Real examples of incorrect polarity in recordings?

Discussion in 'Music Corner' started by Publius, Jul 22, 2005.

Thread Status:
Not open for further replies.
  1. Publius

    Publius Active Member Thread Starter

    Austin, TX
    I'm talking with somebody who believes that most/all recordings preserve correct signal polarity across all tracks. I've heard lots of anecdotal evidence that this is not the case, but to my shock, not one post or web page I've looked at has conclusively said that a CD was recorded in the wrong polarity, or that some tracks from a mixdown are in the wrong polarity (but not others).

    Can anybody here provide objective evidence for a commercial recording with wrong polarity?
  2. Doug Sclar

    Doug Sclar Forum Legend

    The OC
    My opinion is that absolute polarity in pop recordings is arbitrary. First of all there are usually a lot of tracks with little correlation and lots of processing. On top of that many mics have different polarities. There is no real standard as to whether positive pressure on the diaphragm produces positive voltage on pin 2 or 3, or at least there wasn't when I paid attention to such things. Plus mic placement can have an effect on this. There are a whole lot of variables that can screw thing up if one doesn't specifically attempt to preserve absolute polarity. Obviously most engineers will try to maintain relative polarity for stereo correlated mic pickup, but beyond that I think it's a crap shoot.

    I'd say with pure classical or jazz recordings there is a better chance that uniform polarity was maintained, but in pop multitrack recording I'd say it is very unlikely. Many people flip their playback polarity to what they think sounds the best. The problem with this is that if absolute polarity wasn't maintained, what do you listen for. Do you want the kick drum to have correct polarity if that means that other instruments don't?
  3. sharedon

    sharedon Forum Zonophone

  4. Steve Hoffman

    Steve Hoffman Mastering Engineer Your Host

    This thread is about ABSOLUTE POLARITY aka "The Wood Effect". As written about by Clark Johnson.. Not the same as being "out of phase"...

    Try this:

    Or read this from Clark via the Positive Feedback article:
    Yes, Polarity is Absolute—But Only Your Ears Know For Sure!
    by Clark Johnsen

    The topic of acoustic polarity, also known as Absolute Polarity, has cropped up again recently. No surprise—since 1962 it’s lingered in the professional literature, where most researchers claim that polarity matters greatly. But today, who cares?

    For that sorry state of affairs, you can blame the commercial audio press. For whatever reason, hardly a whiff of this vital phenomenon ever appears in those precincts. Ultimate Audio, with two feature articles, became by default an exemplar of polarity awareness—quite so, as ultimate audio cannot be achieved without it! A personal disclaimer: I have often called polarity the sine qua non of correct audio practice. As author of the only book on the topic (The Wood Effect: Unaccounted Contributor to Error and Confusion in Acoustics and Audio, ISBN 0-929383-00-1), which explains everything, I naturally applaud the renewed attention. And I remember how Michael Gindi, an Ultimate Audio contributor, once toured the Stereophile Show chanting, "If you can’t hear the Wood Effect, you can’t hear!" I expect he still stands by that, though nary a peep recently.

    Oh, the Wood Effect. Discovered by Charles Wood at the Defense Research Laboratory in 1957, it was first reported in 1962 in the Journal of the Acoustical Society of America. "Wood used as a signal a sinusoid partially clipped during half of each cycle. The resulting sound had a different timbre when the flat-topped portion was presented to the ear as a rarefaction, than it did when leads were reversed and the flat-topped portion was presented as a compression." There you have it, ladies and gentlemen: An asymmetric sinusoidal signal, presented those two dissimilar ways, with no other distortion, was proven to sound very different each way. Let’s see, what else in the sonic realm generates asymmetric signals? Well, ta da! That would be musical instruments! Which explains how polarity permeates our entire audio world, albeit through negligence, because we hear normal and inverted polarity (more accurately, compression and rarefaction) so very dissimilarly, ,yet rarely are they differentiated by us in playback.

    The abstract to The Wood Effect says it best: "Masked by random combination with other distortions in the music reproduction chain, an unsuspected major contributor has lain hidden: Aural sensitivity to ‘phase inversion’ — the Wood Effect."

    "Music normally creates compression waves. Electronics, however, often invert that natural, positive polarity to unnatural, negative rarefaction, thus diminishing physical and aesthetic impact. The term Absolute Polarity uniquely describes the correct arrival to the ear of wavefronts from loudspeakers, with respect to actual musical instruments."

    "Wrong polarity, when isolated, is obvious to everyone. Its present neglect results from habitual disregard for phase response, especially in loudspeakers."

    Phase accuracy has never much mattered to speaker designers (I might have continued), producing what Richard Greiner has termed "the crossover catastrophe," referring to high-order crossovers that blunt the polarity sense. We must also endure a haphazard engraving of polarity on delivery media, which causes variances from disc to disc, from side to side, from cut to cut, and even within a single cut. Thoroughly permeated by all that confusion, many people throw up their hands in despair and declare that polarity cannot matter. But they are wrong, and therein hangs a tale. Polarity, in its pure state, as with one musical instrument recorded on one channel (designated. a monaural phase effect, or MPE), is perfectly audible, and provably so. Reproduced incorrectly, it becomes what I call "the muffling distortion," because our ears detect an inverted leading transient and suppress the impulse response. Thus the pluck on a guitar string, so striking in real life, becomes dull, muted, and inexpressive. A musician might say, "You can’t hear the fingers, man."

    So? So, to repeat, Absolute Polarity, which earns caps by defining the only true absolute in audio, requires that the loudspeaker (whichever way the recording may be) generate a compression wave whenever the actual musical instrument does. Live music has punch, but music often lacks that aspect when heard over an audio system; that’s your polarity distortion.

    But how to explain? How does impact so often become expact? How sound becomes... muffled? Without a protocol as to how any recorded medium should store positive polarity—a one or a zero, an inside or an outside groove excursion, don’t ask the AES!—the randomness among tape recorders, studio toys and mastering gear is transferred onto delivery media in similar fashion and left for correction by... what? Only our ears! Pretty weird, huh? You actually have to listen to tell which way is right? No meters, no nothing to assist? God! What’s wrong with those audio engineers? Where are they when you really need them?

    To those of us who hear polarity, and we are legion, this whole state of affairs seems ludicrous. Over a low-phase-distortion audio system, no one ever misses the right call. Never! For example, everyone who crosses my own threshold gets it by the second try. Later, granted, people become dissatisfied with the situation and begin to berate the recording industry and the electronics manufacturers, who rarely provide compensation for polarity. Nor is the press exempted from obloquy, along with our vaunted high end "designers," who largely ignore polarity. Those listeners who know about it, however, mostly blame the press for not informing them earlier, and for being more interested in promoting new gear for sale, at the expense of basic principles.

    Readers should be grateful therefore for any assistance at all; and this phenomenon of Absolute Polarity apprises them of a major improvement, available without purchase of anything new. Better sound for free, as I have proclaimed for years. How often do you find that? (Although one could do worse than to invest in the Remote Phono Polarity Control from Lewis Labs. The wired version sells for $400, the wireless for $650. Phone: 305-667-2601).

    But as to why I’m on the case again, there are problems in those Ultimate Audio articles by Mrs. Herron and Fredell—and elsewhere too, notably in an e-zine, in a piece authored by Doug Blackburn, outstanding for its wrong-headedness. These writers (and many others, too) confuse small-case polarity with Absolute Polarity. Also, they use "polarity" interchangeably with "phase." Phase could mean any degree, while polarity denotes precisely the 180° condition. Third, they all seem to think that the Absolute inheres in recorded media, whereas in truth it occurs only at the final impingement point: your ear. To quote Mr. Herron: "Absolute polarity can be lost... and again restored in many places in the recording and playback chain." No, Keith! You mean, simply, polarity. And not "lost," either: polarity never disappears, it only reverses itself, time and again.

    Then Mr. Fredell: "Remember, a musical instrument has one polarity—the absolute polarity." No, Lars; an instrument has only its own polarity. It falls on us to recognize the Absolute at our ears, for right reproduction. Not to beat an almost-dead horse, but here’s Sedrick Harris, too, in Ultimate Audio (Fall 2000): "Note: The absolute polarity of this recording is 180° wrong." No! The Absolute is achieved only case by case; it is not written into law regarding any groove or any gear. Never mind that Harry Pearson, Robert Harley, and scores of others who should know better have made the same mistake, readers by now should recognize their common error, which allows confusion to reign.

    Confusion reigns perhaps nowhere more so than in Doug Blackburn’s peculiar column on Soundstage (October 1999). Let the record show, I once entertained Doug and his droll wife Tracy at The Listening Studio, then back home for dinner, and we got along just fine. Afterwards, he hardly spoke to me again, and published an article about the Vibraplane that, to my mind, quite misrepresented it. Later he wrote a long-winded article opposing polarity without once acknowledging my book or its arguments. I do not take that personally, but since I am about to disrupt his party, the question might arise.

    Mr. Blackburn begins his tale auspiciously enough: "When the scientist selects a test with hidden variables, the test may end up being invalidated later." Quite right, but then: "The experiments that prove polarity is audible have been incorrectly performed for decades." Very well, how? "In every case I’ve seen, these experiments are conducted by reversing the connections somewhere in the audio signal chain." All right, what’s wrong with that? "None of those methods are adequate for evaluating the audibility of polarity reversal." And why not? "When you reverse the connections at a loudspeaker... you also reverse the direction the audio signal is applied to wire in the loudspeaker."

    There you have it, ladies and gentlemen. No one who performed these experiments "ever seems to have evaluated the audibility of reversal of direction of signal travel in wires... All of them made the same mistake—assuming that polarity reversal is the cause of the change in sound they hear." So the trick’s in the metal! "This information would seem to imply that switching polarity could very well be inaudible." How to respond to what I would call errant nonsense? Well, for one thing, I am conversant with wire-direction anomalies. I keep one set of cables hanging around, not because they sound good (they don’t), but to demonstrate the aural effect of reversal, which in this case is fairly dramatic—but nothing like polarity. Doug is correct about wire directionality; where he errs is in attributing the whole of polarity audibility to that phenomenon.

    Other objections to his thesis quickly arise. For instance, what if the reversal is accomplished with an internal switch? Very little wire within a switch! Ah, but Blackburn argues that "reversing the direction the audio signal is applied" to any wire produces his effect, which means that even switching cartridge leads, or bits in the digital domain, will change the performance of all wires following. Likewise, switching even the internal loudspeaker leads right at the drivers—after all, there are still the coil windings in which the signal may be altered by those nonlinearities we (wrongly) associate with polarity. Don’t forget, "reversing the direction the audio signal is applied to wire in the loudspeaker" explains the entire phenomenon. Whatever expanse of wire may be involved following the switch point, we are assured that only it must be blamed for what we hear.

    Let’s inspect that claim more closely. Just what does the writer mean? It’s not "hot" and "cold" that are actually switched. Ground remains grounded, nor do electrons change their general direction of travel with audio polarity reversal. Wires know nothing about acoustic compression and rarefaction wavefronts. In any event, they represent the opposite sides of a sine wave slope, both of which they pass with equal ease. With signals continuously fluctuating, as musical instruments comprise a mix of both modes, are we to suppose that wires somehow differentiate these? That would be the result required by Blackburn’s astonishing, but ultimately insupportable analysis.

    I cannot imagine why he feels the need to deny a perfectly obvious and simple physical principle, thus doing Soundstage readers a severe disservice. Perhaps his confusion arises from that common misapprehension: "You really need to use recordings that are known to be polarity correct." Once again: There are no standards for "polarity correctness" on tapes, records, or CDs. None. Nada. It is a fantasy that leads you astray.

    Another fantasy?

    The estimable Keith Herron, in his article "Is Absolute Polarity Absolute?" (Ultimate Audio, Fall 2000), begins very logically, too. "Is absolute polarity audible? Absolutely!" But then immediately we are told, "But what are you really hearing when you flip that polarity switch—or reverse those speaker wires?" Uh oh! Could it be... something else again? Yes, as it happens. "Did adding the phase inverter change the sound? You bet! But it also changed the measurable distortion products." And beyond that, "Will the speaker cone go in, exactly the same way it goes out?" Those arguments have been heard before, and the only proper reply is, so what? We learn to live with electronic distortion, and in fact some speakers—ribbons, for instance— don’t even have cones, yet reproduce polarity effects clearly. Mr. Herron concludes his sadly obfuscatory article on this somewhat weasely note: "It just might be that distortion products in the recording process cancel or add differently with those in the playback process. Is there anything absolute about this? Absolutely not!"

    The final pitch

    Just a couple things to wind up. The fact that ever-popular high-order speaker crossovers perilously blunt the polarity sense was ignored in these recent articles. And, while Lars’ laudable call for panel (or remote control) polarity switches should be heeded, many such, for whatever reason, barely reveal the polarity condition, even in the vaunted digital domain. The best way to teach yourself—pace, Doug Blackburn—is to reverse all four (or eight) speaker wires, tediously back and forth, to learn the condition, then play it from there. Your rewards will more than compensate the effort, believe me! Remember, no one ever promised you that audio would be easy.

    Finally, I must dispute Lars’ list of polarities by CD label, which he first published several years ago in Fi. While perhaps useful as a rough guide, such happy consistency rarely obtains, except perhaps for some very small labels. Major labels split pretty much 50/50, as the laws of statistics demand whenever no control is exerted over a binary event. To whatever extent Lars may be right, his evaluation applies only to the opening bands; subsequent ones can vary willy-nilly.

    That very confusing 50/50 split was the central revelation of The Wood Effect, remaining as true today as ever, sad to say. For a superlative corroboration of that principle, I offer Lars’ list itself: Its 88 entries for label polarity are tellingly divided 43/45. Eternal vigilance is ever the price of the true Absolute.
  5. Zal

    Zal Recording engineer

    Brooklyn, NY, USA
    I seem to remember working on some Modern Jazz Quartet projects and finding the Vibes out of phase with the rest of the instruments. Some other times as well with piano and drums (stereo miking technics anyone?), but it's not a digital thing per say, it's just the way it was recorded, the physics of sound. Of course, the transfer is important, getting the azimuth right-on, and keeping the channels left and right, consistent with the way the vinyl was cut...
  6. sharedon

    sharedon Forum Zonophone

    Wow. Thank you, Steve!
  7. I notice this effect all the time when I'm doing needle-drops. If I compare of wav files of the vinyl to the CD, they're probably phase opposite at least 1 out of 4 times (meaning the peak on the LP is a valley on the CD -- of course you have to zoom WAY IN to see this). Which is correct -- the vinyl or the CD? Is there any sure way to tell?

    I would have to guess that short, sharp staccato sounds start as a peak rather than a valley. Would this be correct?
  8. CaptainOzone

    CaptainOzone On Air Cowbell

    Beaumont, CA, USA
    I'll venture a guess that this is the kind of thing engineers like Bill Porter paid attention to.
  9. Sean Keane

    Sean Keane Pre-Mono record collector In Memoriam

    Steve, you took the words right out of my mouth.
  10. Publius

    Publius Active Member Thread Starter

    Austin, TX
    Thanks Steve. Clark alludes to a polarity comparison by label compiled by Lars Fredell; do you know any more about that article, or what criterion was used? I'm a little unsure of checking the waveforms myself, because I can think of very few instruments or vocal expressions which have a well-defined polarity as well as a well-defined mic location.

    ... although I suppose that most drums would be mic'd on top, so that the leading waveform is always negative in pressure?
  11. Steve Hoffman

    Steve Hoffman Mastering Engineer Your Host

    Could be, but remember, most modern recordings are a mix of both.

    Can't you just flip and listen and then flip back again? Which ever way sounds the best to you should be the way you listen, correct?
  12. Publius

    Publius Active Member Thread Starter

    Austin, TX
    Well, the long story is a little complicated. :)

    Basically, not only have I heard the difference (and I think I can say one of the two sounds better), I successfully passed multiple ABX tests. However one person has criticized the test on two grounds:
    1. My system itself may be faulty and could have asymmetric distortions between positive and negative polarity
    2. Since recordings are universally mastered in correct phase anyway, and all studio gear is noninverting, the entire discussion is moot except if you somehow hold a preference for the inverted sound. (seriously, that is almost exactly what he said)
    I'm sort of ignoring point #1 for now until I find some good measuring equipment. I know #2 is wrong, but I've been having a really hard time finding real proof of it. ie, examples of recordings which are obviously inverted, concrete evidence of a studio somehow inverting a signal during production, examples of inverting equipment used in studios, etc just don't seem to exist on the Web. And since this fact-finding will eventually make its way back on HA, the evidence need to be objective, so saying "this polarity sounds better so it is probably the correct polarity" can't cut it.

    And so I come to everybody here for guidance. :D
  13. Publius

    Publius Active Member Thread Starter

    Austin, TX
  14. GT40sc

    GT40sc Forum Resident

    Eugene, Oregon

    Since recordings are universally mastered in correct phase anyway, and all studio gear is noninverting, the entire discussion is moot except if you somehow hold a preference for the inverted sound.


    Forgive me, but does this person know anything about electronics at all? Ever been in a recording studio? Looked at a signal on a scope?

    Multi-track recording is not a neat, clean, antiseptic process. It's a damn dark mysterious miracle that we get anything like music on tape.

    Go read "Gizmo" Rosenberg and meditate on the nature of the musical universe, then come back and learn the guts of a U-47.
    If Stephen Paul were still alive, he could set you straight. The truth is stranger than we can imagine...
Thread Status:
Not open for further replies.

Share This Page