phase shift & audio

I just got a 436c up and running from some spare bits I had lying around.

I put a square wave into the unit and get a response which angles up to the leading edge. 2 books I have say that the bass is leading phase wise when this happens.

The compressor sounds OK to my cloth ears.

I'm just trying to get a handle on whether this is necessarily a bad thing ? How does it affect the sound ? Is it just colouration ?

All the books I have explain how to test if this phase shift is taking place but don't seem to evaluate it !

Maybe some light can be shed on this topic ?

On steady tones it's utterly imperceptible. There's limited and highly subjective comment on whether moderate amounts (to the extent which will mutilate and uttely destoy many simple waveforms) may even be audible. -Be aware that the phase shift from sitting to standing in front of a 2-way speaker system will have a colossal effect and most people don't notice.

I say don't worry. About 10° will make a square wave look very funny, and also HF rolloff/limiting will slope the verticals also... and no tube limiter goes to light frequencies after all... (LA2a sidechain excepted! :wink: )

Don't worry, with an Altec, if it resembles a squarewave, you're ahead of the curve!

Keef

Keith,

Looking at a scope trace is there a way to measure the angle of phase shift ?

Also I rolled the sig gen freq around a bit & the response is quite flat using some old Parmeko audio transformers

[quote author="Rob Flinn"]Keith,

Looking at a scope trace is there a way to measure the angle of phase shift ?

Also I rolled the sig gen freq around a bit & the response is quite flat using some old Parmeko audio transformers[/quote]
you could use the x/y mode if you have one.
put the original signal on x and the signal from the output at y. normally you would get a straight line but if you have phaseshift you'll get an ellipse and normally there was a way tom measure and then calculate the phaseshift but i forgot but you can i know.

Thanks folks. I read my audiocyclopedia in a bit more depth last night & it confirms what you're both saying.

Beware of differences between the two wave forms if using the xy approach.
The two wave forms must be exact sine waves.
Your amplified signal might be different if it is being routed into a boat anchor from the years of lore.

Gentlemen;
I'm afraid that you mean a phase shift that is result of low inductivity of some transformer's coil, so you may expect some additional distortions due to core saturation.

I read a paper, though I don't remember where from (will try and dig it out and let you know) about phase shift and frequency response in audio.
The guy had done a lot of experimentation and came to the conclusion that phase shifts themselves are inaudible; it is only the resulting frequency response variations that affect what we hear.
Dunno if thats useful, but seems worth bringing up while on the topic.

[quote author="jeth"]I read a paper, though I don't remember where from (will try and dig it out and let you know) about phase shift and frequency response in audio.
The guy had done a lot of experimentation and came to the conclusion that phase shifts themselves are inaudible; it is only the resulting frequency response variations that affect what we hear.
Dunno if thats useful, but seems worth bringing up while on the topic.[/quote]

Also, if 2 channels of stereo are shifted differently it causes shift of binaural localization, of course, and sometimes more severe than difference in levels.

[quote author="jeth"]I read a paper, though I don't remember where from (will try and dig it out and let you know) about phase shift and frequency response in audio.
The guy had done a lot of experimentation and came to the conclusion that phase shifts themselves are inaudible; it is only the resulting frequency response variations that affect what we hear.
Dunno if thats useful, but seems worth bringing up while on the topic.[/quote]

This could be useful reading. I managed to get the offending 436 to my studio for some audio testing & it is lacking low end noticably. I have it back here now & am going to scope it through to try & see where this bass cut is being introduced.

[quote author="jeth"]The guy had done a lot of experimentation and came to the conclusion that phase shifts themselves are inaudible...[/quote]
This is an oversimplification, without some qualifying statements:

un-referred phase shifts may possibly be inaudible. That is to say that if I put a mono signal through a mono playback speaker, and only have a phase shifted or a NON phase shifted sound at any given instant, the difference may very well be inaudible. Play it in stereo and arrange so that a shift approaching 180° at the low end and I assure you, it's audible! :wink: -In other words, with something to refer to it's audible, but in the absence of a reference, it may be inaudible.

Even so, you have to specify steady-state or transient. While some people insist that phase shift is audible no matter what, my own tomfoolery has led me to believe that steady-state is alost assuredly inaudible, in so far as NOBODY has ever been able to identify phase shift during constant tones... even many hundreds of degrees. This has led me to ignore the significance of small phase shifts for steady-state instances.

Transients however, I'm not sure about. While I myself have not yet been able to correctly identify moderately large phase shifts on large, wide-spectrum transients; I am not yet ready to discount that other golden-eared people might. I haven't yet done enough testing; who knows? -I may yet even hear it myself! :wink:
[quote author="Fred Forssell"]...a recent debate in a discussion forum left me feeling disturbed by the emotion with which some people defend their position on this topic.[/quote]
Most wonderfully worded. It mirrors my own view precisely.
[quote author="Fred Forssell"]The question kept popping into my head, “Why not just listen to it and decide for yourself?”.[/quote]
100% agreement.

Keith

Will definitely have a poke around and try and find the article.
Don't remember his test procedure, but he used a number of "blind" listening tests with variations on phase response between woofer and tweeter I think. I guess his conclusions were a little over simplified, but I think this was intentional to make the point that in the real world of ears and sound reproduction systems what can be heard is secondary results of phase shift, rather than the phase shift itself.

http://www.ethanwiner.com/phase.html

Here it is, not as in depth as I remembered actually, sure there was another article with more test results and analysis.

http://www.hydrogenaudio.org/forums/lofiversion/index.php/t48921.html

Also came across this on another forum.

http://home.att.net/~lemon.j.russell/audphase.htm

Another brief but relevant piece...

http://sound.westhost.com/ptd.htm

And this last one is a good article, though it deals with loudspeaker issues rather than electronic circuits..

Hope thats of use.

As a further note, think I posted here before in my search for suitable circuits for time alignment of speaker systems. Never found anything, though there was some interesting discussion if you search up the thread.
Anyone has any circuit suggestions I'd be happy to hear them, preferably an analog solution.