[chillysalsa]

So, is there any way to tell from a schematic if it inverts? For an integrated design, will it depend if the opamp is inverting? For a discrete design where does one look to distinguish phase?

 

[gaboo]

Quote:

Originally Posted by chillysalsa So, is there any way to tell from a schematic if it inverts?

Yes.

Quote:

For an integrated design, will it depend if the opamp is inverting?

Op amps are not inverting (or not) by themselves, it depends on the circuit. Look in Op amps for everyone, pages 3-3 & 3-4 cover the basic opamp circuits.

Quote:

For a discrete design where does one look to distinguish phase?

For BJT, common-emitter circuit is inverting, while common-collector is not. Most audio circuits are more complex than that, so take a look a tangent's web site, he has links to some nice books. Here is a free one.

 

[gaboo]

The samples sent to me by chillysalsa have been uploaded here:
http://www.hydrogenaudio.org/forums/...howtopic=28774

I had to reduce them to the first 10secs, but I've only used that for the ABX anyway...

 

[sluggo]

Quote:

Originally Posted by gaboo Can anyone get more obsessed with reverse phase?

I think Mr. Cardas has it right. Absolute phase is meaningful only for waveforms where the shape is asymmetric or where the "attack" (to use a musical term) is significantly higher in amplitude than the sustain (drums and cymbals being perfect examples).

I think a well-recorded gunshot would show the difference pretty clearly, and for normal music a quartet of trumpet, bass, drums, and piano would provide a pretty good test of the audibility of absolute phase.

I'm not sure how symmetric the waveform from a singing voice is, but I'm guessing a reversed-phase voiuce would be harder to detect except on plosives such as "p", "t", etc, as someone posted earlier.

I think for many electronic instruments the effect of phase inversion is meaningless, as they have a symmetric waveform and have no parallel in nature anyway - what is a fuzz guitar "supposed" to sound like?

For massed instruments or voices the issue would seem to be moot. There will be phase cancellation at the source, and the recording is simply a collection of the random phase information.

 

[Scrith]

I tried the test file and, after I was able to indentify what to look for by listening a few times, I was able to clearly differentiate between having the Invert Phase button checked in the E-mu PatchMix (semi-blind...I clicked the phase invert button repeatedly while my hand covered it on my screen and then played the music clip in foobar several times and was able to figure out whether I had left the button pressed or not every time). The final cymbal hit in each drum sequence (sorry I don't know the official terminology for this) actually appears to be generating a slightly different note (as in G# vs G, for example) when the phase is inverted (which I assume is the correct sound, since my amp is supposedly un-inverting it).

 

[gaboo]

Quote:

Originally Posted by Scrith I tried the test file and, after I was able to indentify what to look for by listening a few times, I was able to clearly differentiate between having the Invert Phase button checked in the E-mu PatchMix (semi-blind...I clicked the phase invert button repeatedly while my hand covered it on my screen and then played the music clip in foobar several times and was able to figure out whether I had left the button pressed or not every time). The final cymbal hit in each drum sequence (sorry I don't know the official terminology for this) actually appears to be generating a slightly different note (as in G# vs G, for example) when the phase is inverted (which I assume is the correct sound, since my amp is supposedly un-inverting it).

Since you are using foobar, it's trivial to do a proper ABX: just selectr tracks, right click, choose ABX...

 

[rudi]

Quote:

Originally Posted by kevin gilmore So all you people out there with tube amps and sennheiser headphones, flip those connectors around and see if you can hear the difference.

If you make the test in a completly silent room ( anecoic ) you can't hear difference inverting phase. Since sound is a compisition of sine waves, and sine waves make a circular movement, does not matter from which position virtual wheel start to run. In the real enviroment there are enviromental noises and sounds interacting with reproducted sound. When two sounds of same frequency have same phase the resulting is a sum, if phase is 180deg shifetd is a subtraction, changing reproduction phase act on this.
Since the nature, level, frequency and pahse of enviromental noises and sounds are inpredictable and chaotic, the phase of reproducted sound is not a rilevant value in reproduction.

different story in professional recodring and mixing where the relevant signals are mainly deterministc values.

 

[chillysalsa]

Yes, sound is all oscillations, but to say it's a regular sine wave is an oversimplification.

Here's a typical scope plotting a kick-drum beat:

The first thing your ear will hear is either a crest, or a tough in the pressure. Any other music will be asymetrical as well.

This makes a difference in the way it ends up sounding, and is why one can hear the change due to phase inversion.

 

[rudi]

No, your is a semplification, or better a convention to reppreent graphicaly the events in the time domain.
Any strange shape wave is only a composition of pure sine waves, at different frequencies and amplitiude. ( Fourrier)

 

[chillysalsa]

That's true rudi, but I just don't believe my brain is doing a Fourier transform and interpretting music as superimposed pure sinewaves. If you break down the sound into wavelets, you will get some coefficients to the summation equation. But then if you take the inverse to the sound, you get similar coefficients multiplied by -1, or invert the sinewaves by 180º, do you not? There is still a difference between inverted phase even if the brain is doing these Fourier transforms as we listen to music.

I personally think absolute phase means something. When I pluck a guitar string, or hit a drum, I'm convinced the brain identifies some quality of the sound best described as 'impact', from the fact that the first waves to hit your ear drum are of a positive pressure.

But everyone may come to their own conclusion.

 

[Publius]

This is useful. Current psychoacoustic research suggests that hearing is an inherently rectifying operation. Look at the pulse measures about a third of the way down the link: If an impulse is changed from positive to negative amplitude, the electrical pulse response of the ear is delayed by one half "cycle".

If the ear really is rectifying the signal then phase/polarity changes have some definite physical basis for being audible, although it's a matter of debate on exactly how audible it really is.

rudi, signal interactions with environment noises only matter when the signal to noise ratio is very low. For any listening we're caring about, it's going to be at least 60dB overall, and in the worst case at a particularly faint frequency, maybe 20dB. That means that even if the environmental noise at this frequency were completely and always out of phase with the signal, it would only attenuate that frequency by 0.9dB. If it were 90 derees out of phase with the signal, it would move the phase around by 5.7 degrees. And these are pretty much worst case numbers here...

As long as we're actually listening to music, ambient noise has absolutely no impact on phase audibility whatsoever.

 

[rudi]

Quote:

Originally Posted by chillysalsa I personally think absolute phase means something. When I pluck a guitar string, or hit a drum, I'm convinced the brain identifies some quality of the sound best described as 'impact', from the fact that the first waves to hit your ear drum are of a positive pressure. But everyone may come to their own conclusion.

This is right as principle. But immagine to hear your live drum from a distance of 10 mt and to have the beginning of "impact" as a positive signal at as you said,then move 1.5 mt back, now you ar out of phase of 180deg ( if we assume freq. is 100hz ) and your positive impact is now negative, but drum is still a drum ... which one is the right position to listen?...or better from which position microphones have recorded the sound?
If yo listen a live concert from third row or from tenth ( apart of diff sound pressure ) your brain detect the same music even if the phase ar differents.
Since there is no a mandatory distance for listening ( and recording ) the phase problem is not a problem since there are no possible solutions.

 

[gaboo]

Quote:

Originally Posted by rudi If you make the test in a completly silent room ( anecoic ) you can't hear difference inverting phase.

Even if that's true, we're listening to headphones here.

Quote:

Since the nature, level, frequency and pahse of enviromental noises and sounds are inpredictable and chaotic, the phase of reproducted sound is not a rilevant value in reproduction.

I respectfully disagree. Closed headphones limit most of the interference, and provide a predetermined and almost fixed environment. I doubt the shape of human ear varies enough to cause 180 degrees phase variance from person to person. Especially at low frequencies.

 

[gaboo]

Quote:

Originally Posted by rudi This is right as principle. But immagine to hear your live drum from a distance of 10 mt and to have the beginning of "impact" as a positive signal at as you said,then move 1.5 mt back, now you ar out of phase of 180deg ( if we assume freq. is 100hz ) and your positive impact is now negative, but drum is still a drum ... which one is the right position to listen?...or better from which position microphones have recorded the sound? If yo listen a live concert from third row or from tenth ( apart of diff sound pressure ) your brain detect the same music even if the phase ar differents. Since there is no a mandatory distance for listening ( and recording ) the phase problem is not a problem since there are no possible solutions.

That's only true for sine waves. Conside a an ideal (Dirac) impulse: it's phase is unchanged regardless of listening distance.

 

[rudi]

Quote:

Originally Posted by gaboo That's only true for sine waves. Conside a an ideal (Dirac) impulse: it's phase is unchanged regardless of listening distance.

Right, but drum ( like other musical instruments) are much more wave generators then "ideal Dirac Impulse" generators