Question about phase reversal...

[Ironklad Audio]

wow...this one's still going??

 

[vadoom]

Thank you! It is sure lonely when no one understand what one is talking about. Knowing that even one person come around is gratifying.

I have a degree in EE, and at this point I fully support you. I have played devils advocate, twisted and played with this whole concept. But when I go back to the basics, the offset seems to be an unyielding element that keeps phase shift and polarity inversion as two independent concepts.

I must admit that you argued well, kept to the point (pretty much to exhaustion) and stuck to the basics. I compliment you on that. Many people would tend to go astray early on...

I have no idea, to be honest. When you talk about "integrating the wave", you are at the edge of my understanding of mathematics terminology.

I was referring to calculus...
You know, when you integrate x^2, you get 2x. When you integrate a constant, it becomes 0....

I have just stuck to a simple mental picture all along. A phase inversion is nothing more than a phase change of a specific value that happens instantaneously (t=0). It is no different than any other phase change in that the phase rotates around the rest voltage; i.e. the rest voltage is not affected. This means that phase inversion only has the equivalent result of polarity inversion in one special circumstance, when the rest voltage is 0.

I think you meant polarity inversion is nothing more than a phase change of a specific value that happens instantaneously.
Humm... I don't think it is only at point t=0. It can happen at any point of t. It is just that there is no phase shift (same as time delay/shift). It just happen instantaneously.

Then I add the idea that phase change is not limited to a value of 180°. If one wanted to, they could change the phase by any damn degree they wished. There is nothing special about 180° in that regard. It's just another number. And regardless of degree value, the same formula should be used for all of them. One should just need to plug the value into the equation and away you go.

I agree.

Changing the sign, OTOH (a.k.a. multiplying the terms by -1) - which is basically all a polarity inversion is, is an artificial manipulation that has nothing to do with any phase change methodology. The fact that a polarity shift resembles a 180° phase change in some (but not all) aspects is nothing more than a happy coincidence that does nothing to actually make phase and polarity the same thing.

At this point, yes, I agree.



However, I still want to go far away to some mountain and re-contemplate this matter over again... Perhaps Maxwell equations would help?

 

[SouthSIDE Glen]

I think you meant polarity inversion is nothing more than a phase change of a specific value that happens instantaneously.

Nope. One of the resulting effects of a polarity inversion is a 180° phase inversion, but if the rest voltage is not 0, there is also a change in the rest voltage value and in the sinage of the wave values. With a phase change only, those extra changes will *never* occur.

For those familiar dealing with RF, maybe it might be easier to picture if one changes the playing field a little bit with the following analogy. Read my descriptions in this thread and replace the phrase "rest voltage" or "offset" with the phrase "carrier", and replace the term "waveform" with the term "amplitude modulation".

Any phase change to the signal, including an "inversion", would be effected on the modulation wave only, the carrier wave would be left untouched. A polarity inversion, OTOH, would affect the entire transmission, not just the modulated signal, and as such the carrier would be inverted as well.

In such an analogy, what we are now calling DC decoupling to ensure a 0DC rest voltage would be analogus to the supression of the carrier to ensure we are deaiing only with the sideband information. Sure, when stripping the carrier, polarity inversion and phase inversion will appear identical, because all we're dealing with is the modulated signal, but all it takes is the re-insertion of the carrier to break the symmetry and show that they are indeed different.

Before the circuit jokeys come in and object, yes the analogy is not perfect in that carrier waves are AC in nature and not DC biases. But that doesn't break the analogy, the salient points still hold.

Humm... I don't think it is only at point t=0. It can happen at any point of t. It is just that there is no phase shift (same as time delay/shift). It just happen instantaneously.

Sorry, I wan't necessarily referring to your equations when I said t=0. I probably shouldn't have used that equality; I would have avoided that confusion. All I was trying to say is that it was an inversion and not a shift; i.e. that it takes place IN zero time, not AT zero on the timeline .

However, I still want to go far away to some mountain and re-contemplate this matter over again... Perhaps Maxwell equations would help?

I never really wrapped my head around the calculus; my calc class in college was obscenely early in the morning and I kept nodding off .

It's funny that you mention Maxwell, though. At one point during the debate I was considering loosely (analogously, not literally) comparing polarity and phase to electricity and magnetism; under the right circumstances thay appear the same though under most normal circumstances thay exhibit quite different properties. But I backed off because that was an extremely faulty analogy; electricity and magnetisim are far more alike and related than phase and polarity will ever be.

G.

 

[noisewreck]

Wait till we start resorting to imaginary numbers!!

Meh, imaginary numbers on a two dimentional cartesian system are OK. I'll start getting concerned once you guys start discussing phase vs. polarity ramifications from electrons point of view, like whether the issue would be different if they had a different spin, and whether you can truly believe that there is no DC offset in all 31 dimentions.

 

[noisewreck]

Hey Vadoom... your tag line reminds me of Khazanov... "a potom arbuz s'yeli"

 

[SouthSIDE Glen]

Meh, imaginary numbers on a two dimentional cartesian system are OK. I'll start getting concerned once you guys start discussing phase vs. polarity ramifications from electrons point of view, like whether the issue would be different if they had a different spin, and whether you can truly believe that there is no DC offset in all 31 dimentions.

Or whether polarity and phase have no definition until we obesrve them? Or is it that you can't know both polarity and phase precisely at the same time? Or maybe that polarity and phase only appear to be equal undercertain conditions because we are in a universe conducive to human beings who can ovesrve them, but in the universe mext door, polarity more like the color blue and phase is more like the taste of stale coffee, and the closest thing to a human being is Michael Jackson.

G.

 

[The Axis]

telling them that all that button does is invert phase will give them the wrong picture, the wrong impression and the wrong idea of what is actually going on because that is NOT what the button is doing.

Except that, all mathematical theory aside, that is what the "Phase Reverse (or invert)" button of all commonly used audio equipment actually does.

Swapping the phase reverse button just swaps the 'high" lead for the 'low" lead.

As far as phase invert being equal to a 360°shift: That is valid IN THEORY for a single sinusoidal wave; or if EVERY Fourier Transformed frequency component is shifted 360 on an infinitely long or repetitive signal.....something that never eally happens in audio. It would have to by recorded and then digitally processed AFTER the recording. In practice, the high & low leads are simply swapped.

The theory is all fine, interesting, and valid. But you get the same effect as what is called "Invert Phase" by swapping the high lead for the low lead. You can even do it manually.

 

[SouthSIDE Glen]

Feed an analog synth signal that purposely has a nonzero DC component in it (yes, analog sound sculpters can and often do want that) and tell us the flipping the polaity is only flipping the phase.

Remix or remaster recordings that someone else made that has one or more non zero DC components in it becaue the rookie who did the original work used crap soundcars and never even heard of offset, and tell us that flipping the polarity is only flipping the phase.

This thread is a waste of time.

G.

 

[The Axis]

Feed an analog synth signal that purposely has a nonzero DC component in it .....
Remix or remaster recordings that someone else made that has one or more non zero DC components in it becaue the rookie who did the original work used crap soundcars and never even heard of offset, and tell us that flipping the polarity is only flipping the phase.

So there are some electrical conditions that result from faulty equipment that would require a different solution than merely flipping the phase. Who here thinks that a signal with a DC offset sounds different than the identical signal with no offset ?

But that isn't what is being discussed here. Going back to the original question:

My Firestudio and Digimax don't have any phase reversal switches (which kind of upsets me). I'd like to test phase on all of my channels, so my question is this:

In Cubase, in the mixer window, there's an option to flip phase on each channel. If I flip phase in monitoring mode, will it record with an inverted phase, or is it just phase delaying it? Thanks in advance

-Joel

It is flipping the phase in the monitored signal only. The recorded signal is unaffected. There is no "delaying" going on in EITHER signal.

Quote:
I have no idea, to be honest. When you talk about "integrating the wave", you are at the edge of my understanding of mathematics terminology.

I was referring to calculus...

Exactly why there is so much loony misinformation, myth, and "expert opinion" in audio acoustics.

 

[Robert D]

Feed an analog synth signal that purposely has a nonzero DC component in it (yes, analog sound sculpters can and often do want that) and tell us the flipping the polaity is only flipping the phase.

Remix or remaster recordings that someone else made that has one or more non zero DC components in it becaue the rookie who did the original work used crap soundcars and never even heard of offset, and tell us that flipping the polarity is only flipping the phase.

This thread is a waste of time.

G.

Feed it into what? Practically anything you feed it into is going decouple that DC offset, whether it has a phase switch or not. That's where this thread is a waste of time Glen, no offense. I get your theoretical argument. In the real world though, your distinction runs headfirst into a transformer, decoupling cap, or a servo, all of which render it null and void.

I appreciate all the math and postulating you and Vadoom have been doing, and at a higher level, your arguments hold a good deal of validity. But MCI is right, all that math applies only if your gear is busted.

 

[The Axis]

If we really want to open up a theoretical can of worms, let's discuss the usefulness of "dithering" ?

....or not.

 

[noisewreck]

Die, thread DIE!!!!

Or else, I WILL post some stuff with serious DC offsets and have all of you test your transformers, decoupling CAPs and other bits and pieces. Hell, I might even post some stuff with shifting DC offsets, just to make it more spicey for your equipment. Any takers for that test? Geebuss!

I might even put the mess through Kurzweil's WRAP DSP... and if any of you are familiar with that DSP, you'll know the extend of that threat!

To the original question:
The phase invert button and offline process in Cubase just performs polarity inversion. It simply multiplies the original wave with "-1" inverting it around 0-xing. Cubase doesn't have transformers, capacitors and other electronic components.

Die, thread DIE!!!!

 

[SouthSIDE Glen]

Die, thread DIE!!!!

"Thread's dead, baby. Thread's dead."
- Bruce Willis in "Phase Fiction"

G.

 

[vadoom]

Meh, imaginary numbers on a two dimentional cartesian system are OK. I'll start getting concerned once you guys start discussing phase vs. polarity ramifications from electrons point of view, like whether the issue would be different if they had a different spin, and whether you can truly believe that there is no DC offset in all 31 dimentions.

Uhm. That would be a project for quantum physics...

 

[vadoom]

Oh, wait. I thought this thread was still going... My bad


But anyways, to answer any questions why this some of the stuff I might have mentioned was so loony... well, let's just say that forum doesn't necessarily convey tone and intonation very well..........

Maybe I should integrate that.....................