Absolute Polarity in Mastering

[clarkjohnsen]

Originally Posted by Lupo

You should try measuring with the diaphragms inside your head. It's simple if your playback system is up to scratch. The correct polarity (PIPO) sounds bigger and better.

I had written, Not only that, but Track 1 is presented in what the producer calls "Inverted Polarity". I did not argue "correct", but only questioned the logic of his choice.

So why are the tracks alternated?

clark

[24-96 Mastering]

Originally Posted by clarkjohnsen

What needs to be pointed out, if not spelled out, is that the device does indeed work.

When used as instructed, it works as intended. Most of the time.

I have used it, as have others. Have you?

Yes, I have worked with a handheld polarity tester. I assumed my writing about it would imply that.

because when you write "complex/arbitrary signal" you pick the worst case

The issue is not that I'm presenting a worst case (I am not), the trouble is that polarity testers are expecting a best case. The tester is designed to test speaker / driver polarity using a specified test signal, it's not designed to test polarity of a complex signal, playing through a speaker.

In my experience, a polarity tester is very easily confused, even when using the specified test signal. I'm confident that anyone regularly working with those testers will have made the same experience at some point. Feeding it with a much more complex signal, i.e. music (recorded in a room, played back through a speaker array in another room), and expecting to get a reliable results seems very optimistic, to put it mildly.

[dcollins]

Originally Posted by 24-96 Mastering

In my experience, a polarity tester is very easily confused, even when using the specified test signal. I'm confident that anyone regularly working with those testers will have made the same experience at some point. Feeding it with a much more complex signal, i.e. music (recorded in a room, played back through a speaker array in another room), and expecting to get a reliable results seems very optimistic, to put it mildly.

I think is was established long ago in this thread that the results are a reliable and repeatable 50%.


DC

[24-96 Mastering]

Originally Posted by dcollins

I think is was established long ago in this thread that the results are a reliable and repeatable 50%.


DC

Indeed. Silly me

[Nordenstam]

Originally Posted by clarkjohnsen

I had written, Not only that, but Track 1 is presented in what the producer calls "Inverted Polarity". I did not argue "correct", but only questioned the logic of his choice.

So why are the tracks alternated?

I take it that you've deemed the tracks to be alternated by the (incorrect) usage of your polarity testing device?

Let me suggest another method: ears! If you have a polarity invert switch in your monitoring path, hearing both ways should be easy. Pay attention to body and impact, the realness of the sound. It should be obvious that the tracks are PIPO. Can give you specific examples of which sounds to listen to and what they should sound like, but I don't think it'll be necesarry. If you disregard what the measurement tool tells you and try to listen, I think you'll find, like me, that the tracks are all having the same polarity and that is PIPO. (only tried to listen to the first 4 or 5 tracks or so, please give me a word if you find there is something odd with some higher numbered track).

[bob katz]

Originally Posted by clarkjohnsen

What needs to be pointed out, if not spelled out, is that the device does indeed work. I have used it, as have others. Have you? I must assume not, because when you write "complex/arbitrary signal" you pick the worst case; in fact, in many if not most musical presentations there are instruments exposed that give a true reading. (Granted one must know which few by nature create rarefactive transients.)

The device, au contraire, is fully capable of providing reliable results. Try it, you'll see.

clark

If this device actually worked then you would not be seeing 50-50 results on "most CDs". You're preaching wrong information to this crowd and misinterpreting your
"measurement instrument". After having studied thousands and thousands of combined waveforms, as we do every day, mastering engineers on this forum could easily tell you that polarity ambiguity of a composite waveform is a fact of life: the combinations of output of the various musical instruments CAN NEVER be "in phase" because by their nature they are separate and distinct instruments. In viewing composite DAW waveforms, I find it EXTREMELY rare to discover in any recording an isolated single instrument whose unique waveform is observable and not tainted by the room acoustics or the residuals of other instruments. In addition, I strongly doubt that a "go-no go" polarity tester could react to such a complex (non-standard) wave and produce consistent results. And in fact if you were to edit a portion of any recording so as to find (hopefullly) an isolated instrument, even the point at which you begin and end playing this non-standard test signal would confuse your instrument which is designed to detect a standardized test signal. All you have to do is start looking at actual waveforms of the CDs you have been testing and discover the near-impossibility of your measurement task with your "test instrument".

Again Mr. Johnsen will turn these words around and attempt to put them to his advantage by using propaganda and repetition. His modus-operandi is to use repetition instead of argument and conveniently disregard any arguments that do not serve his cause. For example, statements in this thread confirming that acoustical-electrical polarity standards do exist. He disregards them, uses propaganda and repeats his own position. Unfortunately, "Reiteration" is not the same as "argument". Repetition is the sign of a propaganda artist, or someone who will never learn, or someone who simply likes to argue for argument's sake, or more likely, someone who always needs to bolster his own position for his own comfort and never will see the light.

Game, Set, Match. It's over.

BK

[philip]

Originally Posted by bob katz

If this device actually worked then you would not be seeing 50-50 results on "most CDs". You're preaching wrong information to this crowd and misinterpreting your
"measurement instrument". After having studied thousands and thousands of combined waveforms, as we do every day, mastering engineers on this forum could easily tell you that polarity ambiguity of a composite waveform is a fact of life: the combinations of output of the various musical instruments CAN NEVER be "in phase" because by their nature they are separate and distinct instruments. In viewing composite DAW waveforms, I find it EXTREMELY rare to discover in any recording an isolated single instrument whose unique waveform is observable and not tainted by the room acoustics or the residuals of other instruments. In addition, I strongly doubt that a "go-no go" polarity tester could react to such a complex (non-standard) wave and produce consistent results. And in fact if you were to edit a portion of any recording so as to find (hopefullly) an isolated instrument, even the point at which you begin and end playing this non-standard test signal would confuse your instrument which is designed to detect a standardized test signal. All you have to do is start looking at actual waveforms of the CDs you have been testing and discover the near-impossibility of your measurement task with your "test instrument".

Again Mr. Johnsen will turn these words around and attempt to put them to his advantage by using propaganda and repetition. His modus-operandi is to use repetition instead of argument and conveniently disregard any arguments that do not serve his cause. For example, statements in this thread confirming that acoustical-electrical polarity standards do exist. He disregards them, uses propaganda and repeats his own position. Unfortunately, "Reiteration" is not the same as "argument". Repetition is the sign of a propaganda artist, or someone who will never learn, or someone who simply likes to argue for argument's sake, or more likely, someone who always needs to bolster his own position for his own comfort and never will see the light.

Game, Set, Match. It's over.

BK

interesting, so there is no right or wrong. isn't that what I told you six pages ago?

[UnderTow]

Originally Posted by philip

interesting, so there is no right or wrong. isn't that what I told you six pages ago?

No. I do not believe that Bob's post refers to audibility. Bob is referring to the fact that the (not so) SMART instrument is expecting three pulses with specific polarity in a certain order and a certain timing. This (not so) SMART instrument is utterly and totaly overwhelmed by the complexity of a musical signal so it gives a random reading. 50/50 for each polarity.

This 50/50 reading in no ways indicates anything but misuses of an instrument.

Alistair

[clarkjohnsen]

Originally Posted by Lupo

I take it that you've deemed the tracks to be alternated by the (incorrect) usage of your polarity testing device?

You are incorrect sir.

Let me suggest another method: ears!

You are correct sir.

Every track that I've ever tested for its polarity, has been marked. Did it all by ear! Then when the tester fell into my hands I used it to confirm my findings -- which as it happened had been correct.

................It should be obvious that the tracks are PIPO.

PIPO -- Polarity In ≠ Polarity Out. Perhaps you missed the ≠ ?

........... If you disregard what the measurement tool tells you and try to listen,

As explained, the listening (as always with me) came first.

I think you'll find, like me, that the tracks are all having the same polarity.

Are you discussing the Chesky disc I referenced, the one mastered by Mr. Katz? Or just any old disc?

(only tried to listen to the first 4 or 5 tracks or so, please give me a word if you find there is something odd with some higher numbered track).

There certainly is something odd there! The polarities continue to vary. But then, that really isn't odd at all, but quite expected.

clark

[clarkjohnsen]

Originally Posted by bob katz

If this device actually worked then you would not be seeing 50-50 results on "most CDs".

You are incorrect sir.

First, a minor point. I did not say that most CDs exhibit this variance. I said that overall, they exhibit a 50/50 split of polarity. Not only that, I included LPs, tapes and even 78s.

Now: It's because the device actually does work, that I was able to confirm the results I had found earlier by ear; please read my reply to lupo.

You're preaching wrong information to this crowd and misinterpreting your "measurement instrument".

I am interpreting it exactly correctly sir.

After having studied thousands and thousands of combined waveforms, as we do every day, mastering engineers on this forum could easily tell you that polarity ambiguity of a composite waveform is a fact of life: the combinations of output of the various musical instruments CAN NEVER be "in phase" because by their nature they are separate and distinct instruments. In viewing composite DAW waveforms, I find it EXTREMELY rare to discover in any recording an isolated single instrument whose unique waveform is observable and not tainted by the room acoustics or the residuals of other instruments.

Polarity is a Monaural Phase Effect (MPE) as defined in an early JASA article on the topic. Consider a single-channel recording of, oh, a piano. No tainting by other instruments. Are you telling us that no polarity is noticeable?

That would just be wrong!

Ok now add a clarinet, say. Still noticeable, or not? Add a bass. Now what? The fact is that some instruments stand out as polarity-bearers and they can be selected, by ear and by tool. The other instruments, far from adding confusion, tend to confrim the reading -- they all (or mostly, there are exceptions) produce transients and these transients are aligned.

............. Again Mr. Johnsen will turn these words around and attempt to put them to his advantage by using propaganda and repetition. His modus-operandi is to use repetition instead of argument and conveniently disregard any arguments that do not serve his cause.

You're right about the repetition, but what else can one do to beat this concept into skulls?

For example, statements in this thread confirming that acoustical-electrical polarity standards do exist.

No sir, that's asserting that they exist.

He disregards them, uses propaganda and repeats his own position. Unfortunately, "Reiteration" is not the same as "argument".

Don't look now but... you're repeating yourself.

............Game, Set, Match. It's over.

Not quite. We see no reply to my finding that the first six tracks on a disc that lists Bob Katz as the mastering engineer, are of alternating polarity; and that the lead-off track is in the polarity that's defined in the booklet as "inverting".

clark

[clarkjohnsen]

Originally Posted by philip

interesting, so there is no right or wrong. isn't that what I told you six pages ago?

There is no right or wrong, only when considering which way an acoustic transient should be inscribed on a recording medium. There is right or wrong, when considering whether the reproduced transient matches that of the original instrument.

That's been my very point, for pages.

clark

[Nordenstam]

Originally Posted by clarkjohnsen

Are you discussing the Chesky disc I referenced, the one mastered by Mr. Katz? Or just any old disc?

The number one in the chesky sampler and test disk series. It's hard to argue using ears against ears, but I do have a quite nifty listening room and it's no doubt in my ears - most of the sounds on the first tracks does sound fuller and better and more real with the listening path in non-inverting mode. Perhaps your ears are different?


Had a look at the trumpets. On this page, two pictures of typical trumpet waveforms can be found.

The D waveform in this picture is trumpet:


and the E waveform in this picture is a french horn:


Notice that they both have a fast shift from positive to negative, a rising slope from negative to positive with some informative wiggles, another fast drop from positive to negative, etc. The waveform on track 14 on the Chesky disk, the correct polarity, looks the same. The tone also sounds more real, a bit fuller. If your system consistently sounds better on track 16 than on track 14, perhaps you should try to use the polarity tester as it's intended - with the test tones.

[clarkjohnsen]

Originally Posted by Lupo

The number one in the chesky sampler and test disk series. It's hard to argue using ears against ears, but I do have a quite nifty listening room and it's no doubt in my ears - most of the sounds on the first tracks does sound fuller and better and more real with the listening path in non-inverting mode. Perhaps your ears are different?

I wasn't speaking of sound quality. Of course music sounds better when played with regard for Absolute Polarity! I must assume that Track 1 reproduces correctly in your system, which I gather has a fixed (i.e. unadjustable) polarity. But then you're faced with Tracks 2, 4, and 6.

Nor have you (or anyone) explained why the polarity of Track 1 should be what's called "inverted" later on.

For what it's worth, my polarity calls are made on a system that employs "minimum phase" loudspeakers -- single-order crossovers. Anything higher pretty much destroys the polarity sense.

................If your system consistently sounds better on track 16 than on track 14, perhaps you should try to use the polarity tester as it's intended - with the test tones.

First, to repeat, I always use my ears first. Second, one of the intentions of the SMART device was for the very use I put it to.

clark

[UnderTow]

Originally Posted by clarkjohnsen

First, to repeat, I always use my ears first. Second, one of the intentions of the SMART device was for the very use I put it to.

The website says otherwise.

Alistair

[Audiop]

Originally Posted by clarkjohnsen

For what it's worth, my polarity calls are made on a system that employs "minimum phase" loudspeakers -- single-order crossovers. Anything higher pretty much destroys the polarity sense.

Have you measured and confirmed that the speakers actually are min-phase or do you assume that becasue of electrical single pole filters in the x-over?

With most speaker-drivers a first order electrical filter will give a high order acoustic slope with phase distortion.


/Peter

[clarkjohnsen]

Originally Posted by UnderTow

The website says otherwise.

Ah, the website. Yes. Websites -- fount of all truth.

An engineer on the telephone however says that if one were careful one could use it for this other purpose.

And it works!

clark

[clarkjohnsen]

Originally Posted by Audiop

Have you measured and confirmed that the speakers actually are min-phase or do you assume that becasue of electrical single pole filters in the x-over?

Measured? Not personally, no. But I accept the manufacturer's claim. Not only that, polarity is truly easy to hear over them.

With most speaker-drivers a first order electrical filter will give a high order acoustic slope with phase distortion.

More to the point, all drivers are wired in phase and sound together as one -- no push-pull effect that distorts the polarity.

clark

[philip]

Originally Posted by clarkjohnsen

There is no right or wrong, only when considering which way an acoustic transient should be inscribed on a recording medium. There is right or wrong, when considering whether the reproduced transient matches that of the original instrument.

That's been my very point, for pages.

clark

sure, and I think this whole thread proves my point prefectly.

[24-96 Mastering]

But I wore the juice!

[UnderTow]

Originally Posted by clarkjohnsen

An engineer on the telephone however says that if one were careful one could use it for this other purpose.

How amazing! I believe I found a transcript of your telephone conversation on YouTube!

YouTube - Monty Python - Fish License

All in good fun of course. Thanks to Andreas for pointing me to this video (On a totally unrelated subject).

Alistair

[Audiop]

Originally Posted by clarkjohnsen

Measured? Not personally, no. But I accept the manufacturer's claim. Not only that, polarity is truly easy to hear over them.

Most manufacturers are not fully aware of what they are doing and lack a lot of knowledge about the technlogy behind speakers. Would you mind tell me what brand of speakers you use?

The fact that you can easily hear polarity flipping may as well be an indication of high asymetric distortion levels of the specific speakers and may have nothing or little to do with the acoustic summation of the drivers in the design.

More to the point, all drivers are wired in phase and sound together as one -- no push-pull effect that distorts the polarity.

clark

That's not the way it works. Speakers can have all drivers wired with same polarity but still have gross phase distortion that mess up the time domain seriously.. making it a "non-min-phase design".

Speakers with first order electrical slopes use to have such distortion in the time domain.


/Peter

[Nordenstam]

Bumpy for frustration relief.. Just had a DOH moment! Been struggling with an album for days. Was starting to seriously doubt the skills as it just didn't work at all. It had this weird sound no matter what. Until finally remembering the polarity switch! Bam! There it was. All punchy and stuff. Confirmed by looking at the files, drum transients does indeed poke the wrong direction. It literally sucked where it should have exploded! That also explains why the never-used "wrong" setttings for once worked well on the asymmetrical wave shaping used to warm things up. Recalling, inverting, redoing the distortions and re-recording at the moment.. This will keep me polarity paranoid for a while!

post number 909!


edit ps: Time to rename the monitor controller. Invert/normal doesn't quite describe it. The states should read Blow, and this is literally, Suck!

[philip]

Originally Posted by Lupo

Bumpy for frustration relief.. Just had a DOH moment!

that's a pity, I think Peter summed up the thread perfectly.

Originally Posted by Audiop

The fact that you can easily hear polarity flipping may as well be an indication of high asymetric distortion levels of the specific speakers and may have nothing or little to do with the acoustic summation of the drivers in the design.

[dcollins]

Originally Posted by Edward Vinteina

Dont be so ridiculous.

"He is an halibut."


DC

[Nordenstam]

Originally Posted by philip

that's a pity, I think Peter summed up the thread perfectly.

Here's a visual example from the album in question:



What makes you assume that it shouldn't sound different with a polarity inversion?

[pan]

Today, I brought new Monitors in the Studio, set them up and tried to evaluate them by switching between 3 different pairs, 2 actives (1 and 3) and one passive Speaker (2).

Well - the new toy on 3 somehow sounded strange, when switching between sets. The whole Image changed in a way of "perspective"... but only between this pair compared to the others - when I switched the phase in both channels on No.3, the "shift" was gone.

Yes, all amps are wired balanced Pin2 hot. And I will wire the new pair with reversed polarity. By chance the 2 pairs I used to use for some time were wired correctly. If I wanted to have repeatable results in absoloute phase, is this the tool used to measure the polarity:

http://www.gearslutz.com/board/1395485-post2.html ???

[pan]

... the Geekslutz just had the conversation 2 months earlier:

EV RE-20 Repair Issues (Self Repair)

[oldeanalogueguy]

Originally Posted by Lagerfeldt

Playing around with polarity on some asymmetrical waveforms made me think about absolute polarity in mastering.

I was trying to explain the phenomenon to a student and failed miserably in explaining how to detect it (I can explain why), except that 1) it can only be determined by ear, and 2) you need to confirm that it's not a monitoring issue first.

How often do you guys check and switch for the best polarity setting in mastering? I use the word "best" intentionally, since it's not always a case of simply being "different".

On some tracks it makes almost no difference, while it can make a huge difference on other tracks, e.g. a dance track with a saw bass or whatever.

no such thing
delay 180degrees is the same
as flipping the polarity
does a delay in the whole signal sound different
no effin way !!!

do you think because your cd player waited a few millisecs
before it started that somehow the sound will be different
no effin way

you just had to wait a few milliseconds to hear the same thing