James_I
(The truth is out there but not necessarily here)
41
The article says we can assume that absolute phase is inaudible. I don’t think that is true, although I have stood at my preamp playing with the “180 degrees” knob and sometimes it does sound different, but I have never been able to say one is better than the other.
For example, many owners of Magnepan speakers find that the best sound comes from turning them around so that the back faces the listener, because the speaker “grill” (it’s really a giant sock) is thinner on the back. This would be equivalent to reverse absolute phase. Perhaps these listeners also reverse phase.
I don’t hear a difference reversing absolute phase. Most in this thread seem to be referring to it correctly, but I will point out that there is often confusion between absolute phase and relative phase. Relative phase is when left and right are out of phase with each other. Absolute phase is when L/R are in phase with each other but out of phase with the original recording.
When relative phase is reversed, one speaker has + to - / - to + while the other has + to + and - to -. When absolute phase is reversed both speakers have - to + and + to -.
Out of relative phase is dramatically audible. Out of absolute phase is either barely audible or inaudible, depending on who you talk to.
The “grill” is actually the perforated steel sheet that holds the magnets. Turning it around exposes the naked membrane side to the listener. I believe this is how Magnepan makes their speakers today.
James_I
(The truth is out there but not necessarily here)
46
I recommend trying it. I do notice differences on my system. But if you don’t then you’ve got one less thing to irritate your audionervosa! While it is not black and white, I do see that in the couple of hundred albums that I’ve already tagged, one way or the other, that there’s more favorites in the correct column than the wrong one. By fav’s, I don’t mean musically, but ones that I go to for their sound. I find that correcting the wrong ones satisfies this anomaly.
One way to think about it is that the notes start of as a “suck” rather than a “blow”. Once the note has been struck you’re unlikely to hear any difference. But on the initiation of the wave-form, I think that’s a different matter.
Never tried it because not a single piece of hifi equipment I own or have ever considered buying has a phase reverse switch. This fact alone tells you how important it is in the scheme of things.
Mixing desk channels have them but that is to properly assemble the completely random phase input channels that they have to mix.
The only article I have read on the subject suggested that audibility of absolute phase was down to poor linearity (+ and - giving different displacements) in your loudspeakers. You would need a single drive unit to hear this most clearly, as a typical speaker with two drive units or more will of course have all sorts of frequency dependent phase issues due to the crossover.
No, this only tells that you where not aware of this and that most manufacturers see it as pain in the b… and therefore ignore it. Saves money… In the 70s and 80s Technics made many ‘Linear-phase’ speakers (SB-7000A is one), but none of the electronics had phase reversal even though Technics as a company had lots of knowledge about this.
There are a few speaker makers that take this seriously. In the past only a stepped baffle where acoustic centres of drive units could be aligned would be time aligned. The boxes where ugly so no one wanted them (apart from Vandesteens and the like that hid the construction behind grilles and stockings). Your single driver example is of course good. Another example is the Quad ESL-63.
Today there are several new time-coincident speakers available. The Kii THREE, Dynaudio focus XD series and Kef LS50W to mention some. These are all active and use DSP to do cross over and correct timing. Take a look at Stereophile’s test of the Kii. Fig. 1 and Fig. 2 under measurements shows the difference with and without time alignment. In Fig. 2 the Kii behaves like a ‘normal’ speaker. Fig. 1 is with time alignment (and probably other DSP tricks) to make the whole speaker time-coincident. With this kind of speaker, absolute phase of a recording is quite easy to observe. Whether one can hear it or indeed find the change relevant depends a lot on what music one listens to and how.
What you should be asking yourselves, is : ‘What exactly is correct absolute phase in a recording ?’
There really is no answer to that question. Let me try to put that as simple as possible :
I think I understand in this thread, that people believe the following :
A beat on a drum ‘pressurizes’ the room. Therefore, this should be recorded as a positive impulse. Which should be replicated by your source/amp/speaker chain.
Well, it is not a complete room that is pressurized/compressed at once. There is a sound wave travelling through the room. Some parts are being compressed, yup.
But that means that the other parts MUST be de-compressed simultaneously (after all, this extra compressed air has to come from somewhere).
Is the person standing in the decompressed-air-first zone, hearing an incorrect version of the truth ?
Or in real live applications : whether your ears experience a negative or positive impact at first, depends on the distance to the source.
Is the person or microphone on the front row experiencing the ‘truth’, and the person/mic on the middle row a modification of that ?
My answer would be NO. From a technical perspective, this is a strange request. I do understand that flipping phase can give slight sonic differences, as loudspeakers might have different non-linear behaviour for in- and outgoing movement. But this has nothing to do with problems in the recording itself.
Extra food for thought : why would you need exactly 180 degree flipping, and not some other arbitrary value like 68,54 degree ? Are recordings really exactly ‘mirrored’, and if so, why would that be ?
(deliberately ignoring the time-aligning mentioned above, as it is not not related to the original question)
I’m an empiricist at heart. Having said that, the audiophool world is full of inexplicable audible differences. I’m too respectful to assume that just because I can’t perceive a difference that someone else hears, that one or other of us is wrong (let alone the fact that it’s unlikely that we have identical systems/rooms). I don’t know about you guys, but for me one of the things that makes this hobby fun is experimenting, albeit, without gold standard tests (i.e., double blind, which are not the be all and end all, anyhow). I’ve uploaded a list of over 200 albums that I have subjectively decided have “correct” or “wrong” phase, here: Albums that users switch Phase
I’s be keen to hear(!) if anyone hears the differences I do. That way the Roon developers can get a sense of whether adding this function is a worthwhile project, regardless of “theory”.
I agree with some of your insights. Re the pressurized room problem - perhaps this contributes to the reason that we tend to have a listening sweet spot?
As for the binary right, wrong, up until the mixing desk there is the possibility of one channel being correct and another wrong. In a hypothetical 2 channel recording, all other things being equal, flipping the absolute phase will leave one or the other channel out.
Where the idea of absolute phase does have merit, IMHO, is post mixing. There are numerous processes applied to a recording before it is stamped on a CD. should any of these “get it wrong” then flipping the AF gets one back to a closer rendition of what was mixed. Some argue that this is a wide spread (albeit minor) problem and my subjective listening bears this out. approx 60/40.
My reason for this requested feature is that the Yggy resolves the recording to such a degree that this “minor” issue is audible and correctable, hence the feature request. Have you tried this for yourself?
It is not just time alignment itself of course, but also the resulting linearity of phase that lets you hear the the difference. A three or four way speaker with all sorts of un-linearities will mask absolute phase reversals. The speaker can sound brilliant on some material and less so on other types of recordings. I’d suggest that it would be (much) easier to hear absolute phase with the Kii time alignment enabled than without it. In the Stereophile review of Spica TC-60 JA also shows a plot of the speaker’s excess phase which is quite illustrative of what time alignment does to phase.
Time alignment and excess phase matters, but I find it hard to believe that absolute phase makes the slightest difference. A simple point to illustrate: move your head back a few inch, and you will have switched the absolute phase 180 for one frequency, and other amounts for all other frequencies. Lets say the frequency that was changed 180 is 1 kHz, do you honestly believe that the 1 kHz frequency suddenly sounds different? And even if it did, what about all other frequencies?
Don’t confuse this with switching phase for one speaker, which leads to cancellation and matters a lot.
All I can say to this is: did you try it?
If you do not perceive a difference that is fine. The effect is quite subtle. Theories of what it should or should not do does not apply. Practical experience is necessary. Also room correction can be dismissed with theories, but it works, doesn’t it?
Absolute Phase switching is bordering on audiophilum nervosa, a.f.a.I.c.
There used to be a printed list of record companies and their ‘aboslute phase’. It was considered to be the Holy Grail of Audiophilism, and it was almost ceremoniously handed down from one audiophile to another. I can still remember the day I received my copy. I was so happy…
Sorry, couln’t resist to make a little fun of it, but that list does exist and it still circulates. Problem with it is that it is completely outdated. It was made back in the day when record companies still had their own recording studios and/or mastering facilities, so it was easier to ‘predict’ that all CBS recordings would be AP-0 and all Deutsche Grammophon AP-180, for instance. But then the recording studios and mastering fscilities were discarded by the record companies because they were very expensive to maintain, and from that moment on, recordings from any other studio were used, and shipped off to whatever mastering facility was available. Predicting the AP became a game of chance. Then artists started ro record at home and in local studio’s, do some vocal tracks in another studio that has a better mike, and din’t bother to check the AP of separate tracks they mixed to a complete song on two tracks.
You see the problem? There may be tracks with AP-0, AP-180 and AP-0/180 on one album nowadays. Good luck sorting that out
It is true that tracks that are played ‘out of absolute phase’ can sound a teensy-weensy little bit different, but those differences are really, REALLY minimal at best and most of the time it is impossible to hear any aignificant difference at all.
So perhaps you should ask yourself if you need to bother at all. My digital collection contains about 10.000 albums, and I never use the AP switch in my NAD M51 DAC. I’d rather listen to music, instead of listening to sound, but that is a personal choice.
I’ve noticed that this can be a track-to-track problem within an album. Rare on classical albums, more common on pop/rock/jazz. The list you refer to is probably old enough to be vinyl albums? Early CDs transferred from analog tape is more prone to AP errors than newer digital recordings.
I have the joy of an ‘Invert’ button on the remote for my Devialet amp. No problem sorting that out
Regarding ‘teensy-weensy little bit different’ depends strongly on the system used. For me, correct absolute phase means stronger musical expression.
