Phase issues splitting mic/DI bass?

[fat_fleet]

Been playing around with using a Sansamp bass driver, but going from the parallel out to a bass amp and micing that, so I end up with two bass tracks (mic'd and DI). Weird thing is, it's kinda quiet in the overall mix but really pops when I flip the phase on one of the tracks. Is it common to have flipped phase between a mic'd and DI track?

 

[mixsit]

Prettty easy chance to have it happen that way either at the amp or the recording chain. Plus there's a bit of time offset from the mic. The phase effect will only be in the highs if you're close up on the speaker, down into the mids as you move several inches out.

 

[bouldersoundguy]

Been playing around with using a Sansamp bass driver, but going from the parallel out to a bass amp and micing that, so I end up with two bass tracks (mic'd and DI). Weird thing is, it's kinda quiet in the overall mix but really pops when I flip the phase on one of the tracks. Is it common to have flipped phase between a mic'd and DI track?

That's polarity, not phase (in spite of the common mislabeling of polarity switches). There might also be some time misalignment which would cause phase error which would alter the tone if the two signals are panned together. Polarity is in the amplitude domain (vertical on the waveform) and time/phase is in the time dimension (horizontal on the waveform).

Miswiring the speaker in the amp would invert the polarity. The distance from the speaker to the mic would create a time difference.

 

[Steenamaroo]

That's polarity, not phase (in spite of the common mislabeling of polarity switches). There might also be some time misalignment which would cause phase error which would alter the tone if the two signals are panned together. Polarity is in the amplitude domain (vertical on the waveform) and time/phase is in the time dimension (horizontal on the waveform).

Miswiring the speaker in the amp would invert the polarity. The distance from the speaker to the mic would create a time difference.

Take this to the bank.^^^^

 

[fat_fleet]

But if there was a time misalignment (phase error) of half the period of the waveform (180 degrees), doesn't that result in polarity inversion or is that something different?

 

[Steenamaroo]

In examples where sine waves are used, this looks to be the case, but it's not.

If you have two sine waves 180 degrees out of phase playing for ten seconds, they will cancel out completely between the first and last half cycles.
The first and last half cycles are audible because the waves start and end at different times.

Two waves of opposite polarity (that are otherwise identical) will start and end at the start at the same time, and completely cancel out.


I may be wrong here, but I think to say something is 180 degrees out of phase in relation to something else, you have to specify a frequency at which this is true. 360 degrees would be a full cycle at that frequency.

Can anyone confirm this?

 

[fat_fleet]

Yeah I don't know much about all that phase stuff except that it affects mics that vary in distance from the source.
The thing with my two bass tracks isn't really holding me back or anything, I just thought it was weird since I thought phase and polarity were related in the way I described a couple posts up. Guess not tho?

 

[Steenamaroo]

Negatory.

Feet, the best thing to do (apart from trusting your ears) is to set up to record via both inputs, maintain silence, then slap one note. Make sure it's a good clean note with a good attack.

Now zoom in like crazy on the transient of the two recorded wave forms and see if they are out of alignment with each other, or of opposite polarity.

If they have opposite polarity, one will go up at the very beginning, and one will go down.
Invert polarity on one of the tracks.

If they are out of phase, they won't start at the same time.
Measure the amount in ms and use a delay on the early one to compensate, so that all future recordings in that session are fine.

There almost certainly will be a phase discrepancy as one sound is travelling through the air and the other isn't.
They may well be of opposite polarity too.
Have a look and see.

 

[bouldersoundguy]

I may be wrong here, but I think to say something is 180 degrees out of phase in relation to something else, you have to specify a frequency at which this is true. 360 degrees would be a full cycle at that frequency.

Can anyone confirm this?

Yes. Phase shift always refers to a given frequency. A complex signal can't be out of phase by 180° across the board. Phase error from time delay is proportional to frequency. For a given delay increasing the frequency increases the phase shift.

The only time phase and polarity are indistinguishable is with sine waves and some other simple waveforms (triangle, square etc.).

 

[fat_fleet]

Yes. Phase shift always refers to a given frequency. A complex signal can't be out of phase by 180° across the board. Phase error from time delay is proportional to frequency. For a given delay increasing the frequency increases the phase shift.

Ok, yeah that makes sense. So if there was a switch on the bass amp for speaker polarity, switching it would probably give me the same result as hitting the polarity inverter on one bass track?

Now zoom in like crazy on the transient of the two recorded wave forms and see if they are out of alignment with each other, or of opposite polarity.

You know, I've done this before to check for phase issues and, by the time I zoom in close enough to where a difference should be visible, the two wave forms look so different that it's tough for me to tell if they are aligned. What looked like a peak before zooming actually becomes a complex series of peaks and troughs that looks radically different from its counterpart on a different track. I usually just use the logic "If it sounds ok it probably is". Maybe my ears aren't developed enough to hear phase discrepancy, but maybe.... just maybe I'm that rare recording hobbyist that's never had a problem with it!

 

[Steenamaroo]

Ok, yeah that makes sense. So if there was a switch on the bass amp for speaker polarity, switching it would probably give me the same result as hitting the polarity inverter on one bass track?

Yehp. Flipping at any point in the chain will give the same result.
Swapping the speaker wiring, flipping it at the amp, at the preamp, or in the software.; It doesn't matter.

Maybe my ears aren't developed enough to hear phase discrepancy, but maybe.... just maybe I'm that rare recording hobbyist that's never had a problem with it!

Maybe! lol. Go with your ears.

 

[bouldersoundguy]

You know, I've done this before to check for phase issues and, by the time I zoom in close enough to where a difference should be visible, the two wave forms look so different that it's tough for me to tell if they are aligned. What looked like a peak before zooming actually becomes a complex series of peaks and troughs that looks radically different from its counterpart on a different track.

There's a whole lot more going on besides polarity and simple delay. Filters, which can include anything from eq to speakers, do things to phase and polarity that is much more complex than what is caused by delay. You can never perfectly line up at all frequencies two versions of a signal that have been differently processed by the most common types of filters.

I generally go through and find a particularly large initial transient and use that to get me close then fine tune it by ear.