or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › Absolute Phase: The Next Frontier (yeah right)
New Posts  All Forums:Forum Nav:

Absolute Phase: The Next Frontier (yeah right)

post #1 of 36
Thread Starter 
An interesting article...
http://www.gcaudio.com/resources/howtos/absolutephase.html
post #2 of 36
I fail to see how the Fletcher test and the Haas effect makes sense.
post #3 of 36

absolute polarity (really what the article is about) does seem to be audible with test signals in headphones - I  haven't tried music but managed 10/10 ABX with a 2 tone test (posts back to 3/2/14 http://www.head-fi.org/newsearch?Search=SEARCH&action=disp&advanced=1&byuser=jcx&newer=1&numupdates=&order=descending&output=posts&replycompare=gt&sdate=0&search=polarity&sort=relevance&titleonly=0&resultSortingPreference=recency )

 

it doesn't seem that polarity control in  in production is a priority - probably because even if audible with speakers it is hard to tell without switching back and forth


Edited by jcx - 4/19/14 at 5:15am
post #4 of 36
Thread Starter 
I think the switches that say they reverse absolute phase on high end gear actually do more than just reverse phase.
post #5 of 36
Thread Starter 
Quote:
Originally Posted by jcx View Post

absolute polarity (really what the article is about) does seem to be audible with test signals in headphones - I  haven't tried music but managed 10/10 ABX with a 2 tone test

You could tell there was a difference, or you could tell which was which?
post #6 of 36

for me, that there was a difference - I wouldn't expect to pick it up without the ref to compare - ethan claimed the test tone with higher negative peaks sounded slightly flat to his ears - so maybe a good musician would have a good enough internal reference

post #7 of 36
Thread Starter 
I can see how you'd hear a difference when you could snap a/b back and forth. I just don't see how it matters, particularly with music.

I've never been able to tell absolute phase with my speakers. When I was a kid, I once went three days with non matching phase between my speakers and didn't notice it!
post #8 of 36
I've never been able to hear soundstage differences that changed with phase inversion. I do notice more than a perceptible difference with strong bass transients, like a kick drum, especially with headphones. Even when I hear a difference though, it's hard to say which is correct.....you need to know what side of the drum head it was miked on. When I set my system up I just make sure a positive voltage input drives the transducer toward me.
post #9 of 36
Thread Starter 
I can't see how human ears could tell if something was push/pull or pull/push when we're talking about the tiny space of time of a sound wave.

I can see telling a difference when you set them right next to each other lined up.
post #10 of 36
Quote:
Originally Posted by jcx View Post

absolute polarity (really what the article is about) does seem to be audible with test signals in headphones - I  haven't tried music but managed 10/10 ABX with a 2 tone test...

I'd like to see a spectrum analysis of the test tones to be sure that there aren't any differences emerging from any other cause that might be expected to be audible, independent of the phase switching. There's mony a slip 'twixt cup and lip.

Even if absolute phase is recognised to be audible using test tones, l'm still not convinced that the conditions required for audibility maintain in music.

My feeling about absolute phase is that if this has not proved a significant issue in approaching a century of recorded music and broadcasting, then it's not of much significance. There are many amplifiers in everyday use which invert, particularly simple tube amps, and which the users never complain about, at least in this respect.

All my designs are non-inverting, except when they are not for some special reason, such as component count or size. In other words, I don't allow considerations of phase to inhibit the use of particular topologies if I see them as desirable for some other reasons, but I always design to conserve phase if I can do so without significant penalty.

I don't trust the article when it maintains that polarity will be easily detected by listening to the relative sound of speakers while switching polarity. If this were the case, the audibility would be demonstrated absolutely. I keep up with the state of the art pretty much, and news of such doesn't take long to get around.

Looking at his listlist of howtos, I see top of the list 'Why Power Cables Make a Difference'. I didn't bother to read it.

w
post #11 of 36
Thread Starter 
That article claims that absolute phase can be reversed all through a CD from song to song. I'd like to know how that happens. I've never been in a recording studio that didn't have everything hard wired in phase or with plugs that were directional to guarantee phase polarity.

I can't imagine one studio being reversed phase to another either. Plus means plus and minus means minus.
post #12 of 36
I believe only because a kick drum is not a symmetric push pull. I'll try to isolate a clip where I can tell a difference.
Quote:
Originally Posted by bigshot View Post

I can't see how human ears could tell if something was push/pull or pull/push when we're talking about the tiny space of time of a sound wave.

I can see telling a difference when you set them right next to each other lined up.
post #13 of 36

I saw some good info at Hydrogen audio on this.  One way to hear it is to construct an asymmetrical waveform.  One example was 440 hz mixed with 880 hz with the 880 hz signal shifted 90 degrees later (in this case 90 degrees of the 440 wave).  At first you don't notice, but with a little practice you get the impression one is just they tiniest touch louder than the other.  At that point ABX in Foobar is a piece of cake with 9/10 or 10/10 easily possible over and over.   So absolute phase can be clearly audible with some signals.

 

Another way I found to make it a bit more audible signal is some minor compression of such an asymmetrical waveform.

 

There is a plug in in Audacity called a diode processor.  It can be set for various levels of diode-like function far below actual rectification.  At a low setting it isn't too much different than the distortion a single ended circuit does.  It is also pretty easy to hear with any combination of more than one tone.  So pretty obvious this effect can be audible sometimes.

 

Now is it something you could pick out with a musical recording?  If some of the chain was a bit asymmetrical in function (which I bet some mics are at higher levels) followed with a bit of compression along the way, well maybe.  But I haven't been able to ABX any recordings reliably.  Likely one of those cases where it might in some rare instances be barely audible, but such a small difference in practice as a music consumer is a trivial non-issue.   My guess would be if you hear it the most likely place is heavy drum shots.

 

Just a side comment from anyone wondering if their system is correct phase from beginning to end.  You can do the diode processing of a sine wave or white noise in Audacity and save the digital file.  Play it back over speakers at moderate volume.  Put a DC voltmeter across your speaker leads, and see what you read.  A small slowly varying plus DC voltage and you have correct over all system phase.  A small varying negative DC voltage and you have backwards absolute polarity for your system so swap polarity somewhere (like speaker leads).  As to determining the proper polarity for a piece of music I don't have any idea.

post #14 of 36
Ok, I've found a pretty good track that I can detect this on. It is from The Dark Knight Original Motion Picture Soundtrack, track 1, "Why So Serious?".  The section is at 3:26 where the insane sub-bass line kicks in.  About two seconds into the bass line you can detect a warble in the bass that when inverted sounds different.  The original is smoother and a little muted compared to inverted.  After isolating the sensation, I used Audacity to cut out the section and save it regular and inverted, then used Foobar to ABX.  I got this in Foobar the 1st time through using HD558:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/04/19 18:59:44

File A: C:\Users\WebmasterK\Desktop\inverted.wav
File B: C:\Users\WebmasterK\Desktop\original.wav

18:59:44 : Test started.
19:02:43 : 01/01  50.0%
19:03:11 : 02/02  25.0%
19:03:37 : 03/03  12.5%
19:03:50 : 04/04  6.3%
19:04:03 : 05/05  3.1%
19:04:21 : 06/06  1.6%
19:04:33 : 07/07  0.8%
19:04:46 : 08/08  0.4%
19:04:55 : 09/09  0.2%
19:05:27 : 10/10  0.1%
19:08:11 : Test finished.

 ----------
Total: 10/10 (0.1%)

 

links to the files I used:

https://drive.google.com/file/d/0BzGS20n43HCqOUpwS2o1MlZPeW8/edit?usp=sharing

https://drive.google.com/file/d/0BzGS20n43HCqcEVmZ1ZTTTRnUVU/edit?usp=sharing

 

Now, what is correct phase is anyone's guess.  Without a reference/knowledge of the recording setup for the master there is no way to tell that I can think of. 

 

Quote:
Originally Posted by GrindingThud View Post

I believe only because a kick drum is not a symmetric push pull. I'll try to isolate a clip where I can tell a difference.

Edited by GrindingThud - 4/20/14 at 6:35pm
post #15 of 36
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Absolute Phase: The Next Frontier (yeah right)