Re: Sibelius multi-microphone recording

["John P. Hill" <jphill@frank.mtsu.edu>]

Greetings, all...

Bruce makes some interesting points about comb-filtering and multimic
techniques employed in GG's Sibelius recording.  I *promise* to take
further technical debate to private e-mail, but I'd like to address
some issues that I think are either incorrect or oversimplified.

>   "In a recording situation, the microphones don't often pick up 
relections, as well as usually being in a non-reflective environment".

Not true on *both* counts.  Even the best directional mics pick up
sound off-axis; the more expensive ones do it in a more linear fashion. 
In classical music recording, omnidirectional mics are often preferred,
and (by definition) these pick up sound with a 360 degree polar pattern
(ie: ambient room reflections and reverb are *not* rejected).  Most of the
best-sounding orchestral recordings (try Telarc or any film score recorded
by Shawn Murphy) feature these mics in the main array.

Classical music is almost *never* recorded in non-reflective environments!
Even GG's recordings, which were renowned for their close mic perspective,
contain significant amounts of room ambience and/or reverberation.  The
choice of recording venue is *critical* in classical music sessions,
because the whole idea is to be able to integrate the sound of the hall
(or room) with the direct sound of the instruments involved.  Pop music
recording, on the other hand, generally starts from a dry, close
perspective and utilises large amounts of outboard processing to add
various kinds of ambience or reverberation.  Although Gould prefered
a closer-than-usual perspective for both voice and piano (he loved this
aspect of Streisands's recordings) his recordings do contain real room
ambience/reverb.

Another aspect of classical recording technique is that mics are
generally placed further away from the instruments than in pop music
sessions.  If you have a look at the ON THE RECORD video from 1960,
you'll see a set-up employed at the 30th St. Columbia Studios typical
of early GG recordings.  The mics are Neumann M49s and M50s (omnis)
and they are far enough back from the Steinway that they are capturing
*lots* of room ambience along with the direct sound.

Comb-filtering *can* be a problem with spaced mic set-ups when these
are combined to mono.  Again, check out the various issues of the orignal
'55 Goldbergs for a useful example.  But in stereo, a certain amount of
uncorrelated diffuse information *needs* to be in the recording;
otherwise, listeners tend to perceive the sonic quality as "sterile" and
lacking a sense of "ambience" or "spatial envelopment".  When we get to
8-channel digital surround on DVD, we will be able to get some *really*
impressive results with ambience retrieval.

Incidentally, flangers employ LFOs (low frequency oscillators) to
constantly move the comb peaks and nulls across the frequency domain, so
that's another situation altogether.

Gould's use of the shifting pair perspectives is different, because
the intended playback system was *stereo* and the movement of the spatial
perspective (essentially the dry-to-reverberant balance) was linked to
structural aspects of the compostion being recorded.

Just my $.02 worth...


John Hill
Dept. of Recording Industry
Middle Tennessee State University
Murfreesboro, TN  37132