skippy
New member
Bingo. So you were summing the two signals to mono, and not listening to them separately via either speakers or headphones?
D
Digidude824
New member
skippy said:
is that the right or wrong way to do it?
D
Digidude824
New member
and i am using monitors. Both tracks were played str8 up (pan that is), is that what you ment by mono? Should I have panned them far left/right? I guess if both tracks were at 0 with the pan then both speakers were playing both signals of both tracks. If I had them far left and right then they would be seperated, im guessing that this way would be better?
Light
New member
Digidude824 said:
The right way. This is exactly what you should be doing. It is a good idea to listen to it panned out, which will give you a strange, sort of uncomfortable, sound. This is what you must listen for when you are recording with multiple mics.
Light
"Cowards can never be moral."
M.K. Gandhi
Light
New member
bleyrad said:
You are absolutely right. I must apologize.
Polarity is an electrical issue. Pressure is an acoustical issue. I will now go and write that on the chalk board 100 times.
Light
"Cowards can never be moral."
M.K. Gandhi
V
volltreffer
Imply depressed
Don't polarize us...
Hey I read a lot about phase and polarity, but the problem that you seem to miss is the fact that the phase is ALWAYS a frequency dependant problem. The polarity switch can be seen as the very special case of a transfer function of -1 that is a phase shift of 180° for every single frequency... So I would not be too anal on the word, though it may be misleading...
@skippy: you write that the maximum phase shift is 180° - which is DEFINITELY wrong. The phase shift of 180° (and every shift with 180° + x*360°) is just the one with the most obvious effect: total cancellation.
But in fact it is the phase shifts that are bigger than 180° that make the comb filtering sound like comb filtering...
The problem is that the stuff is hard to understand unless you have the right mathematical background, and here we come back to Zeke Sayers questions some months ago about how to plan his life... It definitely is not harmful to have the necessary engineering background to understand all this... And a university engineering diploma might help you not to depend on your recording/mixing/marketing capabilities...
aXel
Hey I read a lot about phase and polarity, but the problem that you seem to miss is the fact that the phase is ALWAYS a frequency dependant problem. The polarity switch can be seen as the very special case of a transfer function of -1 that is a phase shift of 180° for every single frequency... So I would not be too anal on the word, though it may be misleading...
@skippy: you write that the maximum phase shift is 180° - which is DEFINITELY wrong. The phase shift of 180° (and every shift with 180° + x*360°) is just the one with the most obvious effect: total cancellation.
But in fact it is the phase shifts that are bigger than 180° that make the comb filtering sound like comb filtering...
The problem is that the stuff is hard to understand unless you have the right mathematical background, and here we come back to Zeke Sayers questions some months ago about how to plan his life... It definitely is not harmful to have the necessary engineering background to understand all this... And a university engineering diploma might help you not to depend on your recording/mixing/marketing capabilities...
aXel
Light
New member
You do not need an engineering diploma to understand this stuff, all you need is some basic trigonometry. But I do agree that the math is important.
As for the polarity switch altering "phase," well it really does not. Phase relationships are frequency dependent, but a phase shift is a time issue. With out moving the signal in time, you have not changed the phase relationships. A polarity switch can only "fix" one frequency at a time. Every other frequency will be thrown into a different, "non-aligned" non-aligned relationship. It may or may not sound better one way or the other, but it will still have phase incoherencies.
Light
"Cowards can never be moral."
M.K. Gandhi
As for the polarity switch altering "phase," well it really does not. Phase relationships are frequency dependent, but a phase shift is a time issue. With out moving the signal in time, you have not changed the phase relationships. A polarity switch can only "fix" one frequency at a time. Every other frequency will be thrown into a different, "non-aligned" non-aligned relationship. It may or may not sound better one way or the other, but it will still have phase incoherencies.
Light
"Cowards can never be moral."
M.K. Gandhi
skippy
New member
Re: Don't polarize us...
Actually, I didn't say that 180deg is the maximum shift: I'm not sure where you got that. I did say that the maximum cancellation will occur at 180deg, but that's as far as I went with maximums... I kept it to describing the first notch for simplicity, and that's probably where my description left too much out. It's hard to write for both a just-starting-out audience and an up-to-speed audience at the same time...
Of course shifts can be bigger than 180deg- and as you point out, it is the series of additive shifts that cause the "teeth" in the comb. With a given delay, the first notch in the comb occurs at the frequency for which that delay equates to a 180deg phase shift. The second notch is at the frequency for which that delay equates to 540deg of phase shift (360+180), the third at 900deg (720+180), and so on.
Here's the magic. When you push the polarity switch, the teeth in the comb *move*: the first notch occurs at the frequency for which that delay equates to a 360deg phase shift, the second to 720deg, and so on (and you actually also have a parasitic notch at 0Hz for the inverted case, so you have some low bass rolloff that will differ radically from the non-inverted case: this can actually be useful for VLF noise like air-handling systems and stuff, by the way). So that's why you play with both settings. Without moving the mics, you electrically move the notches.
Oh, well. I use so many words that I'm not surprised that some of 'em come out sounding wrong. It is extremely easy to get overly anal on the verbiage: and unfortunately this is one of those cases where common usage really has muddied the waters. Shoot, I have *textbooks* that misuse the term "phase" in this manner, implying a duality that is only correct in certain very special cases... Sigh. We does the best we can with what we got!
I'm in absolute agreement that having the right degree to depend upon for a living, so that you can afford to do your art, is a good move. I'd have starved a long time ago if I had continued trying to make a living with music! Being a professional nerd does pay better...
volltreffer said:
Actually, I didn't say that 180deg is the maximum shift: I'm not sure where you got that. I did say that the maximum cancellation will occur at 180deg, but that's as far as I went with maximums... I kept it to describing the first notch for simplicity, and that's probably where my description left too much out. It's hard to write for both a just-starting-out audience and an up-to-speed audience at the same time...
Of course shifts can be bigger than 180deg- and as you point out, it is the series of additive shifts that cause the "teeth" in the comb. With a given delay, the first notch in the comb occurs at the frequency for which that delay equates to a 180deg phase shift. The second notch is at the frequency for which that delay equates to 540deg of phase shift (360+180), the third at 900deg (720+180), and so on.
Here's the magic. When you push the polarity switch, the teeth in the comb *move*: the first notch occurs at the frequency for which that delay equates to a 360deg phase shift, the second to 720deg, and so on (and you actually also have a parasitic notch at 0Hz for the inverted case, so you have some low bass rolloff that will differ radically from the non-inverted case: this can actually be useful for VLF noise like air-handling systems and stuff, by the way). So that's why you play with both settings. Without moving the mics, you electrically move the notches.
Oh, well. I use so many words that I'm not surprised that some of 'em come out sounding wrong. It is extremely easy to get overly anal on the verbiage: and unfortunately this is one of those cases where common usage really has muddied the waters. Shoot, I have *textbooks* that misuse the term "phase" in this manner, implying a duality that is only correct in certain very special cases... Sigh. We does the best we can with what we got!
I'm in absolute agreement that having the right degree to depend upon for a living, so that you can afford to do your art, is a good move. I'd have starved a long time ago if I had continued trying to make a living with music! Being a professional nerd does pay better...
V
volltreffer
Imply depressed
@skippy: Sorry, if I misunderstood you... But at least there is my hope that some others might have misunderstood this, too and could get some clarity...
@light: If you draw your recording system or even your signal in frequency domain (viewing the amplitude/phase diagram), the only way to multiply by (-1) is the addition/subtraction of 180deg over the ***complete*** frequency spectrum. The thing is that additive cancellations are hard to handle in frequency domain; it is used for serial transfer elements (or maybe the stability consideration of closed loops).
My point was that I do not alter the phase ***relationships*** in a signal by shifting it by 180deg. Therefore any signal with inteference to another (that is NOT total cancellation or double amplitude - i.e. not a x*180deg phase shift over the complete spectrum, according to my definition) will only cuase the interference at other frequency points. Therefore the name of the phase button is at least misleading, as too many people seem to consider it as a silver bullet for phase problems.
The only 'proper' way to solve phase problems would be the use of delays that cancel the time the sound needs to arrive to the second mic. If you record your guitar with one mic close to the grill and another one in 1m distance, you might add a delay of 1/330 s ~ 3ms to the close miced track to compensate for the phase shift(s) that occur. Alas this won't work in the most pratical cases as you'll have more than one sound source (who guarantees that you don't have partial cancellations of the 4 speakers of you 4x12 box...)
Hope I could clarify my last post...
aXel
@light: If you draw your recording system or even your signal in frequency domain (viewing the amplitude/phase diagram), the only way to multiply by (-1) is the addition/subtraction of 180deg over the ***complete*** frequency spectrum. The thing is that additive cancellations are hard to handle in frequency domain; it is used for serial transfer elements (or maybe the stability consideration of closed loops).
My point was that I do not alter the phase ***relationships*** in a signal by shifting it by 180deg. Therefore any signal with inteference to another (that is NOT total cancellation or double amplitude - i.e. not a x*180deg phase shift over the complete spectrum, according to my definition) will only cuase the interference at other frequency points. Therefore the name of the phase button is at least misleading, as too many people seem to consider it as a silver bullet for phase problems.
The only 'proper' way to solve phase problems would be the use of delays that cancel the time the sound needs to arrive to the second mic. If you record your guitar with one mic close to the grill and another one in 1m distance, you might add a delay of 1/330 s ~ 3ms to the close miced track to compensate for the phase shift(s) that occur. Alas this won't work in the most pratical cases as you'll have more than one sound source (who guarantees that you don't have partial cancellations of the 4 speakers of you 4x12 box...)
Hope I could clarify my last post...
aXel
Light
New member
volltreffer said:
That is all correct, except that you should not use the term "phase" or "180 degree" when referring to what the polarity switch does. It does not shift anything, any number of degrees. Degrees of shift, and phase, are strictly time domain issues, and polarity has no time component. What you are calling a 180 degree shift is a simple electronic inversion. All that is positive becomes negative, and all that is negative becomes positive. This may seem like a small issue of nomenclature, but misuse in this area causes a lot of confusion.
Light
"Cowards can never be moral."
M.K. Gandhi
V
volltreffer
Imply depressed
High Light!
Might just be called different in systems engineering. What I learned is that the term 'phase' is strictly correlated to the (amplitude/phase) pair as a function of the frequency. Which comes from the fourier (or in systems engineering rather laplace) transformation. In that case it would be a strict frequency thing. A time domain thing would be a time delay...
But as I said, there may be subtle issues of neither talking in my own language as well as having a (surely related) but not identical engineering background. In any way, I admit without any doubt that the name phase switch is at least misleading, if not wrong...
aXel
P.S.: Edited your name.... Have to excuse... Sorry - these are the things I hate most, but sometimes they DO happen to me...
Might just be called different in systems engineering. What I learned is that the term 'phase' is strictly correlated to the (amplitude/phase) pair as a function of the frequency. Which comes from the fourier (or in systems engineering rather laplace) transformation. In that case it would be a strict frequency thing. A time domain thing would be a time delay...
But as I said, there may be subtle issues of neither talking in my own language as well as having a (surely related) but not identical engineering background. In any way, I admit without any doubt that the name phase switch is at least misleading, if not wrong...
aXel
P.S.: Edited your name.... Have to excuse... Sorry - these are the things I hate most, but sometimes they DO happen to me...
Last edited:
Light
New member
Well, we are both talking about the same thing here, and certainly we both understand the concept. Phase is a time issue which is frequency dependent, at least in the audio world. A phase shift is a difference in time. The degree of phase shift is how the phase shift compares to the specific frequency.
Light
"Cowards can never be moral."
M.K. Gandhi
Light
"Cowards can never be moral."
M.K. Gandhi