dbx 122 - Page 29

28
HOW
dbx
Il
TAPE
NOISE
REDUCTION
WORKS
A
Short
Explanation
To reduce tape noise, the dbx 122 and 124 utilize a
sophisticated version of the classical compressor /expander
(compander) concept. The RECORD processor compresses
the input to the tape recorder by a 2:1 ratio, linear over a
100dB range. Upon playback, the PLAY processor provides
1:2 expansion of the recorder’s output. The expansion is a
mirror image of the compression So the recorded signal
cannot be distinguished from the original audio source,
full dynamics are preserved, and virtually no audible tape
noice is added.
Consider a 100dB program which might have loud peaks
at +18dBm and quiet passages as low as —82dBm. If the
tape recorder has a maximum input level of +15dBm before
distortion
and
a
residual
noise
level
of
—40dBm,
in
other
words, a 55dB S/N ratio which is typical of many good hi-fi
and semi-pro tape machines, there is no way the original
program dynamics can be captured on tape; peaks will be
lost due to saturation, or quiet passages will be lost in the
noise, or a combination of both. dbx encoding will solve
this problem by bringing the maximum signal level down
to +9dBm, and by raising the minimum level to —41dBm
so the recorded program would fall within the usable
dynamic range of the tape. (See Figure 19.) That is, the
quietest signal remains above the noise and the loudest
signal remains below the point of tape saturation.
NOTES:
1. Maximum input levels on tape recorders are specified
at the 3% harmonic distortion point. Because dbx encoding
lowers maximum recorded levels, it greatly lowers distortion
at the same time it eliminates audible noise.
2. The term “‘dB SPL” refers to the acoustic sound
pressure level. The term “dBm” refers to the level of
“sound”
while
it
is
in
electrical
form.
dB
is
a
relative
term,
and
there
is
no
direct
equivalence
of
dB
SPL
to
dBm,