Absolute phase

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
What are people's views on preserving the absolute phase of a system ? Is it audible ?

My current amplifier project uses an inverting, shunt feedback configuration, as this gives best performance. But should I correct for the inversion, either using an inverting input buffer, or by reversing the speaker connections ?
 
Another question:

Is there any difference using these two setups in stereo

1. Two Inverting amplifiers with exchanged LSP connections
2. Two Non-inverting amplifiers with normal plus/minus connections


And what will happen if
Using alt 1) for left channel and alt 2) for right channel
 
I think there's a thread where Nelson Pass suggested to "correct" the absolute phase. I haven't bothered myself trying to find out first hand whether it is audible or not but I had an experience in digital area...

There was a confusion in DSP/DAC chips about signal inversion, and I couldn't find reliable information so I setup 2 circuits, 1 with inverted signal. The difference was clearly audible (may be the unequal circuit played big role there). But what I didn't expect before was, I had no idea which one is the "correct" one. I had a guess (that the one with less dynamic is incorrect), but how can I be sure?
 
"And what will happen if
Using alt 1) for left channel and alt 2) for right channel"

Can sound very bad at high drive levels. This is due to dynamic off-set of amplifier when handling asymetrical signals (music).

Carver (and others) tried this and then dropped.
 
Hi Djk,
I don't understand your logic for arriving at your answer.

The two options require two different amplifiers.
An inverting amplifier on left channel with the loudspeaker connections reversed.
A non inverting amplifier on the right channel with conventional connections to the loudspeaker.

Both these arrangements will give correct absolute phase at the audio output (if the input signal is correctly phased).

Dynamic offset and asymetric signal handling can't interact between two different amplifiers,

Please explain further.
 
It's audible, but subtle. Most people don't notice. Listen to the impact of bass drums, for instance--it's slightly more pronounced when the absolute phase is correct.
If your amp is phase-inverting, reverse the speaker hookups.
This is the one thing that saved Douglas Self from being a complete fool in my eyes. He went on and on and on about how people were delusional when they claimed that they could hear things, then was man enough to admit that he was wrong when it was proved that absolute phase was, in fact, audible. Never had I seen a hard core measurements-are-everything guy admit to being wrong. They inevitably ignore instances when they are proven wrong (audibility of passive components being another case--see Jung & Marsh) and keep right on ranting. I respect him for his admission of error.
In any event, it's a low-level phenomenon. It's not the kind of thing that you should loose sleep over. But, being as easy to correct as it is, why not fix it?

Grey
 
Mr Rollins thanks, i dont understand yours or Mr self's explanation
of ab-phase being audible perhaps, if you could elaborate more
with some hard scientific evidence to back it, i would be more
convinced, i understand it like this it sounds good/better or
different, that makes it a subjective evaluation which is entirely
based on perception in other words if subjective evaluations had
a scientific basis there wouldnt be subjective and objective camps
i respect both you and self, but i think youre both alittle bit confused right now

regards
 
Personally, I'm rather amused that anyone would ever lump me into the same category as Doug Self.
It sounds as though mastertech has gotten himself lost in a philosophical conundrum.
Answer this, Grasshopper, and you will find the answer: How do we know anything at all?

Grey
 
"Please explain further."

Put a 'Y' cord on the input of the amplifier, push a pair of speakers next to each other, remove the grills.

As you increase the drive level with asymetrical program material one cone will move out, and the other move in. EVEN though both channels of the amplifier and both speakers are in phase with each other. It gets really bad when the amplifier clips.

There is an explaination of this at the ESP site.

"but i think youre both alittle bit confused right now"

It's a bit like trying to explain color to a color-blind person, they can see, but they really can't.

We know you aren't deaf, but you really don't hear.

Peoples' like Self are the worst, mainly as they can't hear, they won't bother to use their technical knowledge to figure out why an effect observed by those that can hear, could exist.

The whole thing fairly smacks of the witchcraft trials of yore, and hundreds of years later the decendants of the inquisitors are still unwilling to admit they were wrong (although the church of Rome made a statement a few years ago that they may have been wrong about the astronomer Galileo Galilei).
 
Administrator
Joined 2004
Paid Member
Hi djk,
What you are saying would indicate a DC bias in the music waveform. The first cap in the chain would "DC restore" the signal. The recording engineers should also be able to detect a slight "wind" from the session.

Not unless you are saying that some amplifiers create a DC error when playing music loud. This might be possible.

Just trying to rationalize this.

-Chris:xeye:
 
Since there's no 'official' standard for absolute phase it's a moot point anyway unless you build a phase inversion switch into your gear. All bets are off on multi-mic's recordings as well, it's possible but I've never heard of pop recordings using acoustic impulse tests to assure all the mics (and studios and locations!) used to assemble a mix adhered to pressure positive/voltage positive for example. Absolute phase may be audible but save for minimally mic'd classical recordings I'm not sure it makes sense to say there's a right and wrong way.
 
Administrator
Joined 2004
Paid Member
Hi rdf,
That's one of the things to be considered, but it's not even that easy. After the signal has been eq'd and processed, maybe sent to tape and back, the phase isn't 0° or 180° anymore. It's all over the place. Distance to mics also plays a role.

In a live recording, there are also reflections to deal with. Believe it or not, in PA work, there are phase testers out there. I've used them often enough.

I haven't spent a lot of time over this. I do have a feeling that the phase may be reversed on drums between recordings. I can't obsess over this.

-Chris
 
Hi anatech,

You reminded me of some of the tools we use, such as the phase scramblers common on microphone pre-amps for vocals. The circuits deliberately rotate phase in a frequency dependent manner to remove asymmetry for maximum cut through music. I'm told some studio engineers go so far as to run the entire mix through FM processors to achieve the desired sound. If you've seen what's inside these things absolute phase is the least consideration.
I still think it's audible, just irrelevant due to recording techniques for the vast majority of released product. For the rest (minimally mic'd acoustic) the only solution is an absolute phase reversal switch and preference.
 
"Not unless you are saying that some amplifiers create a DC error when playing music loud. This might be possible."

Yes, it's a design issue. For a partial discussion see the ESP site.

"I analyze when I design or critique."

It is simple matter to mark the CD if it's obviously backwards.

"Is this because one channel inevitably has higher gain and depresses the the B+ (or B-) for the other, "weaker", channel?"

No, at the risk of repeating myself: it is dynamic off-set caused by asymetrical signals. See the ESP site for a partial discussion.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.