-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
KSTR
Major Contributor
Timbre change from polarity flip is readily audible to almost everybody with almost all music that contains asymmetric low frequency waveforms, like electric bass guitar. With speakers, with headphones, at any level, so distortion of the transducers is NOT a false clue.
Actually, I use polarity flip as my go-to warm-up test when doing ABX blind tests... on a bad day I may not get my ususal row of several 10/10 and then I know it makes no sense in proceeding.
EDIT: Snippet of one of my test tracks attached.
It is also well established for decades why we hear asymmetric waveform differently when polarity-flipped.
But a polarity flip also can have impact on soundstage impression (notably phantom source apparant size) and other aspects of reproduction.
How to check recording for "correct polarity"? With classical music or other stuff that was recorded pretty much directly from just a few microphones you may use segments with instruments soloing or standing out that have asymmetric waveforms, like trombone/trumpet... if you have a know good polarity reference track...
Bottom line: there is zero argument about this issue. The impact is small but undebatable. For most music recordings there is no correct polarity, so chose what sounds best to you, or simple don't be bothered. For classical there are lists where the polarity of many recordings is stated in case it still bothers you.
Actually, I use polarity flip as my go-to warm-up test when doing ABX blind tests... on a bad day I may not get my ususal row of several 10/10 and then I know it makes no sense in proceeding.
EDIT: Snippet of one of my test tracks attached.
It is also well established for decades why we hear asymmetric waveform differently when polarity-flipped.
But a polarity flip also can have impact on soundstage impression (notably phantom source apparant size) and other aspects of reproduction.
How to check recording for "correct polarity"? With classical music or other stuff that was recorded pretty much directly from just a few microphones you may use segments with instruments soloing or standing out that have asymmetric waveforms, like trombone/trumpet... if you have a know good polarity reference track...
Bottom line: there is zero argument about this issue. The impact is small but undebatable. For most music recordings there is no correct polarity, so chose what sounds best to you, or simple don't be bothered. For classical there are lists where the polarity of many recordings is stated in case it still bothers you.
Last edited:
KSTR
Major Contributor
For multiway speakers with normal crossovers there is no overall correct polarity as the speaker rotates the phase all the time. You can only have restricted frequency regions where the polarity is right. Some manufacturers use the convention that the lowest way -- the woofer -- should have positive polarity (for less cumbersome integration with subs or other speakers in HT setups).
A step signal (signal going from 0 to 1, and back to 0 after about a second) is good test signal, if the cone goes outward the polarity is positive for the woofer.
A more general test signal that works at any (or a specific) frequency range is an asymmetric waveform, the simplest being a sine plus its second harmonic at a certain phase offset, and then check that with a microphone on an oscilloscope. How do you know the polarity of the mic is correct? Ballon (paper bag) explosion ;-).
There also are hardware devices the PA guys are using. Nowadays we even have phone apps for polarity checking but I haven't tested any of these.
solderdude
Grand Contributor
Blue is left, red is right channel.
Look at the scale on the left it is around -180 degrees.
The squarewave scope shots below are perhaps clearer.
Below the square waves from Edition XX (which are in phase)
What are normal crossovers?
There were a bunch of speakers designed to have low order crossovers where the phase was not flipped along the stack of drivers.
In DSP land and active crossovers, it gets a lot easier.
Thanks. I am aware of the x-over part but was thinking that the 80 Hz-3 kHz region, midwoofer, should be in correct polarity. A step signal should be possible and look at the woofer movement. Will try someday.
Hi Guys
I’m not sure if this will help but it is to do with ‘phase’
forums.stevehoffman.tv
Hope it helps.
Regards
Pegwill
I’m not sure if this will help but it is to do with ‘phase’
SH Spotlight - Play the Kinks "Waterloo Sunset" in stereo, check and fix or improve your stereo playback with it!
We've talked about this before but try it. Play the stereo WATERLOO SUNSET and see how far you can get the background singing to go. The recording...
Hope it helps.
Regards
Pegwill
KSTR
Major Contributor
Speakers with linear phase crossovers are few and far between, not more than 1% or so. "Normal" refers to what applies to vast majority (the definition of normal).
restorer-john
Grand Contributor
As in the activation of the switch? Sure, there will be a minor 'click' when a D/A converter inverts the polarity.
restorer-john
Grand Contributor
Just throw a asymmetic clipped sine through your system and look at ('scope) the signal at the loudspeakers. Or an impulse.
If your system preserves absolute polarity, it will show. (doesn't mean that in can be inverted and inverted again all the way through).
I swear I've heard absolute polarity issues in the past, but with decent gear (speakers/headphones), and music, not test signals, it's really a non-issue for me.
If your system preserves absolute polarity, it will show. (doesn't mean that in can be inverted and inverted again all the way through).
I swear I've heard absolute polarity issues in the past, but with decent gear (speakers/headphones), and music, not test signals, it's really a non-issue for me.
Are you saying you can listen to some music and say "oh this sounds like the phases is inverted"?
I can' t. but maybe you ears and especially your memory how it supposed to sound is better.
Also i can think this effect has to do with Asymmetrical non linear distortion.
In your speaker and also in your ear. so some ears have more Asymmetrical distortion than others.
in particular "Stiffness of suspension" eardrum is also not linear and symetrical.
KSTR
Major Contributor
Testfile of sensitivity to absolute phase and phase rotation.
If you hear a steady-state tone you're phase-deaf. If you hear a cyclic timbre change pattern you're not.
If you hear a steady-state tone you're phase-deaf. If you hear a cyclic timbre change pattern you're not.
Thanks. My neighbours hate it.
The louder i play it easier i can hear it.
tuga
Major Contributor
With music that I am familiar with I can sometimes hear when polarity needs flipping.
Some tracks sound like they’re a bit congested and in my head (with speakers) and tonally somewhat off (mostly the bass).
tuga
Major Contributor
I had played with polarity inversion many years ago when I owned an Audio Alchemy 3-box CD player.
Recently I gave it another go with HQPlayer and was surprised by how much difference it makes with some tracks
Recently I gave it another go with HQPlayer and was surprised by how much difference it makes with some tracks
OP
Deleted member 23982
Guest
- Thread Starter
- #57
Oh i have such a test track too for 3d spacing, its " Burn the Witch by Radiohead " really a good one to test for it with those "echo-effects", and yes phase had an impact to it too, tho for me its still a "general impact" that affects everything, more or less (with my speakers atleast)Hi Guys
I’m not sure if this will help but it is to do with ‘phase’
Hope it helps.SH Spotlight - Play the Kinks "Waterloo Sunset" in stereo, check and fix or improve your stereo playback with it!
We've talked about this before but try it. Play the stereo WATERLOO SUNSET and see how far you can get the background singing to go. The recording...
i will give waterloo a go later too
Edit: if i would guess i would say the main vocals of waterloo sunset are with inversed polarity, is that right? (to create (or extend) this kinda "all over the place" effect)
Last edited by a moderator:
Nope. I was referring to the signal itself.
Apesbrain
Addicted to Fun and Learning
@Pegwill , @Ghoostknight , I think you may be confusing "absolute polarity" and "phase". Easy mistake, but they are not the same thing.
The definitive source on this subject is The Wood Effect by the late Clark Johnsen.
A simple AA battery across rour speaker will cause motion of the woofer as it is not blocked by capacitance in most cases. Forward means the battery is correct, and reverse means the battery is in reverse. This gets the bass in phase with non-inverting gear.
An easy Impulse response measurement, with tweeter disconnected, will tell if the system/gear is inverting or non-inverting, once you've done the battery test.
Depending on the slopes and summation of the speaker xover, you cannot rely on the tweeter to be normal or reversed, as the connection is preferably whichever keeps the speaker in phase through the xover.
If you have a full-range single driver speaker, the impulse will tell you.
Even supposed 1st order inherent designs may have the tweeter reversed to get good summations due to the inherent tweeter rolloff being second order and 3rd order with highpass in place.
An easy Impulse response measurement, with tweeter disconnected, will tell if the system/gear is inverting or non-inverting, once you've done the battery test.
Depending on the slopes and summation of the speaker xover, you cannot rely on the tweeter to be normal or reversed, as the connection is preferably whichever keeps the speaker in phase through the xover.
If you have a full-range single driver speaker, the impulse will tell you.
Even supposed 1st order inherent designs may have the tweeter reversed to get good summations due to the inherent tweeter rolloff being second order and 3rd order with highpass in place.
Similar threads
