Testing audiophile claims and myths

[jcx]

I listened only to Sample 2, normal and then inverted. They sound the same to me. I listened on decent quality Sony headphones.

Usually a difference can be heard after inverting polarity only with low frequency content. And as I explained earlier, it's due to non-linearity in the speaker or headphone drivers, not the ear's ability to distinguish polarity.

--Ethan

 
ethan - please make the clarification of what can be heard in careful tests designed to point up the differences and what is perhaps a "practical" decision to ignore polarity in normal music mastering
 
polarity as I point out below (previously in this thread) can equivalent to substantial harmonic vs fundamental phase shift for asymmetric waveforms
 
and the audibility has been tested by recognized researchers, published in Audio Engineering Society Journal
 
http://www.aes.org/e-lib/browse.cfm?elib=3824
 

 ~~ On the Audibility of Midrange Phase Distortion in Audio Systems~~
 
The current state of our knowledge regarding the audible consequences of phase nonlinearities in the audio chain is surveyed, a series of experiments is described which the authors have conducted using a flexible system of all-pass networks carefully constructed for this purpose, and some
conclusions are drawn regarding the audible effects of midrange phase distortions. It is known that the inner ear possesses nonlinearity (akin to an acoustic half-wave rectifier) in its mechanical-to-electrical transduction, and this would be expected to modify the signal on the acoustic nerve in a manner which depends upon the acoustic signal waveform, and so upon the relative phase relationships of the frequency components of this signal. Some of these effects have been known for over 30 years, and are quite audible on even very simple signals. Simple experiments are outlined to enable the readers to demonstrate these effects for themselves. Having satisfied ourselves that phase distortions can be audible, the types of phase distortions contributed by the various links in the audio chain are surveyed, and it is concluded that only the loudspeaker contributes significant midrange phase nonlinearities. Confining the investigation to the audibility of such phase nonlinearities in the midrange, circuitry is described which enables such effects to be assessed objectivbely fo their audible consequences. The experiments conducted so far lead to a number of conclusions. 1) Even quite small midrange phase nonlinearities can be audible on suitably chosen signals. 2) Audibility is far greater on headphones than on loudspeakers. 3) Simple acoustic signals generated anechoically display clear phase audibility on headphones. 4) On normal music or speech signals phase distortion appears not to be generally audible, although it was heard with 99% confidence on some recorded vocal material. It is clear that more work needs to be done to ascertain acceptable limits for the phase linearity of audio components-limits which might become more stringent as improved recording/reproduction systems become available. It is stressed that none of these experiments thus far has indicated a present requirement for phase linearity in loudspeakers for the reproduction of music and speech.
 
Authors: Lip****z, Stanley P.; Pocock, Mark; Vanderkooy, John

 
and since no one seems to actually look back or even clik on links:

  I put a a pair of polarity test files in the alt.zip over on diyAudio since I can't attach them here
 
 http://www.diyaudio.com/forums/everything-else/54596-audibility-absolute-phase-5.html#post3841138
 
also included the LTspice file that generates the .wav
 
edit: rethought dither and concerned that I may not have LTspice waveform data compression turned off so I reposted as audio.zip in the same diyAudio thread

 
 

  https://web.archive.org/web/20110101113016/http://www.ocf.berkeley.edu/~ashon/audio/primer1.htm does concentrate on phase coherence - some of the info supports absolute polarity discrimination - page down past the missing graphs

 
Although not in large numbers, previous research in investigation of the audibility of phase distortion has proven that it is an audible phenomenon. Lip****z et al. [7] has shown that on suitably chosen signals, even small midrange phase distortion can be clearly audible. Mathes and Miller [8] and Craig and Jeffress [9] showed that a simple two-component tone, consisting of a fundamental and second harmonic, changed in timbre as the phase of the second harmonic was varied relative to the fundamental. The above experiment was replicated by Lip****z et al., with summed 200 and 400 Hz frequencies, presented double blind via loudspeakers resulting in a 100% accuracy score.

the 2nd harmonic relative phase shift test does amount to a test of polarity inversion audibility when the phase between the fundamental and the 2nd shift relatively by 180 degrees
 
don't confuse the 180 degree phase shift of symmetric waveform, 1/2 period delay with polarity inversion - polarity can only be seen with a asymmetric waveform - requires a specific harmonic structure
 
200 Hz + 400 Hz/90 degrees sines 1:1 amplitude:
 

 
no amount of phase shifting, time delay of the above waveform will let you put the positive and inverted version on top of each other - both have 0 DC component, and a absolute polarity
 
reading diyAudio and hydrogenaudio threads on the subject shows much confusion about the basic terms, assumptions

 

[EthanWiner]

ethan - please make the clarification of what can be heard in careful tests designed to point up the differences and what is perhaps a "practical" decision to ignore polarity in normal music mastering

I'm not promoting that people honor or ignore absolute polarity. My entire point is simply that it's not audible, and when it is it's due to the speaker drivers and not how we hear.

the audibility has been tested by recognized researchers, published in Audio Engineering Society Journal

That article addresses all-pass filters and their frequency-selective phase shift. That's not usually audible either, though it can be in certain cases. But phase shift has nothing to do with polarity.

--Ethan

 

[jcx]

 ...But phase shift has nothing to do with polarity.

 
didn't look at the waveforms I posted? shift the 400 Hz 2nd harmonic by 180 relative to the fundamental and you get the equivalent of inverting the polarity of the 2-tone waveform
 
as I explained this can be done with frequency dependent phase shift - like Linkwitz XO - equivalent to all pass network phase shift - the test conditions are applicable to polarity inversion if the relative phase shift of the harmonic is enough
 

 ...not audible, and when it is it's due to the speaker drivers and not how we hear.

 
is not what Lip****z concludes
 
just read what I carefully collected, think about it without prejudging - the evidence is against both of your comments - absolute polarity detection is a property human hearing for some signals - and seems unimportant for music

 

[EthanWiner]

didn't look at the waveforms I posted? shift the 400 Hz 2nd harmonic by 180 relative to the fundamental and you get the equivalent of inverting the polarity of the 2-tone waveform

What amplifier or other audio device shifts 800 Hz by 180 degrees while leaving 400 Hz alone? Further, that shift is benign unless you combine the source with the output, which would never happen. You're really reaching here.

the test conditions are applicable to polarity inversion if the relative phase shift of the harmonic is enough

Phase shift can be audible if it's extreme, such as many thousands of degrees at midrange frequencies. If the lows arrive 1/4 second after the highs, well sure, you could hear that. But audio gear doesn't have anywhere near that much phase shift, and what phase shift does occur is generally at the highest and lowest frequency extremes. So again this is a reach.

the evidence is against both of your comments - absolute polarity detection is a property human hearing for some signals - and seems unimportant for music

LOL, I ask yet again for an example Wave file that shows absolute polarity being audible.

--Ethan

 

[jcx]

still haven't read without your preconceptions filtering everything - go to your DAW, add the 200/0 degree + 400 Hz/90 degree sines 1:1, then shift the 400 Hz by an additional 180, add to same 200 Hz/0 degree sine - look at the waveforms - think about their polarity
 
any can do this at home, in free tools - Audacity, math tools that can export .wav like SciLab, XCAS, languages like Python with SciPy - I used a free Spice circuit simulator because I know it well
 
you are missing the point of my examples of frequency dependent phase shift - they are to show that the Lip****z paper's test conditions are relevant to the question of polarity for specially constructed signals
 

 ~~ LOL, I ask yet again for an example Wave file that shows absolute polarity being audible

 
more evidence you haven't fairly read my posts - the diyAudio link where I attached the zipped .wav files only shows a single download of the file as I write this - and at least one of the spice .asc file downloads was me
 
http://www.diyaudio.com/forums/everything-else/54596-audibility-absolute-phase-5.html#post3841138
 
can't say the difference is immediately obvious to my ear - but I believe the Lip****z paper results, have heard 1st hand from John Allen that students quickly trained over a few hours of testing to hear some multi-way loudspeaker's XO phase shift in DBT - which was not a welcome result given his advocacy of 3-way or higher frequency partitioning to reduce loudspeaker IMD
 
if you grant that the frequency dependent phase shifts referred to in the paper are audible - then you only have to do the simple waveform summing I describe to see the equivalence to polarity being audible
 
 
EDIT it turns out that I can hear the difference, just can't articulate it - foobar2000 abx, motherboard Realtek HD sound, HD600 direct from computer headphone connector
 

 

[bigshot]

How many angels can dance on the head of a pin?
 
If it isn't able to be heard by human ears or even reproduced by the best transducers, who cares?
 
I have to say that although a lot of people seem to put stock in foobar test results on the internet, I honestly don't see how that reflects on any particular aspect of sound unless an unbiased third party verifies it and presents all the information on how the test files were prepared and the methodology behind the test. I can create two nice files (one with a nice big click in it) and generate reams of 10 out of 10 report cards.  Not to besmirch anyone's integrity, but unconscious bias isn't the only thing that needs to be eliminated to get a fair test. Systems can be gamed.

 

[wakibaki]

  EDIT it turns out that I can hear the difference, just can't articulate it - foobar2000 abx, motherboard Realtek HD sound, HD600 direct from computer headphone connector
 

 
I'd like to see an fft of the recaptured audio. I guess realistically that's a bit much to ask. This is a stress test for a transducer and indeed for the whole system and might reasonably be expected to reveal an assymmetric response, so the fact that you can hear a difference doesn't necessarily mean that much. It could just mean that your system is not transparent in this sense.
 
We understand and accept, of course, that there are limits to the stress that reproducing systems are exposed to in normal use, and limits to the quality of transducers, so this is not necessarily a serious reflection on the overall quality of your equipment, particularly when used for listening to music.
 
w

 

[jcx]

  How many angels can dance on the head of a pin?
 
If it isn't able to be heard by human ears or even reproduced by the best transducers, who cares?
 
I have to say that although a lot of people seem to put stock in foobar test results on the internet, I honestly don't see how that reflects on any particular aspect of sound unless an unbiased third party verifies it and presents all the information on how the test files were prepared and the methodology behind the test. I can create two nice files (one with a nice big click in it) and generate reams of 10 out of 10 report cards.  Not to besmirch anyone's integrity, but unconscious bias isn't the only thing that needs to be eliminated to get a fair test. Systems can be gamed.

I did supply all the pieces in my posts collected just above, files linked on diyAudio since we don't seem to be able to attach them here at head-fi
 
I invite others to look, listen, inspect the .wav in Audacity or whatever you are familiar with, post your own results - not at all claiming posting my results is absolute proof
 
I could get out the Radio Shack SPL meter with a cardboard "adapter" for a indication within a few dB - but all of the sound system, mixer and foobar's volume sliders were each below -3 dB and the HD600 is a very begin 300 Ohm load with very good linearity
 
the test files are 330 mVrms (or 33%), 670 mV peak - well below digital clipping  (LTspice maps .wav full scale range to +/-1.0 Vpp)
 
 
I believe Ethan is simply wrong - the human psychoacoustic, and animal physiologic auditory nerve electrode probe evidence support our hearing being able to discern absolute polarity below a few kHz
 
 
if you believe that the relative phase of a musical tone's harmonics contribute to timbre in addition to the relative amplitudes then you already are more than halfway to the audibility of polarity
 
 
foobar2000, ABX plugin lets many ears repeat the tests - put your eyeballs on the .wav, ears on the sound, create the test waveforms yourself - instead of carping about "possibilities"
 
 
its hard to see how small amounts of distortion could affect the results - for the asymmetric test wav 2nd harmonic is already 100%
 
only asymmetric clipping could possibly "explain" - and at the <90 dB SPL (wouldn't drown out normal conversation) level I doubt any clipping is occurring

 

[bigshot]

I'm not referring to you specifically, but the person being tested shouldn't really be the one conducting the test. I can post tests stating that I can see dead people and post charts saying I can to back it up, but unless an independent authority verifies the evidence, I could just be talking through my hat and fabricating the proofs. Again, I am not specifically referring to your test, but the general trend of posting foobar test results as "proof".

 

[jcx]

try getting specific then - look at my files, listen, tell us where problems are or admit that many have wrongly "learned" about absolute polarity - the posted refs are respected researcher's findings
 
if you keep harping on the "possible errors" I could feel insulted when I have provided the specifics, they are directly accessible, testable by you - particularly for someone with audio production tools, experience
 
despite the Science saying otherwise somehow much of the pro audio world has come to believe absolute polarity is never audible - not that it is just seldom easily audible, mostly safely ignored
 
 
while I couldn't tell immediately, my total "training" probably amounts to <10 min , few minutes at a time over a few days out of the past week - possibly the change to HD600 vs SA5000 helped too
and I have only long abandoned high school band level musical training - no mixing/sound production experience, 50+ year old ears

 

[bigshot]

Are you saying that with your files you think I'll be able to hear the difference between absolute phase and reversed absolute phase?

 

[analogsurviver]

Are you saying that with your files you think I'll be able to hear the difference between absolute phase and reversed absolute phase?

This is a LONG story - IF it was from the start put right, it would have been mandatory and by now non-issue. In an ideal world, perhaps.
 
In real world, there is ANY number of people who simply do not want this to be spread; equipment manufacturers, producers, record companies, etc, etc - because it forces everybody to work with better equipment with greater commitment - not necessarily resulting in better financial compensation for their effort. It is FAR  cheaper to declare (or "order" a paper by a esteemed member of AES ) absolute polarity to be inaudible - than rectifying ALL the faults in one's equipment and ways to use it.
 
On GOOD equipment, absolute polarity reversal is easily audible. Whether or not one is willing to take the commitment - that is another question.

 

[bigshot]

I have good equipment, no problem about that, so it should be easily audible to me, right?

 

[analogsurviver]

I have good equipment, no problem about that, so it should be easily audible to me, right?

It depends what you are using. 
 
It is one of the reasons why I insist on using my 80 mm full range speakers - they can not play very low or extremely high, they can not go loud (enough with some music - certainly not Mahler) - yet they are inherently uncapable of funny things with phase etc - such as most 2 and more way dynamic speakers. They have no trouble whatsover in discerning absolute phase - nor do Stax Lambda Pro headphones.
 
These speakers truly excel with vocal music - if the choir does not consist mainly of equivalents of  Rebroff ... ( when he swithes on his "subwoofer").
 
 

 

[jcx]

 ~~The experiments conducted so far lead to a number of conclusions. 1) Even quite small midrange phase nonlinearities can be audible on suitably chosen signals. 2) Audibility is far greater on headphones than on loudspeakers. 3) Simple acoustic signals generated anechoically display clear phase audibility on headphones. 4) On normal music or speech signals phase distortion appears not to be generally audible, although it was heard with 99% confidence on some recorded vocal material

 
again look at the waveform I construct - "phase distortion" of 180 degrees of the 2nd harmonic is equivalent to changing the polarity of the tone - so the JAES article I quote is directly relevant - and they say use headphones
 
 
Daisuke Koya's 2000 Master's Thesis, AURAL PHASE DISTORTION DETECTION also warns that it is much more difficult to hear with loudspeakers - so use headphones
 

...In a broad sense, the average correct responses for the loudspeaker-based listening test were significantly lower than for the headphone-based test...
...Table 5.1 indicates that even for the headphone listening test, phase distortion audibility was of very subtle nature...
...Table 5.2 indicates for the loudspeaker listening test that the overall phase distortion audibility was increasingly difficult as compared to the headphone listening test...

 
 
and this is head-fi - we should care even if the effect can only be heard in headphones - because we use headphones