Old 23rd June 2013
  #1
Lives for gear
 
therock's Avatar
 

Thread Starter
Phase

I'm curious, When tracking guitars with multiple mics.
At what ms. interval do you start to hear potential phase shift?
Old 23rd June 2013
  #2
Gear addict
 
MichalS's Avatar
 

Take guitar signal, duplicate it to another track put short delay on it feedback 0, mix 100% delay and you will hear every single millisecond, If you don't have delay compensation in your daw you will hear even the delay just being plugged. With microphones it's more complicated as the sound of particular mic kicks in, and sometimes the phase shift defines the sound itself. If you don't want phase shift problems 3:1 rule of mic placement is your friend.
Old 23rd June 2013
  #3
Lives for gear
 
therock's Avatar
 

Thread Starter
thanks, these mics are both at the same distance (almost)
So when I zoom in very very far I can see they are off a few samples.
So my question should be, how many samples before you can hear
a difference. I can simply fix the problem by moving one mic closer or whatever.
Or nudge one of the waves in the DAW

I'm just curious has any scientific study every been done at what sample interval humans can start to notice phase cancellation? You'll notice picture on the left
the track look aligned but when you zoom in really really close down to almost the final Sin wave you see it's off .001 of a second. Obviously this is never going to cause audible cancellation but at what point does it?
Attached Thumbnails
Phase-zoom3x.jpg   Phase-zoom4x.jpg  
Old 23rd June 2013
  #4
The Audio Whisperer
 
donsolo's Avatar
Phase cancellation isn't something the human ear can detect. In fact, the human ear can't hear delay that's less than 25-30ms depending on the person. 0-25ms is a all comb filtering from phase cancellation. Comb filtering happens whenever you have two separate mics. Some will only comb 1 or 2 frequencies, some will comb whole sets of frequencies (which you can calculate on your own)

The best way to figure out what you're losing is to invert the polarity/phase on that second track. If you can't hear anything after that, then you're not losing anything when you put them back together. If there's a lot of high frequencies, then nudge it around until you minimize what's being cancelled.

That's how to test for what is being filtered out. It becomes a game of "what sounds better" when flipping polarity/phase after following that 3:1 rule.
Quote
1
Old 23rd June 2013
  #5
Lives for gear
 
couch11's Avatar
 

Why are you trying to record two exact signals? The point of multimiking guitar cabs is for different tonal variations between the different microphones, the placement of those microphones, AND the comb filtering from using two microphones. Think of the comb filtering as EQ. Depending on the distance between the mics, you can boost or cut different frequencies to taste just by moving one mic in relation to the other.

By trying to have two exact signals perfectly in phase alignment, all you are doing is making your source louder.
Quote
1
Old 23rd June 2013
  #6
Lives for gear
 
therock's Avatar
 

Thread Starter
Quote:
Originally Posted by couch11 View Post
Why are you trying to record two exact signals? The point of multimiking guitar cabs is for different tonal variations between the different microphones, the placement of those microphones, AND the comb filtering from using two microphones. Think of the comb filtering as EQ. Depending on the distance between the mics, you can boost or cut different frequencies to taste just by moving one mic in relation to the other.

By trying to have two exact signals perfectly in phase alignment, all you are doing is making your source louder.
thanks

I use a dynamic in conjunction with a condenser, sometimes a ribbon and use that one the best sits in the mix.
Old 23rd June 2013
  #7
Lives for gear
 
therock's Avatar
 

Thread Starter
Quote:
Originally Posted by donsolo View Post
Phase cancellation isn't something the human ear can detect. In fact, the human ear can't hear delay that's less than 25-30ms depending on the person. 0-25ms is a all comb filtering from phase cancellation. Comb filtering happens whenever you have two separate mics. Some will only comb 1 or 2 frequencies, some will comb whole sets of frequencies (which you can calculate on your own)

The best way to figure out what you're losing is to invert the polarity/phase on that second track. If you can't hear anything after that, then you're not losing anything when you put them back together. If there's a lot of high frequencies, then nudge it around until you minimize what's being cancelled.

That's how to test for what is being filtered out. It becomes a game of "what sounds better" when flipping polarity/phase after following that 3:1 rule.
but if they are less than 30ms that's less than 1°, If I flip the polarity, now
they will be 179° out of phase. So I guess I should reword the question.

is less than 1° noticeable? can some people hear comb filtering at his interval? I can't hear it
Old 23rd June 2013
  #8
Lives for gear
 
Tinderwet's Avatar
Quote:
Originally Posted by therock View Post
can some people hear comb filtering at his interval? I can't hear it
Then it's good.
Old 23rd June 2013
  #9
Lives for gear
Quote:
Originally Posted by therock View Post
but if they are less than 30ms that's less than 1°, If I flip the polarity, now
they will be 179° out of phase. So I guess I should reword the question.

is less than 1° noticeable? can some people hear comb filtering at his interval? I can't hear it
You can't directly convert miliseconds to degrees, because phase shift is frequency dependent.
At 44.100 Hz sampling, you could theoretically hear phase shift caused by 1 sample delay, but you might not hear it on electric guitar, because filtering happens at a very high frequency, which is may not be produced by the amplifier.

You won't hear 1° phase shift, but 1° at 10 kHz is way less than 30 ms.
Old 23rd June 2013
  #10
Lives for gear
 
therock's Avatar
 

Thread Starter
Quote:
Originally Posted by jetam View Post
You can't directly convert miliseconds to degrees, because phase shift is frequency dependent.
At 44.100 Hz sampling, you could theoretically hear phase shift caused by 1 sample delay, but you might not hear it on electric guitar, because filtering happens at a very high frequency, which is may not be produced by the amplifier.

You won't hear 1° phase shift, but 1° at 10 kHz is way less than 30 ms.
so how do devices like Little Labs IBP work. Are they multiband and shift independently respective each to frequency?
Old 23rd June 2013
  #11
Lives for gear
Old 23rd June 2013
  #12
Lives for gear
 
therock's Avatar
 

Thread Starter
Topic:
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
 
Register FAQ Search Today's Posts Mark Forums Read

SEO by vBSEO ©2011, Crawlability, Inc.