Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!security!genrad!wjh12!n44a!ima!inmet!muller
From: muller@inmet.UUCP
Newsgroups: net.audio
Subject: Re: Orphaned Response - (nf)
Message-ID: <795@inmet.UUCP>
Date: Fri, 27-Jan-84 05:48:42 EST
Article-I.D.: inmet.795
Posted: Fri Jan 27 05:48:42 1984
Date-Received: Tue, 31-Jan-84 03:24:06 EST
Lines: 91

#R:decwrl:-490100:inmet:2600031:177600:5972
inmet!muller    Jan 26 16:45:00 1984

There are several possible reasons for your experience, all related to
sound reinforcement phemonena, not signal processing.  Without more info
tho', we can't say zactly what caused it.  You said there was a monitor at
each end of the stage...were they really monitors, i.e., speakers for the
performer's benefit, or were they actual reinforcement speakers for the
audience?  (I suspect the latter.)  This might make a difference...
   The first possibility is that you were hearing feedback effects.  Yes,
feedback can produce significant frequency coloration well below the
self-sustain level, e.g., +3 to -6 (?) dB at 3 dB below self-sustain.  
(We can call this amplitude response rather than frequency response, if
you prefer, tho' some relative phase shifts also occur.  These numbers
were recalled from my fuzzy memory of having actually plugged munbers
into the equations once...and we don't know how close the engineers were
actually running it anyway.)  These levels are well within your and
others' discernibility.  This is caused by the interference between the
coherent sounds FROM 4 sources (2 speakers, 1 piano, 1 voice), and INTO
2 mics.  As a recording engineer, you probably have never run into the
FROM problem, but may have had to fight the INTO problem.  I would guess
that the vox mic was mixed hotter than the piano mic, and thus was the
real offender.  This problem would be more likely if the speakers on the
stage were true monitors.  Other contributing factors include the off-
axis rejection characteristics of the mics and the angle at which they
were pointing.  Unfortunately there are many so-called engineers doing
reinforcement who don't understand wave phenomena at all.
   The second possibility involves your seating position.  The relative
distances to the speakers was probably such that you were hearing both
at similar amplitudes (assuming the speakers were pointed at you, not
at Mr. Winston) if you were near center.  Similar amplitudes is, of
course necessary for significant interference to occur.  If, then, one
speaker were perhaps just a few decimeters (give or take one order of
magnitude!) closer than the other, PRESTO! instant frequency-dependent
interference, again at a level you could most certainly discern.  The
summation of 2 equal coherent signals would give +6 dB, while the
cancellation of those frequencies out of phase could give -10, -100,
or -EVERYTHING! dB, if the amplitudes were similar enough.  Ironically,
those people seated to the sides probably had better sound than you did,
since they "heard" one speaker much more than the other.  There are 
several solutions to this: 1) use a hall with a center aisle, thus no
seats have this problem until you get further back from the stage (the
problem zone widens as you go back, but those folks won't feel like they
got ripped off by paying for a choice seat which did't sound so good!, 
2) run a stereo mix - this works for every sound source on the stage 
except those performers right at center stage, who will probably still
be panned equally into each channel, but it can accentuate the feeling
of being-off-to-the-side for those people with side seats (stereo mix
probably was not done at your concert, I would guess), or 3) use a
better speaker placement, i.e., only ONE (if you are mixing in mono),
placed above the performer, with a full spread angle (180 degrees or
whatever is necessary) - this also has the benefit of increasing the
throw of the coherent sound from the speaker into the reverberant field
of the room.  This last solution is rarely done partly becuz it isn't
practical with portable (for hire) sound systems, and partly because
engineers have a fixation with two speakers (ya' know, high-end stereo
stuff, and all), and sometimes becuz multi-speakers are needed for
audience coverage.  This last solution also has the benefit of allowing
some of the patrons to experience the proximity effect, making them
think they are hearing the performer directly rather than the PA system.
Unfortunately a lot of (even pro) sound reinforcement engineers don't
understand this either. 
   As you can see, the problems of SR'nfrcmnt are usually the result of
the actual sound fields, and typically dominate the smaller electronics
or signal processing concerns that SR'prdctn involves.  Worrying about
relative phase response between frequencies in an amplifier, or whatever
is misguided when you can have +/- 5 dB in amplitude response from the
sound field, lots of phase shifts in microphones, as much as 100%(!)
harmonic distortion in (admittedly the cheapest) PA speakers, poor
frequency/dispersion characteristics and nonconstant efficiencies 
(varying impedance) with volumn or frequency, etc., that comes from
optimizing a speaker for high efficiency/high power handling ability.
Couple this to the fact that anyone can buy PA stuff any then consider
that a rigorous quantitive analysis of any concert setting is not
practical (so that experience often counts for more than theory) and
you can see why there is variability in concert PA system performance.
I'm not trying to criticize either sound engineers or your own thoughts
on recording, rather just trying to put things into perspective.  Notice
that I haven't even touched on the question of PERCEPTION.
   Most of what I have said I can substantiate with experience.  I am
a semi-pro musician, semi-pro sound reinforcement engineer, and a
(professional) signal processor (underwater acoustics) and scientist
(seismology).  I HAVE probably spent more time thinking these problems
than most people in the SR'nfrcmnt business.
   Yeah, I know I misspelled voluum, and a few other words....
              Jim Muller
              8 Bates Rd East
              Watertown, MA 02172
Yes, interference between the direct and speaker sound is real possibility
though it is more likely that the requisit similarity of amplitudes 
for any frequency was obtained between the two speakers.