Sound System Design 128534.4.3.3 Mix-Minus and Matrix Mixing
The automatic mixer in Fig. 34-40 includes a matrix
output. This enables a mix-minus system for conference
rooms or other multimicrophone systems, Fig. 34-41.
Each talker at this conference table has his or her own
microphone and loudspeaker. The signal from a given
talker’s microphone is amplified to a greater degree in
loudspeakers that are farther away from the talker. This
is a natural way to make the talker’s voice heard well at
any point around the conference table. In addition, this
system helps control feedback because the talker’s
voice is not amplified into his or her local loudspeaker
and only slightly into nearby loudspeakers. In combina-
tion with normal automatic mixer functions, the
mix-minus approach can make effective sound rein-
forcement possible in a large conference room. This
approach is also valuable for audio or video teleconfer-
encing systems.
A mix-minus system is complex but the calculations
are simple if taken one microphone at a time. Consider a
single talker and microphone. Assume that the listener
seated next to this talker can hear the talker unaided by
the sound system. Now, use inverse-square law, Eq.
34-1, to calculate the loss from this nearby listener to
the farthest listener. Amplify the signal to this farthest
listener (using the loudspeaker nearest to that listener)
to make up for the loss. Do the same for each remaining
listener. Now, repeat the process for the second talker
and so on. A spreadsheet is a useful tool for keeping
track of these calculations and the required settings in
the matrix.34.4.3.4 Problems in Automatic MixingDespite sophisticated circuitry, ambient noise may still
turn a microphone channel on at the wrong time.
Another problem is coherent input signals, that is, sig-
nals that are in-phase and have similar waveshape,
which may fool the mixer and allow it to raise its gain to
a feedback condition. Nearly coherent signals may
arrive at the microphones from a slammed door, for
example.
An obvious problem with all automatic mixers is that
they do not know when a new talker approaches the
microphone. Thus, the mixer cannot readjust a micro-
phone level for a loud-versus-quiet-voiced talker. A
compressor or AGC (automatic gain control) circuit
could be added to the mixer to adjust the level to
compensate but this would defeat the number of open
microphones (NOM) function and could cause the
system to go into feedback. An experienced human
operator is the best solution to this problem.
Despite their problems, automatic mixers are
extremely useful, and a well-designed system with an
automatic mixer is more likely than ever before to be
audibly transparent to an audience.34.4.4 Signal-Processing ComponentsMost signal-processing functions are now performed by
software modules in a DSP device. However, it is still
possible to purchase individual signal-processing
devices and the functions are the same whether they are
performed by a separate hardware device or a software
module in a DSP. For reasons of clarity, this section
considers separate hardware signal-processing compo-
nents. However, as discussed, the functions can be per-
formed equally well by multifunction DSP devices.Figure 34-40. An automatic mixer with matrix output and
DSP. Courtesy Lectrosonics.
Figure 34-41. A simplified mix-minus system.
Ceiling loudspeakers
0 dB^ 3 dB^ 6 dB^ 9 dB^ 12 dB^ 18 dB mutedTalkerMicrophone