<<What you can do is to place the two AT3032 microphones on a short
bar, one pointed due left and one pointed due right. The separation
is determined by the length of the microphones. The spacing gives
some directional cues, but the non-obvious part is that the narrowing
polar pattern at high frequencies also gives you some directional
cues. An omnidirectional microphone of the size of an AT3032 is
typically down several dB for sounds 90 degrees off axis above 5 kHz
or so. I've tried this, and it works.>>
Yes, despite the dictionary meaning of the word "omnidirectional",
all omni mics become directional to some degree with increasing
frequency. Some become perfectly cardioid in the upper 2 octaves.
However, because of this, a 180 degree orientation is not optimal, as
it will not provide a solid & defined center image. So, although the
end to end array listed above will be tremendously physically
convenient, it's not a good sort of stereo, i.e. left & right will
not be coherent, as they would with a forward orientation of the
mics. A good way to determine how realistic your stereo field is, is
to face a road while a vehicle drives past from far left (or right)
to the far other side. On playback there should be an absolutely
linear sense of motion from one side through a well defined center to
the opposite side. If the car stays mostly in one speaker, then
abruptly jumps to the other side this may be dramatic but it's not a
realistic depiction of the stereo field. This is a problem I hear
with many of my older head-mounted (quasi-binaural) lavaliere
recordings, which is a similar set up to a Jecklin or boundary
separated omni array, & this dictates why the boundary can easily be
too big for good stereo. Omni mics may have more latitude for
effective spacing than directional mics, but the laws of physics
still apply.
Scott Fraser
|