I just learned from testing, and three experts on the subject confirmed, that <s>all</s> at least SOME Oktava microphones are wired with reversed polarity!!!


This polarity inversion results in the signal being 180 degrees out of phase. Pins 2 and 3 are swapped on (at least some) Oktavas. If you talk into the mic monitoring via headphones, it will sound fuller with the polarity inverted. All other mics in my experience sound fuller via that test with the polarity normal.

This is apparently "by design", since it is consistent, even though afaik no other mic manufacturer does it that way. It doesn't necessarily mean the mics suck, it just means that you should flip the polarity on them via a preamp switch or inverting adapter by default.

Why? Because when an impulse blows air onto the mic capsule, it should translate to the speaker downstream blowing air in the same direction onto the listener's ear. Blowing air shouldn't translate to air being sucked back. I have a Chesky test CD (JD37) that shows just how disappointing a trumpet sounds with the polarity reversed...as if the trumpeter is sucking the air back into the trumpet (the disc is from the 80's and is screwed up in that they also reversed channels L and R, but importing the file into a DAW and properly inverting it yourself demonstrates the issue). This same degradation happens to a lesser extent on any signal. It's actually easy to verify this on a good enough monitoring system with a good enough signal, but many people have yet to realize that.

Scott Dorsey (of the famous "Dorsey MC-012 mod") writes on rec.audio.pro, 10/19/2002:

But I do worry about it, Scott! And I'm not sure what other mics have this problem. It is a problem. One that's very easy to fix of course, you just have to remember to do so.

Interestingly, to my knowledge, correcting the polarity isn't a standard part of any of the Oktava mods that are available. I just confirmed that my 219 <s>and four different 012's all</s> (EDIT: apparently the 012's I tested were actually pin 2 hot) had the polarity reversed. I have to wonder whether Oktava's reputation could have been subtly damaged all these years by this odd reversal. Why are they being made, and left, inverted?

Thanks for the info. I'll be checking mine...

There was a time before the concept of "positive pressure / positive polarity on pin 2" was the "standard" Western pro audio convention it is now.

I mean, isn't it more intuitive that positive pressure should result in a positive pulse on the highest numbered pin of an XLR? - As in: pin 1 = ground, pin 2 = minus and pin 3 = hot? I would argue that this should be the standard. But hell, here in America we use still use goofy pounds, inches and feet instead of the Metric system which is really intuitive.

Those wonderful Ruskies thought pin three should be hot and I can see their logic.

Anyway, phase is a relative thing. You should always flip it around till it sounds best whether in a solo mic "Elvis / Geetar" deal or in a 12 mic mega drum mix.

This is just one of the reasons I love Oktava - they're different, they do their own thing, they don't copy and people can deal with it by using their own ears.

Who cares. easy to fix, but good to know I guess. If you didn't notice it before, why should it matter now. Flip phase on all channels anyway and see what sounds better.

My life is out of phase, so Oktava mics fit it perfectly.

I had the same issue with a RED mic I purchased. I only discovered it because I was doing an A/B comparison with another mic and noticed the phase difference. I then tested it against all my other mics (GeFell, BeyerDynamic, SP, etc) and it was the only one with reversed phase. I contacted the RED company, and after sending me 3 mics, they finally got me one that was phased correctly.

...I will point out that they shipped each of the mics to me via "overnight" service and did not charge me a dime. They compensated me very well for helping discover the little issue.

Fukkin' awesome mic and company!!

(...you may want to check any RED mics you own, btw)

-mr moon

Also of note, if you use the rear of a fig-8 ribbon mic for tonal variation, you should technically flip the polarity.

There is an audible difference when flipping the polarity of a fig-8 room mic when mixing with a close mic. This can be used creatively, where you have a choice of the direct sound or the first reflection being in phase with the close mic.

Anyway, as far as the polarity thing, I don't always think about what the "correct" polarity of a given channel should be. I push the polarity button in and see if it sounds better. If it does, it stays in.

OK I've done further tests, and I'm trying to get some definitive answers on this... Ideas of what I'm seeing are coming and going by the minute...I'll eventually get it hang in there... =)

In any case, I think we should all know the polarity of our mics! This will at the very minimum inform us when we're doing multi-mic setups. It will also greatly affect a vocalist's self-perception when they are hearing themselves in headphones. While I'm all for listening to every option at all times, in practice it's always good to know there is an issue and what the predicted outcome ought to be so that you can narrow in on something quickly.

I now have a more definitive test via hand-drawn waveforms sent out speakers and back into the mic. This is revealing lots of interesting things...

In some cases, condenser mics sounded fuller in the headphone mix with the polarity inverted, yet they were still pin 2 hot. This means that the polarity that goes into the cans might not necessarily be the best polarity to go onto tape. I'm not sure why this was...the mics are the same distance from me, but perhaps they have different response times.

That effect fooled me into thinking the MC-012's were pin 3 hot as well. My new test seems to indicate that those mics are pin 2 hot, at least my 1999ish TSR MC-012. So only the Mk-219 is pin 3 hot in the locker I have in front of me, however, I did see reports via Google that an MKL-2500 was pin 3 hot from the factory.

Note that inverting the polarity inside the mic is merely a case of resoldering two wires. Delicate work in many mics (and certainly the tiny wires in the MK-219 I'm not touching) but well feasible.

My MK012s have a reversed phase too (compared to some other mics I have, e.g. Shure, Schoeps, Audio Technica...).

My Mk012's are reversed as well.

It is better to blow than to suck.

Jim Williams
Audio Upgrades

Another thing to keep in mind is that some devices (mixer, headphone amp, FX processor) may invert polarity between their input and output also. They should get a ruler whacks to their hands for doing this, but it isn't unheard of.

So do we have to switch it internally or do we just flip phase?

You can just flip phase. I've put a note in the mic's case to remind me.

OK so after sleeping on it I realized that an asymmetrical sawtooth wave would be a perfect test of polarity. You can generate one and then invert it and that will reveal all. I may have known all of this ten years ago and forgot it!

Doing so revealed the polarity of the mics instantly (by playing the two waves out a speaker and into the mic, then zooming in on the waveforms). Of course the sawtooth isn't perfectly replicated after going through a speaker, the air, and the mic. But it's clearly legible. Again, my Oktava MC-012's, bought from The Sound Room as a FMSP in the late 90's, were pin 2 hot like the rest of my mics, but my MK-219 bought from Banjo Mart in the mid-90's was pin 3 hot.

I also discovered that the Mixcubes I had recently wired up were wired inverted...durrr...nothing wrong with the Mixcubes, I just screwed the wrong wire into the wrong terminal. Nice to fix that!

Of most interest was the readily apparent difference in sound of the sawtooth and its inversion. The saw that diagonals up and then immediately drops down was clearly bassier when heard through speakers than its inversion (which diagonals down and then immediately goes up). I think this makes sense when you think about the physics of the drivers.

However, this difference varied with distance from the speaker. Since my reflection points (L, R and above) are all treated with GIK 2" panels the difference was clearer in the mix position. This may be a good test to see how effective the monitoring setup is. I think you ought to be able to hear a difference...but maybe with the ideal speaker, you couldn't hear a difference?

However, there was no difference between the two saws visible to me on a spectrum analyzer! Either before or after going through the speaker/air/mic! This is an example of how important the ear/air/electronic boundary is to what we hear. Since most waveforms are asymmetrical, this issue doesn't just go away. The polarity we record and play back is critical in an absolute sense, not just a relative one.

It's easy to create your own test file and test your equipment for this issue. I used a 200Hz saw wave generated by Pro Tools Signal Generator and then duplicated it and used the Invert audiosuite function. Shuffling them side by side (trimming to zero crossings) and looping playback let me hear, and record through mics, the different polarities. The waveform display looks somewhat like a DC offset making it easy to see the polarity coming back in, and you can also zoom in.

the alan smart C1 was pin 3 as well.
i remeber constantly trying to interfaceit with various studios and having to get the xlr phase barrels.
i finally flipped it internally.

I think it shows your speakers are not ideal. In fact, only few speakers are close to ideal.

Doesn't this just show room issues?
With the ideal speaker, you wouldn't hear a difference, at least not on a continous wave: Our ears can only "measure" the frequency, not the absolute phase.
Might be different with transients, I'm not really sure about that.

But the point peeder is making is regarding a saw (asymmetrical) wave. I did a quick and simple experiment. Created a simple saw wave on my SE-1x synth. Put it through a mixer channel, played a sequence and switched the phase button in and out. In phase, lot's of bass and top, out of phase, less bass, more mids, very "nasally" sounding.

There are acoustic intstruments that create asymetrical sound waves, such as brass, reeds, etc. Mic polarity can be a noticable factor when recording such instruments, even on a solo, mono track.

What difference does it make if a unit is pin 3 hot, if you send a balanced signal through? None.


THE REEL THING - Real Music Production

heh

"Balanced" doesn't mean a signal somehow has no polarity!

Place a pin 3 hot and pin 2 hot mic as close together as possible, facing the same way, and record something into both at once.

Check mono compatibility.

What balancing does is it inverts the signal, sends it simultaneously on a separate (but tightly aligned) wire, and then re-inverts it to double its signal at its original polarity. This nulls most noise it picked up along the way, but it re-constructs the signal at its original polarity!

The point made by The Reel Thing is correct, if a little out of place in a thread about mic polarity! Which pin is hot makes no difference on a signal that just passes through a box. If the box has inserts or direct outs in the path then it will be a (potential) problem.

That's sort of like saying "If I write something backwards, it'll still come out backwards whether I fax it or xerox it or print it."

I don't want to make my listeners read something backwards!

!ti t'nid ,KO mih rof dekroW ?sdrawkcab nettirw saw taht mubla munitalp a dah icniV aD taht tuo tniop uoy dluoW ?did I fi ti ekil uoy dluow woH

Yes this is a problem. Absolute polarity matters (do the saw test) as does relative polarity (do the mono compatibility test). Are you grasping that?

Put it and another mic on the same source, zoom in and check to see if they are reversed from each other. I've used an Oktava that was phase reversed as well. I just thought it was a mistake, and I'd phase flip on input whenever I used it. It still sounded fine.

The QC on the Oktavas left a lot to be desired. I remember GC was taking a lot of them back. If you have a good one, they're happening -- they just had some duds.

A saw is not as asymmetrical as you might think. And it *is* continuous in that it doesn't have any noise component in the beginning (like a trumpet might have).
What you did when putting that saw through your mixer and speakers was just finding out that your mixer and speakers are not ideal. As I pointed out in my last post, it is quite normal for many speakers not to be ideal, so this doesn't mean your equipment sucks. But your result does not affect the theory behind it, which has been proven experimentally with closest-to-ideal speakers. If you can hear a difference between inverted and non-inverted phase in a SINGLE and CONTINUOUS signal, it's some component in the signal chain. It's not the ears reacting differently.

When an instrument creates an asymmetrical wave, this sound isn't just a biassed saw or other continous thing. It's the attack (ie. lots of transients and noise components) which can maybe make a difference, and this should be examined in experiments as has been the hearing of continous signals.
We need to always remember the difference between theory (sine/saw/square/white-pink-brown noise) and practice(music/speech/nature sounds/etc.).

A saw wave has a gradual rise and a sharp decline or vise versa. It is most definitely asymetrical, and most definitely has a different tonal characteristic when it is phase reversed. As I said, I created a saw wave on my analog synth and reversed the phase. the difference is NOT subtle, whether through the monitors or through phones. I am quite aware of the difference between theory and practice. My findings are based on practice. If I am micing my synth through a bass amp, then the polarity of the mic becomes a factor.

I don't understand your argument about the speakers not being perfect. Until the end listener either can except digital signals directly to his brain, or every listener has a "perfect" listening enviornment, then phase DOES matter with regard to the tone of a solo mono track.

"A saw wave has a gradual rise and a sharp decline or vise versa. It is most definitely asymetrical, and most definitely has a different tonal characteristic when it is phase reversed."

Kinda sounds like you are comparing apples to oranges here with respect to polarity of a saw wave. The transient is no different whether the wave is swinging positive or negative. The DV/T is the same regardless of polarity. If you are saying that when you reverse the polarity of your mic, the waveform becomes reversed, putting the transient at the end of the period, you have much bigger issues than the polarity of the signal. At that point you are talking about time travel.... And besides - do you really know how many inverting stages are between your mic and your recorder? If it is an odd number, (some equipment is and some isn't), if the polarity of the mic is reversed, it is actually hitting your recorder at correct polarity. That is the main reason most pres have a polarity reversal function in the first place.

I am not saying that the wave reverses when the polarity is reversed. I am saying that a saw wave is asymetrical (pkautzsch eluded to the fact that it wasn't), and I am saying that there is a definite tonal difference when reversing phase on some mono signals. Granted, it may be because the speaker itself has an asymetrical response by design, but there is an audible difference which I hear in real world applications.

My argument is that this phenomenon exists, and as long as the end listener is using the same asymetrical speakers as I am, then phase matters to me. This is why I check both positions of the polarity button on every track and if I hear a difference, I choose the one that sounds best to me.

"What balancing does is it inverts the signal, sends it simultaneously on a separate (but tightly aligned) wire, and then re-inverts it to double its signal at its original polarity. This nulls most noise it picked up along the way, but it re-constructs the signal at its original polarity!"

Almost... It doesn't really re-invert anything. The reason that a balanced signal voltage doubles is because the transformer or opamp at the receiving end is picking up the DV/T as the wave swings across 0VDC or whatever DC bias the circuit is operating at from its positive swing on the one wire, to its negative swing on the other. Unbalancing the signal at that point would provide only one-half the voltage because it is only the positive or negative swing on one wire relative to the zero point alone....if that makes any sense.

Cool.....as long as you replace the word "phase" with "polarity," I can accept that! It was unclear from peeder's post about the backwards writing what exactly was being discussed...

Not sure where you're going with this...I think you're missing the forest for the trees. Answer me this: what is the point of balancing, if not to null out noise picked up during transmission? And how would one null that noise other than to blend two inverted signals (one of which was pre-inverted so it would effectively double via constructive interference)? And will a signal sent over a properly balanced connection be re-constructed at its original polarity, at the opposite polarity, or will it have "no polarity"? heh