Absolute phase, why does it matter?

Sorry, now you know, how my fingers feel after trying to keep up, however drinking form a fire hose is refreshing at the least T.

 

FWIW all of WMs stuff is irrelevant to 'absolute phase'....

I'm very surprised that people here aren't hearing differences, it's standard audiophoolery, much more accepted than funny mains cords and other ZG staples.

I want to have absolute phase right for the same reason I want left to be left and right to be right. And getting the latter wrong doesn't affect the sound or image quality at all...

Paul

 

Paul, I can't remember saying any of my above posts had anything other than do to with phase variant shifts & attemping to redress the balance of the phase inversion. (if a tall possible)

 

Getting L/R right may not affect sound or image quality but I wan't to hear the violins to my right and the cellos to my left and it's certainly noticable if you get them swapped around!

The thread is asking "why" is absolute phase standard audiophoolery? Whether it's accepted or not I don't really care. I've heard significant differences with mains cords "and other ZG staples" but not an iota of difference with reversing phase.

Michael.

 

I agree Michael, I have never heard a difference with absolute phase.

When you consider the phase inversion that goes on in the studio and mastering suite it's hardly surprising. With regards to WM's posts, the amount of phase shifts and distortion that exists in just about all pre power amps is simply a fact of life. Unless of course, everything is done in the digital domain

 

Agree with Mike about L/R - but not entirely sure about absolute phase.

I think absolute phase may make a difference to low frequency stuff - where the impulse we're listening to is no more than a few cycles (at most) of the fundamental frequencies. Think about the time-domain signal of e.g. a drum - it's clearly asymmetric, heavily damped, basically low frequency, so it's not inconceivable that it's different if polarity inverted.

At one demo, the dealer was clearly concerned to get the absolute phase between the pre and power right, although I didn't notice much difference. He didn't seem to think it was the same thing as inverting the phase from the CDP though - a contention that makes no sense to me whatever.

All of that cut'n'paste by WM is completely irrelevant to the current discussion. The following has possibly more (although still vanishingly small) relevance - the difference is that I (and I guess most people here) realise it

The only thing that may have any relevance is the response of an electrolytic cap (being polarised). However any circuit that does show such a dependence is clearly broken - as it can't accurately reproduce AC signals. Actually, I guess it's possible that devices with strong even order distortion (e.g. valve amps ) may show an absolute phase dependency.

(And for the digital obsessives, phase shifts are inherent in most causal filters whether implemented digitally or analog. Digital does however allow implementation of FIR filters that may be less phase-dispersive than typical analog)

 

Interesting Graham.

I guess with a name like Max Planck one should expect him to beleive in the relevence of all this.

 

Wow! This thread has developed impressively since I last checked. Special bonus points to WM for draggin quantum mechanics into the frey

I'll just add my tuppence worth to some of the simpler points...

I agree entirely with michaelab's response to this. In addition though it occurred to me that even if you did only hear one half cycle, or particularly noticed the first half cycle, whether this would be a compression or rarefaction ( you being pushed or pulled ) will depend on where the microphone was sited relative to the top of the drum won't it? I may be wrong, but aren't mics usually held above the surface of the drum, which would I suspect mean the first half cycle of the recorded signal will be negative (the drum skin being pushed away from the mic)?

On the issue of "phase inversion" it strikes me that what I think buttons labelled as such actually do is invert the polarity of the signal; is that right? If indeed signals are not symmetric, which I guess they can't be since in general they are decaying resonances, then I don't believe there is a phase change which would cause the same effect.

What is important is that the relative phases of the different frequencies is preserved as the signal is passed through the recording and sound reproduction chain. Phase shifts resulting in merely introducing a time delay, the same at all frequencies, have no effect on the sound.

Assuming you don't normally sit with your back to your speakers I suspect you've got your left and right mixed up there

 

DOH! :shame: You're correct of course Violins on the left and cellos on the right.

Michael.

 

erm.......have you lot been attending college over the summer .......this dumb cookie stayed at home.....


You alright Mike - you seem to be blushing a lot lately - must be that jazz lark you've been getting into recently.....its messing with your head

 

If you don't listen to orchestral recordings where this is the convention then how would you know?

To be absolutely clear, because there does seem to be some confusion still, 'absolute phase' refers to which way up the signal is, you change it by inverting the signal, either electronically or by swapping the plus and minus speaker connections around (either at the amp or speaker, not both....). The effect (if any) cannot be replicated or compensated for by any other phase processing.

Because I have so many power amps and speaker cables I'm going to rip a selection of likely tracks and produce a CD with normal and inverted versions, for research purposes...

I had this demo'd to me long long ago so I know that the effect can be relevant, but there's no reason why it would always be.

Paul

 

I also listen to TV programs, and very occasionally DVD's through my HiFi, when it can be obvious of course . Also I did actually once with recent speaker cable demos get left and right mixed somehow, and I spotted it on a non-classical recording just because things weren't the usual way round. It was most odd actually, in that my brain kept expected a sound to come from one side and it came from the other!

I thought that might very well be the case... Out of interest does anyone know why in HiFi circles this is referred to as phase?

 

Sorry but are you guys missing the point here. Inverting phase will not swap channels over, you are simply swapping positive for negative on each channel. try it on your speakers. See Left is still left, right is right.

Absolute phase is held by many audiophools to be essential to maintaining the signals integrity, but it is primarily software dependent I believe. the argument goes along the lines that, provided you maintain the correct phase throughout your amplification and speaker interface, inverting the phase at the source will have an affect.

If a recording was mastered with it's phase inverted, then switching it a further 180 degrees at the source output will ensure that the recording is not phase inverted at the listening position. Hence the phase invert facilities on many high end CD players and dacs, to allow switching. In theory one setting will sound better than the other depending on the disc being played.

Can't say I have ever heard a difference, but then maybe I'm listing to the music by mistake.

 

Nope, not confused, we're just following on from Paul Ranson's earlier comment. The issue of "phase inversion" is nothing to do with L/R swapping as you say.

BTW, now I'm sure what we're talking about, I can see there could be something in this phase inversion, since as I said above I think audio signals can't be completely symmetric. Whether the difference is large enough to be audible though is a completely different matter, and to be honest I can't be bothered to go and check. Whatever, I'll stick to assuming that the recording is made with playback in mind, such that using no inversion gives the "real" sound.

 

I couldn't hear any difference when I swapped the +/- leads over on my speakers. Since I couldn't hear any difference I could have just left them like that as, for the DAC64 it would have been "correct". However, since I also use my 2 stereo speakers as the two front speakers in my AV system it would have been a bad idea because then the fronts would have been out of phase with the centre and that certainly would have been noticable

Michael.

 

I think it's confusing to use degrees here. There is no phase shift at all.

Paul

 

Try 90

 

No Tone - Paul is right here I'm afraid. Talking about angles/degrees is implying a phase shift which isn't what's happening. The waveform is not being shunted forwards or backwards 90 (or whatever) degrees, it's being turned upside down.

In the case of a sine-wave, a 180deg. phase shift is the same as inverting it but that's not true for most "normal" sound waveforms.

Michael.

 

I give up really never have fucking guessed, one day you fuckers might just get a Joke, scary thing on a hifi forum I know. :cry: