Flipping phase on vocals in the mix

Recording Techniques, People Skills, Gear, Recording Spaces, Computers, and DIY

Moderators: drumsound, tomb

User avatar
Marc Alan Goodman
george martin
Posts: 1399
Joined: Tue Oct 28, 2003 7:57 pm
Location: NYC
Contact:

Post by Marc Alan Goodman » Mon May 14, 2012 8:55 am

vvv wrote: In other words, if something on Track A is on the beat, and something on Track B is on the beat, you might want them both moving the speaker in the same direction on that beat.

Now, feel free to make fun of me but please have mercy on my moms on this Mutha's Day. :twisted:
Or you may want them moving in oposite directions so that the cone doesn't have to extrude as far to produce the sound :)

Both situations have their advantages.

As far as the op, I'll stick with my original thought: if it sounds good, do it!

Andy Peters
re-cappin' neve
Posts: 602
Joined: Mon Mar 12, 2007 8:45 pm
Location: Sunny Tucson

Post by Andy Peters » Tue May 15, 2012 12:29 pm

losthighway wrote:I have always understood phase to only be relevant in so far as it is interacting with other elements. Re: Is there headphone bleed?

I have a hard time understanding a sonic difference flipping phase on something that was recorded in isolation unless there was some weird problem like DC offset.
The problem is that people keep using the word "phase" when they really mean "polarity."

The button on the console with the phi symbol flips the polarity. Flipping "phase" requires some more electronics, and of course it requires one to think about phase with respect to frequency.

Remember that phase is a time-domain concept. Consider that you have an identical signal fed to two speakers. One speaker is 5 feet away and the other is 20 feet away. When you listen to both speakers at the same time (with levels adjusted so at your head, the two are the same), you hear the sound from the close speaker first (obviously). There is a phase difference between the two signals, which is a function of frequency. At some frequencies, the signals are in phase, meaning they combine coherently and you get a 6 dB increase in level (two signals of the same level added together give a 6 dB bump). At some other frequencies, the two signals are exactly out of phase, meaning they cancel completely. At all other frequencies in between, you have some amount of reinforcement or cancellation.

If you look at a frequency-response (amplitude) graph of the two signals combined, you'll see the peaks and nulls and it looks like a comb, hence the term "comb filtering," which describes what you hear.

If you change the distance of the speakers, or if you move your head around, you change the time-of-arrival difference, and thus you've "shifted the phase."

So, sure, flip the polarity and see if it sounds better or difference. But remember you are not changing the phase. You can change the phase by moving the mic with respect to another mic, or by electronically delaying the signal with respect to another signal.

-a
"On the internet, nobody can hear you mix a band."

Seditionary
audio school graduate
Posts: 19
Joined: Tue Oct 27, 2009 6:12 am
Location: Los Angeles

Post by Seditionary » Tue May 15, 2012 12:39 pm

also -- a good test is to solo the vocal. if it has more low end with the phase flipped... might be out of phase in the first place. if it has less low end with the phase flipped... well then you are flipping it out of phase.

is it possible you are using some gear that is 3 pin hot? we had a lot of 3 pin hot gear i found and had to modify to be 2 pin hot to not flip the phase. you might be compensating for something else going on in your chain?

cheers.

ashcat_lt
tinnitus
Posts: 1094
Joined: Thu Jan 04, 2007 1:54 pm
Location: Duluth, MN
Contact:

Post by ashcat_lt » Tue May 15, 2012 2:19 pm

Seditionary wrote:also -- a good test is to solo the vocal. if it has more low end with the phase flipped... might be out of phase in the first place. if it has less low end with the phase flipped... well then you are flipping it out of phase.

is it possible you are using some gear that is 3 pin hot? we had a lot of 3 pin hot gear i found and had to modify to be 2 pin hot to not flip the phase. you might be compensating for something else going on in your chain?

cheers.
How's that then? If any track, when soloed, has a different frequency response in one absolute polarity compared to the other then there's a problem in your system somewhere. I'm almost willing to accept that some folks can hear the difference in whether the initial attack pushes the speaker toward or away from you (I know that I can't, but...), but it just plain can't change the overall frequency response without interacting with something else.

Seditionary
audio school graduate
Posts: 19
Joined: Tue Oct 27, 2009 6:12 am
Location: Los Angeles

tests

Post by Seditionary » Tue May 15, 2012 7:59 pm

whichever position has most low end is the one that is in phase. if the sound is thin... then you can assume the polarity is not correct.

mic test --

take a 2nd mic that works -- move it next to the mic you are wanting to check phase on. put both mics together so that their capsules are touching and speak into them. try to give them equal gain / volume on the preamps. then flip the phase on the mic that you are testing for phase. choose the option that gives you the most low end. whichever position sounds like it has the most low end is the position that is in phase. that is one sure fire way to hear it.

if you want to do a phase test for the actual chain (mic excluded) -- record a synth or sine wave, etc through a preamp and/or cabling you know to be 2 pin hot... then pass it through the chain that might be out of phase... zoom in on the waves, match them up, and see if they changed 180 degrees or not. if the valleys turn into peaks, something in the chain is out of phase.

dfuruta
re-cappin' neve
Posts: 697
Joined: Fri May 08, 2009 11:01 am

Re: tests

Post by dfuruta » Tue May 15, 2012 9:55 pm

Seditionary wrote:whichever position has most low end is the one that is in phase. if the sound is thin... then you can assume the polarity is not correct.
If you only have one mic picking up a particular source, flipping the polarity will have no effect on the amount of low end. How could it? There's nothing else to cancel with...

Seditionary
audio school graduate
Posts: 19
Joined: Tue Oct 27, 2009 6:12 am
Location: Los Angeles

Post by Seditionary » Wed May 16, 2012 4:30 am

Yeah, if you can't tell a difference, you should use a second reference mic to check phase. Put the capsules together... Described in my post above :)

User avatar
vvv
zen recordist
Posts: 10173
Joined: Tue May 13, 2003 8:08 am
Location: Chi
Contact:

Post by vvv » Wed May 16, 2012 4:43 am

Hmm, just corresponding with a English friend, ...


... I wonder if the OP meant "that flipping phase" ...


:twisted:
bandcamp;
blog.
I mix with olive juice.

User avatar
farview
tinnitus
Posts: 1204
Joined: Tue Aug 31, 2004 1:42 pm
Location: St. Charles (chicago) IL
Contact:

Post by farview » Wed May 16, 2012 7:35 am

Seditionary wrote:Yeah, if you can't tell a difference, you should use a second reference mic to check phase. Put the capsules together... Described in my post above :)
Yes, but why? If you can't tell the difference (which you shouldn't be able to) there is no point in checking it. Of course, with your method, it assumes that the second mic is is phase to begin with...

ashcat_lt
tinnitus
Posts: 1094
Joined: Thu Jan 04, 2007 1:54 pm
Location: Duluth, MN
Contact:

Post by ashcat_lt » Wed May 16, 2012 8:13 am

Yeah. If there's one vocal on through one mic on one track then flipping polarity should make no difference as far as overall frequency response. That's a period.

There might be a very subtle difference in the "impact" of the transients. I find it very questionable, but people say they can hear it.

But there's the question: is it really just one vocal on one track? Is there ant parallel processing happening in this mix? If the polarity flip happens post send (not sure why this would be, but I've never tested my DAW for that. Have you?) it could make a noticeable difference when the effect is mixed with the original. Also, if there is parallel processing with un- or poorly compensated latency then polarity flipping could also make a major difference.

User avatar
vxboogie
pushin' record
Posts: 284
Joined: Wed Aug 06, 2008 11:25 am
Location: OH

Re: tests

Post by vxboogie » Wed May 16, 2012 8:17 am

dfuruta wrote:
Seditionary wrote:whichever position has most low end is the one that is in phase. if the sound is thin... then you can assume the polarity is not correct.
If you only have one mic picking up a particular source, flipping the polarity will have no effect on the amount of low end. How could it? There's nothing else to cancel with...
+1000 - phase and polarity are relative and don't matter if you are talking about a single source.

Also, people are saying that it matters if the wave pushes or pulls the mic or speaker, but they aren't taking into account that the sounds are AC not DC. So they are pushing and pulling during every cycle(frequency). So my stand is that the attack or transients aren't affected they way people are saying either since during the attack(think amplitude not frequency) phase there are multiple positive/negative cycles(frequency) that are happening.

To me this is the same as does a red or blue 57 sound better...

Totally different story when you add in other mics, though.
Mark - Listen, turn knob, repeat as necessary...

dfuruta
re-cappin' neve
Posts: 697
Joined: Fri May 08, 2009 11:01 am

Post by dfuruta » Wed May 16, 2012 8:44 am

It also might be worth pointing out that, on some instruments, the "correct" polarity isn't obvious. For example, are you sitting in front of or behind the drum kit? Should the kick (& snare & toms) start with a rarefaction if you pan things drummer perspective?

I can't hear the difference in isolation anyway...

Chris_Meck
pushin' record
Posts: 209
Joined: Sat Feb 28, 2004 2:40 pm
Location: Kansas City
Contact:

Post by Chris_Meck » Fri Jun 01, 2012 12:29 pm

I use the polarity flip trick a lot in live sound, usually when stage volume is proving difficult for me to get the vocals out front. I don't exactly get HOW it works, but a lot of the time, flipping the polarity on the lead vocal will make it pop right out front without turning it UP.

It doesn't ALWAYS work, but it USUALLY works. it can also work with a particular instrument that's a lower than ideal signal.

I know it's a different situation than the OP is bringing up, but maybe it's related.
greetings from Flyover Country...

User avatar
Nick Sevilla
on a wing and a prayer
Posts: 5575
Joined: Mon Mar 03, 2008 1:34 pm
Location: Lake Arrowhead California USA
Contact:

Post by Nick Sevilla » Fri Jun 01, 2012 11:50 pm

One thing on "absolute" phase, not to be confused with polarity.

Usually, when you strike an instrument, or excite it into making the sound, it wil normally have some part of it which moves, and its' movement causes the air on its' surface to move away from the instrument, which is what we call the "sound" transmission.

Typically, whatever this initial mechanical movement IS, we want the speakers at the final end of our signal chain to move in the SAME direction.

So, imagine a drum skin, which when struck by a beater, moves outwards away from the beater, away from the foot which struck it. This is looking at the drum skin from in front of the kit.

We would call this initial movement the initial excitation phase of that drum skin.

If we are to use this drum skin in a traditional musical mix, we would imagine this drum skin to be in the imaginary "front" of the listener, and we would make sure that the speaker reproducing this sound will move in the SAME direction when the electrical signal reaches it, matching that absolute phase of the insturment.

In contrast, if an instruments' initial mechanical movement is AWAY from the intended listener, then we would expect the speaker to move again, in the same direction of the mechanical movement which created the sound.

Now for the fun, very simplified explanation of Polarity in an electrical wire:

Polarity has to do with the ELECTRICAL properties of a balanced electrical signal travelling through a balanced wire, such as we have in an XLR 3 pin cable. In a system, we typically decide to assign pin 1 to ground, pin 2 to "hot" or positive voltage swing, and the pin 3 to the negative voltage swings. These two voltages, although are complete waves within the wire, are only used to allow a longer cable run, by splitting the positive and negative portions of the original waveform, into two polar opposite voltages, which travel at the same time down the wire, and reject RF inteference which may exist in the air around the wire. The two voltages are usually half the amplitude of the original signal.
Once these two polar opposite signals reach the mic preamp, they are again combined and sent down, combined back into the original signal and amplitude, then amplified within the preamp as much as necessary for further processing. Typically the reconstructed signal is turned into a DC signal flow, with only a voltage wire and ground, for most applications, which allows for very small components to process it to be used. Like our modern transistors.

As you can read from the above, Polarity has nothing to do at all with the sound itself, it is only concerned about what voltages are going down a wire.

I hope this clears some misconceptions on what is going on.

Cheers
Howling at the neighbors. Hoping they have more mic cables.

dfuruta
re-cappin' neve
Posts: 697
Joined: Fri May 08, 2009 11:01 am

Post by dfuruta » Sat Jun 02, 2012 10:48 am

Nick Sevilla wrote: If we are to use this drum skin in a traditional musical mix, we would imagine this drum skin to be in the imaginary "front" of the listener, and we would make sure that the speaker reproducing this sound will move in the SAME direction when the electrical signal reaches it, matching that absolute phase of the insturment.
Like I asked above, would you have the polarity the other way if you're panning from drummer perspective? Neither is really correct in real life; it depends where one is standing.
Polarity has to do with the ELECTRICAL properties of a balanced electrical signal travelling through a balanced wire, such as we have in an XLR 3 pin cable. In a system, we typically decide to assign pin 1 to ground, pin 2 to "hot" or positive voltage swing, and the pin 3 to the negative voltage swings. These two voltages, although are complete waves within the wire, are only used to allow a longer cable run, by splitting the positive and negative portions of the original waveform, into two polar opposite voltages, which travel at the same time down the wire, and reject RF inteference which may exist in the air around the wire. The two voltages are usually half the amplitude of the original signal.
Once these two polar opposite signals reach the mic preamp, they are again combined and sent down, combined back into the original signal and amplitude, then amplified within the preamp as much as necessary for further processing. Typically the reconstructed signal is turned into a DC signal flow, with only a voltage wire and ground, for most applications, which allows for very small components to process it to be used. Like our modern transistors.
This is incorrect. Both pin 2 and pin 3 carry the full signal (both the positive and negative swings); the difference is that pin 3 has the full signal with inverted polarity. Why would we want to do this?

As the signal travels down the cable, it's going to pick up some noise. The idea is, both conductors (in close proximity) will pick up the same noise. Then, when you get to the destination, we invert the signal on pin 3* (so, now the polarity of the musical signal is the same as on pin 2), and then add the two signals. Since we flipped the polarity of the noise that got picked up by pin 3 (same as the noise that got picked up on pin 2), the noise on pin 2 and 3 cancel each other out, while the musical signal adds to itself (i.e. gets ~6dB stronger). So, we get rid of the noise we don't want, and the signal we want is reinforced. This is the magic of balanced signals...

DC doesn't enter into this, also. If the signal got turned into DC, there wouldn't be any musical information left.


* this isn't quite right; look up "differential amplifier" on wikipedia if you want the real story

Post Reply

Who is online

Users browsing this forum: No registered users and 78 guests