G2 setup.

Phase inversion can happen in some cases. Some Japanese products like Accuphase and Luxman (note: not including Esoteric) adopt XLR pin 2 cold standard, but I believe it is more common to have source products that have XLR pin 2 hot. So using these together via XLR results in phase inversion by default. Luxman has a polarity inversion button, but Accuphase does not. Then there are many albums that are recorded in inverted phase, making this issue complicated - or even worse, compilation albums may have different phase for different tracks.

Back to our tests. In the particular tracks we tried, we came to a conclusion that a correct phase gives more audible details. Yet there were also cases where some people (subjectively) like the incorrect phase more for particular tracks. (This may also be an example of that the most accurate audio presentation is not necessarily the most enjoyable for a listener's subjective taste.)

So, do you guys hear phase inversion? Not intending to start a flame war. This is more like a poll.
 
wklie said:
Some months ago we did some exhaustive testing before shipping Lumin X1, and that included phase inversion feature testing. A product manager thought he could not hear phase inversion. Yet when I played some bass-heavy music (e.g. Billie Jean from Thriller SACD, if I remember correctly), he was surprised that he could actually hear it, and the drum sounded so different in our Vivid Giya G2 setup.

Phase inversion can happen in some cases. Some Japanese products like Accuphase and Luxman (note: not including Esoteric) adopt XLR pin 2 cold standard, but I believe it is more common to have source products that have XLR pin 2 hot. So using these together via XLR results in phase inversion by default. Luxman has a polarity inversion button, but Accuphase does not. Then there are many albums that are recorded in inverted phase, making this issue complicated - or even worse, compilation albums may have different phase for different tracks.

Back to our tests. In the particular tracks we tried, we came to a conclusion that a correct phase gives more audible details. Yet there were also cases where some people (subjectively) like the incorrect phase more for particular tracks. (This may also be an example of that the most accurate audio presentation is not necessarily the most enjoyable for a listener's subjective taste.)

So, do you guys hear phase inversion? Not intending to start a flame war. This is more like a poll.
Click to expand...

YES.

On my RoomPlay playlist, IIRC, out of 38 tracks, 17 or so are inverted. With Audirvana, it is corrected automatically when the iTunes comment column for that track is marked ‘invert phase’.

Also works that way with Pure Music.

IMO, having it to be automatic for designated tracks is great, or at least it needs to be quick & easy to switch absolute polarity.


Sent from my iPhone using Tapatalk
 
wklie said:
Yet there were also cases where some people (subjectively) like the incorrect phase more for particular tracks. (This may also be an example of that the most accurate audio presentation is not necessarily the most enjoyable for a listener's subjective taste.)
Click to expand...
This is consistent with controlled phase tests, where listeners could detect "a difference"...but preferences, i.e. which was "better", was inconsistent.
Outside of audiophile/studiophile fantasy/imagination, there is no such thing as "accurate" stereo.:)
http://www.aes.org/e-lib/browse.cfm?elib=9136
At best, we create a rough facsimile, the mind/memory fills in many gaps and hopefully, some enjoyment is derived.

cheers,

AJ
 
Jim Smith said:
YES.

On my RoomPlay playlist, IIRC, out of 38 tracks, 17 or so are inverted. With Audirvana, it is corrected automatically when the iTunes comment column for that track is marked ‘invert phase’.

Also works that way with Pure Music.

IMO, having it to be automatic for designated tracks is great, or at least it needs to be quick & easy to switch absolute polarity.


Sent from my iPhone using Tapatalk
Click to expand...

YES indeed.

Over the years, I have observed a couple of what I find interesting things about absolute polarity:

1) Most of the people most aware of absolute polarity are serious recording engineers using somewhat minimalist recording techniques - Jim Smith, Peter McGrath and Dave Wilson to name a few.

2) The more time correct the system, the more obvious the differences.

3) In older multi-miked recordings, the polarity can indeed be preferred one way or the other as the incoming signals that were ultimately mixed did not necessarily have consistent polarity.

4) There are those who hear correct polarity (see #1) without the need for an A/B.

As always, ymmv.
 
AJ Soundfield said:
This is consistent with controlled phase tests, where listeners could detect "a difference"...but preferences, i.e. which was "better", was inconsistent.
Outside of audiophile/studiophile fantasy/imagination, there is no such thing as "accurate" stereo.:)
http://www.aes.org/e-lib/browse.cfm?elib=9136
At best, we create a rough facsimile, the mind/memory fills in many gaps and hopefully, some enjoyment is derived.

cheers,

AJ
Click to expand...

I agree to some extent - it seems to be recording (and system set-up) related.

I have NEVER had a client who could not hear the difference and express a correct acoustic polarity preference for the majority of the inverted tracks from my RoomPlay playlist. Of course, those cuts may feature transients such as from pianos, drums (most any drum, but especially rimshots, struck wood blocks, etc.), guitar plucks, upright bass (such as Brian Bromberg's Wood II, not only for the bass note attacks, but also for when he uses the fret board as a percussion instrument).

Vocals are harder to diagnose. First, the system set-up needs to be really dialed in (an exceedingly rare occasion, IME). Correct acoustic polarity seems to give some vocalists a bit more presence, as if they are more forward in the sound stage. It's easier with close mic'd recordings, IMO. With some recordings, I simply cannot be sure of which is better.

I remember the first time I noticed this phenomenon - it was 1985, and I was setting up a pair of Tympani 1Ds for a client who was coming in to my audio shop that afternoon. The recording? It was a Willie Nelson LP (sorry, I forget which one). His guitar was in opposite polarity to his voice (!)

Although I never knew how it happened, I assumed the recording engineer simply didn't insure that the guitar pick-up was the same polarity as the mic (back then, some well-liked European mics had reverse polarity to the US standard).
 
With my Horn/SET system, I find it pretty easy to hear phase inversion. Sometimes it is the electronics. my C-J MET-1 says in its user manual that it inverts phase, for example. Others don't say. I've had a Herron preamp for many years that has a phase button on its remote - easy to change from my listening chair.

Part of the problem is on multidriver speakers, I understand that sometimes one driver is phase inverted compared to another driver in the speaker. In heavily multimiked recording sessions one can easily have some instruments recorded in opposite phase to other instruments.

There is a pretty famous story of a non-phase problem that sounded like a phase problem. There is an out semi-audiophile recording of the Haydn and Hummel Trumpet concertos with Gerald Schwarz both conducting the orchestra and playing the trumpet solo. Many listeners complained that his trumpet sounded out of phase with the orchestra. Photos taken of the session explained why. Schwarz was facing the orchestra when conducting, even when he was playing the trumpet solos - so the trumpet was facing away from the microphones!

Larry
 
astrotoy said:
..

Part of the problem is on multidriver speakers, I understand that sometimes one driver is phase inverted compared to another driver in the speaker. In heavily multimiked recording sessions one can easily have some instruments recorded in opposite phase to other instruments.


Larry
Click to expand...

These are different things.

There is a difference between electrical phase (crossover to driver wiring) and acoustic phase (summed output at the listening distance). Many multi-driver speakers wired this way have no issue whatsoever showing absolute polarity differences. Others all wired in electrical polarity but with other crossover and or driver location issues have a tougher time showing absolute polarity.

The multi-miking thing ime seemed much more prevalent on older recordings when less attention (no attention?) was paid to absolute polarity. Jim's Willie Nelson experience is a perfect example.

Assuming proper choice of mike pattern for given spots, attention to polarity in set up and all instruments playing toward the capsules (as opposed to the Gerard Schwarz example), it should be pretty consistent. Then again, the best laid plans..
 
AJ Soundfield said:
An old issue, to say the least: http://www.enjoythemusic.com/magazine/bas/1208/
Click to expand...



The nice thing about absolute polarity inversion it is that it is free. Many pieces of gear offer polarity inversion switches. Worst case if the gear you have doesn't have a switch, you could flip polarity at the speaker cables (ugh). So if you feel it does matter and you can hear it, simply make the adjustment.

If you don't feel you hear it, your system doesn't show it, or it doesn't matter to you, there really is no issue at all.
 
AJ Soundfield said:
An old issue, to say the least: http://www.enjoythemusic.com/magazine/bas/1208/
Click to expand...

Shanefield reported his tests using MG-2s.

As the person who actually wrote the MG-2 set-up manual, I'd guess that he had set-up issues that would just about completely smear polarity inversion.

If the speakers were closer than 5' or so from the front wall, or if the front wall was reflective in any way, or if he allowed them to remain tilted backwards as they came from the factory, as well as many more potential set-up errors, the polarity/time component was smeared badly.

FWIW - in those days I never - not once - heard a set of Maggies set up in any audiophile's home that were anywhere close to their true potential.

If that observation held true for Shanefield, I am not surprised that he had difficulty hearing the issue at hand.

But maybe he was the lone exception in North America, and he was right, and those of us who say we have heard it for years are fooling ourselves.

Sorry, that's more than enough of this bickering from me - it's out of line with what was requested by the OP. So I'm out.

:hey:
 
A very recent example... last night... playing around with different cables on this new to me McIntosh amp and I accidently got one speaker out of phase... very obvious issue... tubby bass, while there was great channel seperation, there was almost no center fill. I am very sensitive to phasing so I knew something was not right.. tinny sounding, no body...

Found the one channel out of phase... what a difference.
 
Randy Myers said:
A very recent example... last night... playing around with different cables on this new to me McIntosh amp and I accidently got one speaker out of phase... very obvious issue... tubby bass, while there was great channel seperation, there was almost no center fill. I am very sensitive to phasing so I knew something was not right.. tinny sounding, no body...

Found the one channel out of phase... what a difference.
Click to expand...

What you describe is not absolute polarity plus or minus but channels out of phase with one another.
 
Absolute Polarity can be inverted in Roon’s DSP section. Long discussion about it in their community forum. Lots of opinions on whether it’s worth messing with or not.
 
Jim Smith said:
Shanefield reported his tests using MG-2s
I'd guess that he had set-up issues that would just about completely smear polarity inversion.

If the speakers were closer than 5' or so from the front wall, or if the front wall was reflective in any way, or if he allowed them to remain tilted backwards as they came from the factory, as well as many more potential set-up errors, the polarity/time component was smeared badly.
Click to expand...
:)
Dr Shanefield's (RIP) results were based on his tests conditions and your results were based on your test conditions.
YMMV

cheers,

AJ
 
Willco said:
Absolute Polarity can be inverted in Roon’s DSP section.
Lots of opinions on whether it’s worth messing with or not.
Click to expand...
Yep. No "Absolute" conclusions about "better/worse"....just like all controlled tests. "Controlled" with cognizance of confounders as aptly described by Dr Shanefield.

cheers,

AJ
 
How does one tell if the phase is inverted, or not? I use balanced connections between all gear, and none have a phase inversion button.
 
AJ Soundfield said:
Yep. No "Absolute" conclusions about "better/worse"....just like all controlled tests. "Controlled" with cognizance of confounders as aptly described by Dr Shanefield.

cheers,

AJ
Click to expand...

Better/worse is never the correct question in such tests. As with any valid test, ask the wrong question and the results are no better than the question. Better/worse is a secondary value judgement, not a confirmation/denial of an identifiable difference.
 
metaphacts said:
Better/worse is never the correct question in such tests. As with any valid test, ask the wrong question and the results are no better than the question. Better/worse is a secondary value judgement, not a confirmation/denial of an identifiable difference.
Click to expand...
No one is denying "a difference", not Lipshitz, Shanefield, anyone. So yes, it is 100% correct to then ask if better/worse/preferred
 
You must log in or register to reply here.
Back
Top