Amplifier/Pre-amplifier phase and phase inversion. How do you check it?

[walpurgis]

How do you detect or check for amplifier phase and phase inversion? I don't know how you'd tell, so I'm asking. Presumably an oscilloscope comparing a reference signal would tell you, but most of us don't have that kind of gear.

[Jimbo]

Hi Geoff,

Sorry I can't give you a technical answer as I am not a techie but you can hear the difference. Not as obvious as speakers out of phase but it is audible.

I believe there are only a few amps on the market that invert phase so for most folk it will not be a concern.

[Firebottle]

Maybe you've identified a gap in the market -

The numero uno super duper hi-fi phase checker, yours Sir for only £399........

Actually it wouldn't take a lot of gubbins to make one, two LEDs to show phase same or phase inverted.

:cool: Alan

[Barry]

How about knocking up a circuit design for us Alan? We could put it in the DIY section.

The great majority of pre-amp/amplifier combinations are non-inverting, but there are some where the ouput of the pre-amp is deliberately inverted to compensate for the inversion of the power amp.

[Paul Hynes]

Hi Geoff,

This is a tricky subject, as you have no control over any signal processing used at the recording stage in many recordings. The signal of instruments and voices can pass through various processing stages to add effects and equalisation some of which can invert phase and/or add time delays. Ideally a positive pressure change on the microphone diaphragm will produce a positive electrical voltage and a negative pressure change would produce a negative electrical voltage and any additional signal processing after this would maintain this phase relationship.

Assuming a recording where the phase response has been maintained throughout the recording chain, phase inversion of the signal by the replay equipment should be noticeable. The overall transparency of the replay system will also be a factor to consider and a transparent system would make it easier to perceive any difference between non-inverted and inverted phase than a system that is relatively opaque.

To me, once again assuming a recording where the phase has not been inverted, a non-inverted signal sounds more right than one where the phase relationships have been inverted. There appears to be more presence and body to the performance. For instance, a kick drum waveform leading edge should provide a positive pressure change giving a more dynamic punch than if it were inverted to give a negative pressure change.

Without an oscilloscope to check the replay equipment signal phase you are left with listening comparisons. Give it a try and if you can perceive differences.

Regards
Paul

[Barry]

The audibility, or otherwise, of absolute phase was discussed in the first part of a post (http://theartofsound.net/forum/showthread.php?11839-Absolute-phase-and-integrated-cartridge-arms) I made some time ago. I couldn't hear any difference.

[AlexM]

I haven't tested exhaustively in any way, but I can say that the only time I have heard phase inversion have an audible effect is with a phase inversion test track featuring a male voice speaking, recorded with a single mike. For non purist multi track recordings, I'm not sure what correct phase even means - for which track? With what microphone placement and what mic preamp? and in comparison with what DI source?. Given that this is highly unlikely to be consistent there is no 'correct' absolute phase in a recording to preserve.

My preamp inverts phase so I reverse the polarity of the speaker leads to maintain whatever is coming from the source. In fact, I'm not sure if the phono stage is phase inverted or not (and I don't really care as it sounds great anyway)

Cheers,
Alex

[awkwardbydesign]

For instance, a kick drum waveform leading edge should provide a positive pressure change giving a more dynamic punch than if it were inverted to give a negative pressure change.

That's the only time I've ever been able to say with certainty that I've heard the difference. It was repeatable too, but that was with sealed box bass (Heybrook HB3). The abuse I got on another forum for mentioning it was incredible. Mind you, that forum is infamous for that kind of thing; cables anyone?

[337alant]

Ahhh (Heybrook HB3) Phatt Bass, loved them speakers

Alan

[walpurgis]

Hmm. Many replies, but no definitive answer. I sort of expected that would be the case.

I wonder if there's a little box of tricks on the market anywhere at a sane price that can show phase?

[Barry]

Why don't you try reversing both speaker leads? If you notice a change then you can hear the effect of polarity inversion. Choose which orientation you prefer and stick with it.

IMO you may notice a change with some LPs/CDs, but not all. Multi-miked recordings probably loose any polarity alignment they had, if they had any at all.

[walpurgis]

Yeah, been there (inadvertently sometimes). Having one speaker out of phase is interesting. I wonder how many home systems are cocked up by that? Thousands I reckon (especially ones owned by girlies. I'll probably get a whack round the ear from Bev now :D).

[Jimbo]

I think speakers out of phase is more noticeable than amps out of phase, in my experience. So no wonder many folk may not be even aware that they may have not wired up their amps in correct phase.

That is if they invert phase in the first place.

[Barry]

Why do you think that should be James?

[Clive]

All I can add to this is that polarity matters in my system as I have active bass so given that most T-amps / Temple amps are inverting I have to allow for this when setting up my bass amps. Likewise anyone with line driven subs will need to account for this.

[hifi_dave]

Phase inverting in amps or pre-amps is subtle at best and depends on the recording and even tracks in a recording. Usually, it manifests itself as a slight difference but nothing major.

Reverse phase where one speaker is wired pos to neg is very audible and even uncomfortable to the ears when you are seated in the centre position. To me, it feels like my brain is being pulled around, it also reduces bass. Very noticeable.

[fiddlemaker]

I think speakers out of phase is more noticeable than amps out of phase, in my experience. So no wonder many folk may not be even aware that they may have not wired up their amps in correct phase.

That is if they invert phase in the first place.

Are you talking about both speakers being wired in reverse phase or just one? As Dave says the latter is very noticeable, the former not nearly so much (and to many of us mere mortals not at all).

[Jimbo]

Phase inverting in amps or pre-amps is subtle at best and depends on the recording and even tracks in a recording. Usually, it manifests itself as a slight difference but nothing major.

Reverse phase where one speaker is wired pos to neg is very audible and even uncomfortable to the ears when you are seated in the centre position. To me, it feels like my brain is being pulled around, it also reduces bass. Very noticeable.

Bang on Dave - hope this answers your question Barry.

[Jimbo]

Are you talking about both speakers being wired in reverse phase or just one? As Dave says the latter is very noticeable, the former not nearly so much (and to many of us mere mortals not at all).

Hi John, just the one speaker.

Regarding phase inversion with amps, as Dave says it is not always noticeable but can be more noticeable when playing CD rather than Vinyl.

[Barry]

Bang on Dave - hope this answers your question Barry.


I thought you were distinguishing between reversing both speaker leads and reversing the polarity of both input signal leads.

I never thought the discussion was about the speakers being out of phase.

[Jimbo]

Think I might have got some crossed wires somewhere:lol:

[Mr Kipling]

Yeah, been there (inadvertently sometimes). Having one speaker out of phase is interesting. I wonder how many home systems are cocked up by that? Thousands I reckon (especially ones owned by girlies. I'll probably get a whack round the ear from Bev now :D).

Strangely enough, Steve Hoffman on his forum advises out-of-phase speakers to set them up in the room. You position them to get the greatest diffused sound. I'd never heard of it myself and always used a mono signal which obviously shows up phase problems, to set my speakers up.

[Barry]

Yes - that is the better method: always easier to achieve a good, focused, centre image, than trying to set the system up for the most diffuse image (! :scratch:).

[Jimbo]

:scratch::scratch:

[Barry]

What I was trying to say is when you have wired the two speakers in phase and position the speakers so that on playing a mono record, the centre image can clearly be heard to be nicely focused. Using out of phase speakers, how can you tell when you have achieved the "greatest diffused image"?

It's a bit like saying one can easily see when an optical image is in focus, but when out of focus, how do you judge the 'maximum blurr'?

[Mr Kipling]

What I was trying to say is when you have wired the two speakers in phase and position the speakers so that on playing a mono record, the centre image can clearly be heard to be nicely focused. Using out of phase speakers, how can you tell when you have achieved the "greatest diffused image"?

It's a bit like saying one can easily see when an optical image is in focus, but when out of focus, how do you judge the 'maximum blurr'?

Yes, I know. That's what I was thinking. How can you tell?

[camtwister]

The original post enquired about tools for checking absolute polarity. There are a number of manufacturers of test equipment, aimed at the pro-audio market. Use of these is particularly convenient when operating large arrays of drivers in concert situations, in cabinets that are not exclusively used in the same system, either during a tour or festival season, for example.

Here is one such device:

http://www.laaudio.co.uk/product_pc90_1372.aspx (http://www.laaudio.co.uk/product_pc90_1372.aspx)


You may notice that I refer to polarity rather than phase. Consideration of sine-wave orientation in two dimensions is a useful method of determining the absolute polarity of simple domestic systems. However, as discussed by a number of other contributors, maintenance of absolute polarity at the recording stages of processing energy from musical instruments is not straightforward and is often not a significant consideration. I refer to my post in the thread linked to in post 6, by Barry.

“During the recording sessions that I've engineered and the gigs that I've mixed, management of signal polarity from mic to console is limited to 180 degrees, via either phase changing cables or inverting switches on the desk.
Management of phase between instrument and mic is additionally limited by the engineer's skill in mic placement. Not that a totally phase coherent response between the incident waveform and the mic diaphragm is necessarily desirable or entirely achievable. It may not be desirable in terms of frequency response and it may not be possible when close miking due to rippling effects on the diaphragm or ribbon.

Most non-orchestral percussion instruments are recorded by more than one transducer. For example, a bass drum in rock music is recorded by a minimum of two mics. One will be placed near the centre or rear of the drum and will be a large diaphragm type or boundary effect type for capture of bass frequencies. A second, different mic - capable of withstanding high SPL - will be placed near the batter skin, and will capture the 'click'. The phase relationship of the output signal of these two mics is complex. It's quite possible to have two or three mics per instrument contributing to a multi-track mix, which is then manipulated via several further processes. Once all of the ingredients in the pudding have been kneaded and baked, to say that the net movement of drivers in a playback system is directly proportional to the original phase relationship of an individual recorded instrument and transducer is too simplistic.

Ambient real-time recording, still regularly employed in orchestral sessions, is similarly non-linear. Consider a classic coincident or spaced pair arrangement of microphones placed tens of metres from the performers. The phase differences between the two mics are exactly what the brain uses to decipher stereo image and soundstage. However, the incident pressure wave on each mic is an interference of direct and reflected sound, the phase relationship of which is rather knotty. So, even before dynamic processing such as compression, fx processing such as digital reverb and all the mastering hoo-ha, phase coherence isn't coherent.”

However, having made such a statement, I always rig my hi-fi system to reproduce absolute polarity, because I am able to make a subjective decision for this preference when listening to percussive or plosive sounds from a single instrument, if there is significant low frequency content.

In the related thread to this (http://tinyurl.com/kob8l3r) an example of unexpected results due to what may be the phase behaviour of electrical signals in a circuit is given in post 52.

It is best to think of an electrical signal in the models that concern us, not as a two dimensional wave, but as current density. This is simple enough to conceptualise as a three dimensional quantity. However, it is a vector and specifically a 4-vector when charge is computed, in Minkowski space.
Consider also, that charged particles in a conductor do not have zero kinetic energy even when net potential across the conductor is at equilibrium. A dynamic model of the Fermi energy of the electrons, incorporating molecular interaction and skin effects for example, requires higher dimensional analysis. Hence, interpreting phase relationships of alternating electrical signals, even at audio frequencies, is not straightforward and simplistic macroscopic analogies will fail. This reinforces the requirement, at the evaluation stage of a good design process, for ‘listening’ to be included in the feedback loop in addition to calculating and measuring.

[fiddlemaker]

A dynamic model of the Fermi energy of the electrons, incorporating molecular interaction and skin effects for example, requires higher dimensional analysis. .


Please can you explain what you mean by this?

[anthonyTD]

hi All,
This subject has been discussed a few times here on AOS over the years, and the same answers have come up consistently.
We have no control over absolute phase as a consumer, as others have stated, a lot will be dependent on the engineer, the recording equipment being used, and finaly, the replay equipment we use, even the amplification musicians use sometimes have phase inversion switches built in!!! what we do have control over' and which is very important is making absolutely sure that the equipment we use to replay the recordings on have exact phase in relation to each channel, be it speakers, and outputs from source, to power amp,and preamplification etc. As long as you have established this, then by all means experiment with the phase changes you have control of, ie; both signals of each channel in phase or both with their phase inverted, [whether you do this at source, or speakers] see which way round suits your system, and your listening enviroment as a whole.:)

[anthonyTD]

Excelent post!
The original post enquired about tools for checking absolute polarity. There are a number of manufacturers of test equipment, aimed at the pro-audio market. Use of these is particularly convenient when operating large arrays of drivers in concert situations, in cabinets that are not exclusively used in the same system, either during a tour or festival season, for example.

Here is one such device:

http://www.laaudio.co.uk/product_pc90_1372.aspx (http://www.laaudio.co.uk/product_pc90_1372.aspx)


You may notice that I refer to polarity rather than phase. The modelling of sine-wave polarity in two dimensions is a useful method of determining the absolute polarity of simple domestic systems. However, as discussed by a number of other contributors, maintenance of absolute polarity at the recording stages of processing energy from musical instruments is not straightforward and is often not a significant consideration. I refer to my post in the thread linked to in post 6, by Barry.

“During the recording sessions that I've engineered and the gigs that I've mixed, management of signal polarity from mic to console is limited to 180 degrees, via either phase changing cables or inverting switches on the desk.
Management of phase between instrument and mic is additionally limited by the engineer's skill in mic placement. Not that a totally phase coherent response between the incident waveform and the mic diaphragm is necessarily desirable or entirely achievable. It may not be desirable in terms of frequency response and it may not be possible when close miking due to rippling effects on the diaphragm or ribbon.

Most non-orchestral percussion instruments are recorded by more than one transducer. For example, a bass drum in rock music is recorded by a minimum of two mics. One will be placed near the centre or rear of the drum and will be a large diaphragm type or boundary effect type for capture of bass frequencies. A second, different mic - capable of withstanding high SPL - will be placed near the batter skin, and will capture the 'click'. The phase relationship of the output signal of these two mics is complex. It's quite possible to have two or three mics per instrument contributing to a multi-track mix, which is then manipulated via several further processes. Once all of the ingredients in the pudding have been kneaded and baked, to say that the net movement of drivers in a playback system is directly proportional to the original phase relationship of an individual recorded instrument and transducer is too simplistic.

Ambient real-time recording, still regularly employed in orchestral sessions, is similarly non-linear. Consider a classic coincident or spaced pair arrangement of microphones placed tens of metres from the performers. The phase differences between the two mics are exactly what the brain uses to decipher stereo image and soundstage. However, the incident pressure wave on each mic is an interference of direct and reflected sound, the phase relationship of which is rather knotty. So, even before dynamic processing such as compression, fx processing such as digital reverb and all the mastering hoo-ha, phase coherence isn't coherent.”

However, having made such a statement, I always rig my hi-fi system to reproduce absolute polarity, because I am able to make a subjective decision for this preference when listening to percussive or plosive sounds from a single instrument, if there is significant low frequency content.

In the related thread to this (http://tinyurl.com/kob8l3r) an example of unexpected results due to what may be the phase behaviour of electrical signals in a circuit is given in post 52.

It is best to think of an electrical signal in the models that concern us, not as a two dimensional wave, but as current density. This is simple enough to conceptualise as a three dimensional quantity. However, it is a vector and specifically a 4-vector when charge is computed, in Minkowski space.
Consider also, that charged particles in a conductor do not have zero kinetic energy even when net potential across the conductor is at equilibrium. A dynamic model of the Fermi energy of the electrons, incorporating molecular interaction and skin effects for example, requires higher dimensional analysis. Hence, interpreting phase relationships of alternating electrical signals, even at audio frequencies, is not straightforward and simplistic macroscopic analogies will fail. This reinforces the requirement, at the evaluation stage of a good design process, for ‘listening’ to be included in the feedback loop in addition to calculating and measuring.

[camtwister]

"A dynamic model of the Fermi energy of the electrons, incorporating molecular interaction and skin effects for example, requires higher dimensional analysis."
Please can you explain what you mean by this?


Yes John. I'm suggesting that multivariable calculus in four dimensions (or greater) is required to model alternating current densities.
The reference to Fermi energy is an acknowledgement of the quantum behaviours of fermions during conductance and if you were minded to explore alternating current densities at this level, you would employ wave functions in three dimensions or higher. It's of interest that probability flux is described by functions that are very similar (at my level of understanding) to those that describe electromagnetic behaviour.
The modelling is further complicated by interaction at a molecular level that is intrinsic to the crystal structures in the conductor.

I'm procrastinating before undertaking the washing up, so don't give my ramblings any significance. I'm trying to point out that the phase relationships of alternating currents in a system are a bit more complex than two dimensional graphs of sine waves. Hence, at the risk of kicking about with semantics, a differentiation in the use of "polarity" and "phase" may be helpful.