MAKING CASSETTES SOUND BETTER
DOLBY B-TYPE NOISE REDUCTION
Dolby B-type noise reduction was developed in 1968 to address the hiss
of slow-speed consumer tape-recording formats, such as 3-3/4 inches per
second open-reel, compact cassette, and, later, the stereo linear tracks
on VHS video cassette tape.
Magnetic tape can only hold so much signal. If you put too much signal
onto the tape, it will overload or "saturate." The louder the signal
that is being recorded, the closer the tape becomes to being
"saturated." If you were recording loud (high level) signals and
at the same time you boosted them significantly for the purpose of noise
reduction, the tape would over-load and the recording would become
distorted and harsh sounding. Making matters worse for today's high-tech
music styles, the bass and treble extremes of the audible spectrum of
magnetic tape have even less capacity than the midrange.
"If it isn't broke, don't fix it"
It is very important that the use of noise reduction not take away from
the ability to record loud sounds by unnecessarily boosting them. Loud
sounds already hide noise on their own, so any increase would only mean
the overall recording level would have to be turned down to prevent
distortion. All Dolby Noise Reduction systems apply the principle of
"least treatment" in their design. This means that the loud signals are
not processed, so there is no chance of causing tape overload when Noise
Reduction is being used.
Least treatment essentially means that if there is no benefit to be
gained by changing the audio signal, then don't change it. This is the
reason that the compression-expansion process is only used on softer
(low level) signals (see Figure 2). For example, suppose you want to
record a solo, loud rap on a bass drum. The main part of the drum's
sound is at low frequencies and may already approach the tape's overload
point. Even so, the drum is not able to hide the hiss noise in the
treble range. The Dolby B-type noise reduction encoder does not provide
a boost to the drum signal, so the tape will not over-load (saturate).
However, it does provide a full boost over the rest of the frequency
spectrum. During playback through a Dolby B-type decoder,there is
effective noise reduction and no tape over-loading. Dolby B-type Noise
Reduction does this by using a SINGLE, SLIDING COMPRESSION-EXPANSION
BAND of frequencies. A sliding band is basically a filter that can shift
its frequency breakpoint up and down. In B-type noise reduction, the
frequency where the noise reduction action just begins can change from
about 300 Hz all the way out to 20,000 Hz.
When the audio signal is very low or contains no treble frequencies, the
band slides to the lowest frequency point (all the way to the left in
Figure 3), giving a maximum of 10 dB noise reduction above 4,000 Hz. As
the band slides to higher frequencies, less and less of the spectrum is
covered, so there is less effect. The band will not slide up, however,
until the sounds being recorded are loud enough to hide the noise on
their own. By moving the sliding band quickly enough to follow the
music being played, the full effect of the noise reduction is
maintained. In Figure 3, the sliding band slides up (in frequency) out
of the way of the loud bass drum signal (the "dominant frequency"), and
keeps the noise reduction working at higher frequencies, where tape hiss
would otherwise be audible.
In all Dolby Noise Reduction systems, the key to the operation is in the
positioning of the sliding bands. In order for the final decoded sound
to be faithfully restored, the bands in the decoder must track the
positions of the bands in the encoder as closely as possible. The
recorded audio signal tells the decoder how to operate. If, for some
reason, the level or the frequency response of the encoded signal is
changed before it reaches the decoder, mistracking of the sliding bands
will occur. How audible this becomes has to do with several factors,
including the nature of the music, the listening conditions, and the
sensitivity of the listener. By limiting its overall range to 10 dB,
B-type Noise Reduction is not very susceptible to audible mistracking.
Still, it is advisable to make sure that the tape formulations used with
any recorder are compatible with its design, and that the proper
settings are used for bias and record calibrations.
TECHNICALLY SPEAKING
In Figure 4 you can see that the compression expansion process is only
operating at middle-level signals (not real loud, or real soft
passages). At high signal levels (loud passages), there is no dynamic
action and the system acts as a unity-gain amplifier. At low
signal levels (soft passages) the system acts only as a fixed-gain
amplifier. By restricting the system's compression-expansion action to
only the middle-level signals, distortion caused by overshoots is
minimized. See Figure 5 for a Block Diagram of Dolby B-type Noise
Reduction.
Dolby B-type noise reduction is now the standard in tens of millions of
cassette decks,mini-component systems, boom boxes, personal portables,
and car stereo components; it provides 10 dB of noise reduction above
about 4,000 Hz.
Copyright © 1996 Dolby Laboratories, Inc.
--- QScan/PCB v1.18b / 01-0249
---------------