frank w.scott wrote on Sun, 06 November 2005 16:31 |
- Full pan left and right: Completely out of phase, jumps between -1 and 0. Extreme wide stereo (false sounding).
- Pan left 50%, right 50% Jumps between 0 and +1, in phase although not completely.
- Pan left 50%, right 0% Completely in phase, stays solidly at +1
- Pan left 20%, right 20% Completely in phase, stays solidly at +1
So, anyway it seems I can have my cake and eat it too.
|
No one else seems to have pointed out the fallacy in this experiment. You seem to be missing the fundamental science behind what is going on here.
In your first "completely out of phase" test with the hard panning, you are seeing the actual signals' phase relationship. Almost completely out of phase. This is generally not a good thing unless a phasey special-effect is desired.
As you are panning them closer towards the center you are NOT removing signal from left and right, you are simply adding more of the left signal to the right and vice-versa.
Big surprise, the signal becomes more and more in-phase as left and right become more and more the same. If you used two fully out-of-phase sine waves you would achieve the same result. Complete nulling of the signal, but no out-of-phase content at all when both are panned center. Maybe a better test is 1k on each side and 2k on one side only. As you pan inwards, you'll be left with only 2k signal if the 1k components were at an equal level. Now imagine that at integral-multiple relatitionships to the mic distances up and down the frequency spectrum...
The fact is, there is massive comb-filtering happening if the mics are in front of the same instrument and one mic is polarity flipped. The fact that this sounds ok to you notwithstanding, the conclusion that the phase error doesn't happen as you pan inwards is incorrect.
The 20% pans and 0%/20% are only revealing the limited resolution of the phase analysis.
If you want a more natural sound, this would be the dead last thing you should do. The sound will change based on panning, but also based on speaker distances, listener's rooms, your head position, etc. after the recording is a done deal. This is why keeping a modestly accurate phase relationship using multiple mics is a goal for better sound.
If the proximity effect is bothering you, perhaps try omni mics? Intentionally comb-filtering with left and right pairs sounds dangerous to me as well.
Chris