|
|
Post by zonkola on Jan 24, 2021 15:52:31 GMT -6
I see these terms used interchangeably, but I believe it actually works like this:  |
|
|
|
Post by ragan on Jan 24, 2021 16:26:54 GMT -6
I see these terms used interchangeably, but I believe it actually works like this: View AttachmentPolarity refers to the positivity/negativity of a single signal. At any given point, flip the polarity and everything positive will be negative and vice versa. Phase has to have to some reference. It could be referencing one signal to another (as we do in multi-mic'd situations like a drum kit) or it could be in reference to some point in time (like zero). Moving signals around in time will change their phase compared to a given reference. In any case, when you've got multiple signals in some medium, they will interfere with each other. Constructive interference will increase amplitude (like in the top and bottom of your chart), destructive interference will decrease amplitude (like in the middle of your chart). The top and bottom of that chart are going to produce the same result, by the way, a doubling of amplitude. "Sounds dope", as is often the case, is just louder here. |
|
|
|
Post by jeremygillespie on Jan 24, 2021 16:46:25 GMT -6
You’ve also got to consider that the chart shown works great with a sine wav, but once you move away from that these things are frequency dependent. As always, use your ears. Nudging tracks around in the daw on a harmonically rich instrument with lots of microphones can lead to some really messy results.
|
|
|
|
Post by zonkola on Jan 24, 2021 16:57:35 GMT -6
Yep, the chart is an attempt to clarify the difference between the two terms. Most real-world phase issues won't involve sine waves or be as neat and tidy as a 180 degree shift, but the basic concepts still apply.
|
|
|
|
Post by matt@IAA on Jan 24, 2021 17:39:59 GMT -6
The reason it gets confusing is because a periodic signal 180 degrees out of phase with itself is the same thing as polarity reversal. Only because with a periodic signal absolute time reference drops out.
|
|
|
|
Post by drumsound on Jan 24, 2021 17:40:14 GMT -6
I see these terms used interchangeably, but I believe it actually works like this: View AttachmentPolarity refers to the positivity/negativity of a single signal. At any given point, flip the polarity and everything positive will be negative and vice versa. Phase has to have to some reference. It could be referencing one signal to another (as we do in multi-mic'd situations like a drum kit) or it could be in reference to some point in time (like zero). Moving signals around in time will change their phase compared to a given reference. In any case, when you've got multiple signals in some medium, they will interfere with each other. Constructive interference will increase amplitude (like in the top and bottom of your chart), destructive interference will decrease amplitude (like in the middle of your chart). The top and bottom of that chart are going to produce the same result, by the way, a doubling of amplitude. "Sounds dope", as is often the case, is just louder here. This is the crux of the biscuit. PHASE has to do with frequency and time, POLARITY is literally switching positive and negative. We use the Polarity flip to hear differences in Phase relationships. Once you get into sliding tracks around in a DAW, you are altering Phase Relationships, which is also what happens if you move microphones, though the results are often very different. |
|
|
|
Post by matt@IAA on Jan 24, 2021 17:41:54 GMT -6
To make matters even more confusing there is a signal phase relationship that’s in the frequency domain and not the time domain and has nothing to do with distance or polarity. 😁
|
|
|
|
Post by ragan on Jan 24, 2021 17:48:53 GMT -6
To make matters even more confusing there is a signal phase relationship that’s in the frequency domain and not the time domain and has nothing to do with distance or polarity. 😁 Do you mean in acoustic waves or are you only talking about electrical signals and the frequency-dependencies that come from various inductances and capacitances? |
|
|
|
Post by zonkola on Jan 24, 2021 17:56:52 GMT -6
That's what I tried to illustrate in the chart, but I apparently did very poorly.  But if there's a "phase" switch on a mic preamp that actually time shifts a signal, I haven't run across it yet. Don't get me started on vibrato vs tremolo... |
|
|
|
Post by matt@IAA on Jan 24, 2021 18:31:06 GMT -6
Do you mean in acoustic waves or are you only talking about electrical signals and the frequency-dependencies that come from various inductances and capacitances? Same thing. Any system described in the frequency domain can have a phase-amplitude relationship that is frequency dependent. It's the same for audio, the math is the same. You can use the exact same math that describes the resonant frequency of an RLC filter to describe the behavior of a rotating shaft with an eccentric unbalance going through a critical speed (frequency). Or a guitar string when you hit a harmonic. At the critical frequency, a node is created, and the waveform of the system goes through a 180 degree phase shift. The sharpness of this phase shift depends on the damping of the system. In mechanical systems that means once the system goes through a critical speed the motion is different. With an audio system that means that a, say, 30 Hz signal going through the same circuit simultaneously may leave the circuit out of phase from a 30 kHz signal. Maybe a few degrees....maybe 180 degrees. |
|
|
|
Post by ragan on Jan 24, 2021 18:41:50 GMT -6
Do you mean in acoustic waves or are you only talking about electrical signals and the frequency-dependencies that come from various inductances and capacitances? Same thing. Any system described in the frequency domain can have a phase-amplitude relationship that is frequency dependent. It's the same for audio, the math is the same. You can use the exact same math that describes the resonant frequency of an RLC filter to describe the behavior of a rotating shaft with an eccentric unbalance going through a critical speed (frequency). Or a guitar string when you hit a harmonic. At the critical frequency, a node is created, and the waveform of the system goes through a 180 degree phase shift. The sharpness of this phase shift depends on the damping of the system. In mechanical systems that means once the system goes through a critical speed the motion is different. With an audio system that means that a, say, 30 Hz signal going through the same circuit simultaneously may leave the circuit out of phase from a 30 kHz signal. Maybe a few degrees....maybe 180 degrees. Right, I get that, it was the "that's not in the time domain" that was throwing me. If we take, say, a capacitor, the current through it is going to be the time derivative of the voltage across it (scaled by its capacitance). It's, of course, often easier to analyze its behavior in the frequency domain but it's also very much existent in the time domain. Same with the rotating shaft, no? If you critically/over/under dampen it you're affecting its harmonic motion over time, yeah? Or maybe I misunderstand what you mean by "not in the time domain". |
|
|
|
Post by jeremygillespie on Jan 24, 2021 18:56:41 GMT -6
That's what I tried to illustrate in the chart, but I apparently did very poorly. But if there's a "phase" switch on a mic preamp that actually time shifts a signal, I haven't run across it yet. Don't get me started on vibrato vs tremolo... The majority of that misunderstanding can probably be attributed to Leo Fender. |
|
|
|
Post by zonkola on Jan 24, 2021 19:22:23 GMT -6
That's what I tried to illustrate in the chart, but I apparently did very poorly. But if there's a "phase" switch on a mic preamp that actually time shifts a signal, I haven't run across it yet. Don't get me started on vibrato vs tremolo... The majority of that misunderstanding can probably be attributed to Leo Fender. Yep, that's my suspicion as well—Leo or one of his marketing guys didn't consult a dictionary when they rolled out the Stratocaster. |
|
|
|
Post by zonkola on Jan 24, 2021 19:30:44 GMT -6
Do you mean in acoustic waves or are you only talking about electrical signals and the frequency-dependencies that come from various inductances and capacitances? Same thing. Any system described in the frequency domain can have a phase-amplitude relationship that is frequency dependent. It's the same for audio, the math is the same. You can use the exact same math that describes the resonant frequency of an RLC filter to describe the behavior of a rotating shaft with an eccentric unbalance going through a critical speed (frequency). Or a guitar string when you hit a harmonic. At the critical frequency, a node is created, and the waveform of the system goes through a 180 degree phase shift. The sharpness of this phase shift depends on the damping of the system. In mechanical systems that means once the system goes through a critical speed the motion is different. With an audio system that means that a, say, 30 Hz signal going through the same circuit simultaneously may leave the circuit out of phase from a 30 kHz signal. Maybe a few degrees....maybe 180 degrees. While I figured there'd eventually be references to room resonance and comb filtering, I have to admit that "rotating shaft with an eccentric unbalance" wasn't on my bingo card for this thread. It does bring back fond memories of installing a crankshaft in a 289 V8, though. |
|
|
|
Post by matt@IAA on Jan 24, 2021 19:59:51 GMT -6
Haha - my background is as a mechanical engineer in the field of turbomachinery. Rotating unbalance is my jam. 😁
|
|
|
|
Post by matt@IAA on Jan 24, 2021 20:02:31 GMT -6
Right, I get that, it was the "that's not in the time domain" that was throwing me. If we take, say, a capacitor, the current through it is going to be the time derivative of the voltage across it (scaled by its capacitance). It's, of course, often easier to analyze its behavior in the frequency domain but it's also very much existent in the time domain. Same with the rotating shaft, no? If you critically/over/under dampen it you're affecting its harmonic motion over time, yeah? Or maybe I misunderstand what you mean by "not in the time domain". The not in time domain is because the phase shift is for frequencies relative to other frequencies, not in time. Time is the same. Or in the case of a mechanical system the phase is not time dependent but in how the motion is in time (what direction the vibration / amplitude is in, not when it happens). they’re related but not the same. |
|
|
|
Post by ragan on Jan 24, 2021 21:10:37 GMT -6
Right, I get that, it was the "that's not in the time domain" that was throwing me. If we take, say, a capacitor, the current through it is going to be the time derivative of the voltage across it (scaled by its capacitance). It's, of course, often easier to analyze its behavior in the frequency domain but it's also very much existent in the time domain. Same with the rotating shaft, no? If you critically/over/under dampen it you're affecting its harmonic motion over time, yeah? Or maybe I misunderstand what you mean by "not in the time domain". The not in time domain in is because the phase shift is for frequencies relative to other frequencies, not in time. Time is the same. Or in the case of a mechanical system the phase is not time dependent but in how the motion is in time (what direction the vibration / amplitude is in, not when it happens). they’re related but not the same. Hmm. I don’t really get it. What are a couple of terms I could look up, particularly in the electronic paradigm? |
|
|
|
Post by matt@IAA on Jan 24, 2021 21:18:02 GMT -6
This page is cool. lpsa.swarthmore.edu/Bode/BodeWhat.htmlBut imagine a bode plot of a peaking filter which has (of course) a phase shift. Then imagine sending a compound sine wave of two frequencies, one lower frequency than the filter phase change and one higher - but both with the same period. The lower one will be phase misaligned to the higher one due to the circuit transfer function. Now.. is that a time effect or frequency? If you move the mic, does the phase mismatch change? But... what if you change the frequency of the input signals? Then, to really drive it home, it *looks* like time with periodic signals because time and frequency are the same thing. But what if it isn’t a periodic signal? A kick drum with low and high frequency components - happening once - and through a filter they will be out of phase. But not with respect to time. |
|
|
|
Post by ragan on Jan 25, 2021 2:10:38 GMT -6
This page is cool. lpsa.swarthmore.edu/Bode/BodeWhat.htmlBut imagine a bode plot of a peaking filter which has (of course) a phase shift. Then imagine sending a compound sine wave of two frequencies, one lower frequency than the filter phase change and one higher - but both with the same period. The lower one will be phase misaligned to the higher one due to the circuit transfer function. Now.. is that a time effect or frequency? If you move the mic, does the phase mismatch change? But... what if you change the frequency of the input signals? Then, to really drive it home, it *looks* like time with periodic signals because time and frequency are the same thing. But what if it isn’t a periodic signal? A kick drum with low and high frequency components - happening once - and through a filter they will be out of phase. But not with respect to time. Thanks Matt. I’ll check this out. Edit: I read that page. It is indeed a fun, interactive little tool. I mean, I’m marinating in phasors and Bode plots and transfer functions and moving between time and frequency domain with Laplace/Fourier transforms and whatnot. It’s not those principles I was confused about. I’m just still trying to figure out how any of this is divorced from time. A system responds differently to different frequencies precisely because a given excitation is asking it for certain behaviors over time and it can’t facilitate them all equally. Mass on spring, compound periodic voltage, whatever, the various frequency components that make up a given excitation are going to be handled differently, thus the phase shift. I’m still missing where any of this detaches from the time domain. |
|
|
|
Post by ragan on Jan 25, 2021 2:15:23 GMT -6
...a compound sine wave of two frequencies, one lower frequency than the filter phase change and one higher - but both with the same period. But I don’t get this. Frequency and period are reciprocals (f = 1/T) so how is anything periodic going to have a different frequency but the same period? |
|
|
|
Post by matt@IAA on Jan 25, 2021 7:43:38 GMT -6
Sorry it was late and brain no work good. Should have said different frequencies and same phase.
|
|
|
|
Post by drumsound on Jan 27, 2021 22:15:26 GMT -6
That's what I tried to illustrate in the chart, but I apparently did very poorly. But if there's a "phase" switch on a mic preamp that actually time shifts a signal, I haven't run across it yet. Don't get me started on vibrato vs tremolo... The majority of that misunderstanding can probably be attributed to Leo Fender. Explain... |
|
|
|
Post by the other mark williams on Jan 27, 2021 23:56:28 GMT -6
The majority of that misunderstanding can probably be attributed to Leo Fender. Explain... Tremolo = change in volume Vibrato = change in pitch Yet... The Strat had a "tremolo bar" when in fact what it was changing was the pitch of the strings (which is really vibrato). And then he compounded it with his amps that had a "vibrato" circuit which was in fact actually changing the volume of the signal (which is really tremolo). And yet he changed the world forever. |
|
|
|
Post by drumsound on Jan 28, 2021 9:01:10 GMT -6
I didn't realize that Leo was the one that called in a Tremolo bar. I do know that some amps say tremolo and some say vibrato, and that always seemed strange to me.
I'll give Leo a pass though because he got SO MANY THINGS RIGHT.
|
|
|
|
Post by the other mark williams on Jan 28, 2021 10:33:46 GMT -6
I didn't realize that Leo was the one that called in a Tremolo bar. I do know that some amps say tremolo and some say vibrato, and that always seemed strange to me. I'll give Leo a pass though because he got SO MANY THINGS RIGHT. AMEN. |
|
But if there's a "phase" switch on a mic preamp that actually time shifts a signal, I haven't run across it yet.
