Audio Engineering

(Barry) #1
Data Compression 587

of the scale factors are suffi ciently close or when the encoder anticipates that temporal
noise masking will hide the consequent distortion. The layer 2 algorithm also improves
performance over layer 1 by representing the bit allocation, the scale factor values, and the
quantized samples with a more effi cient code. Layer 2 coding also added 5.1 multichannel
capability. This was done in a scaleable way so as to be compatible with layer 1 audio.


MPEG layers 1 and 2 contain a number of engineering compromises. The most severe
concerns the 32 constant-width subbands which do not refl ect accurately the equivalent
fi lters in the human hearing system (the critical bands). Specifi cally, the bandwidth is
too wide for the lower frequencies so the number of quantizer bits cannot be specifi cally
tuned for the noise sensitivity within each critical band. Furthermore, the fi lters have
insuffi cient Q so that signal at a single frequency can affect two adjacent fi lter bank
outputs. Another limitation concerns the time frequency–time domain transformations
achieved with the wave fi lter. These are not transparent so, even without quantization, the
inverse transformation would not perfectly recover the original input signal.


19.6 MPEG Layer 3 ......................................................................................................


The layer 3 algorithm is a much more refi ned approach. Layer 3 is fi nding its application
on the Internet where the ability to compress audio fi les by a large factor is important in
download times. In layer 3, time to frequency mapping is performed by a hybrid fi lter
bank composed of the 512-tap polyphase quadrature mirror fi lter (used in layers 1 and
2) followed by an 18-point modifi ed cosine transform fi lter. This produces a signal in
576 bands (or 192 bands during a transient). Masking is computed using a 1024-point
FFT: once again more refi ned than the 512-point FFT used in layers 1 and 2. This extra
complexity accounts for the increased coding gains achieved with layer 3, but increases
the time delay of the coding process considerably. Of course, this is of no account at all
when the result is an encoded .mp3 fi le.


19.6.1 Dolby AC-3


The analogy between data compression systems and noise reduction has already been
drawn. It should therefore come as no surprise that one of the leading players in audio
data compression should be Dolby, with that company’s unrivaled track record in noise
reduction systems for analogue magnetic tape. Dolby AC-3 is the adopted coding
standard for terrestrial digital television in the United States; however, it was actually

Free download pdf