
Dolby Digital Audio Decoder
11-17
11.5.2 Dolby Digital Decoding
The Dolby Digital decoding ow is shown in
Figure 11.5 and described
in the steps following the gure.
Step 1.
Synchronization and Error Detection
The Dolby Digital bitstream format allows rapid synchronization
by the decoder. The 16-bit sync word in the bitstream has a
2.5% probability of false detection.
When a synchronization pattern is detected, the decoder
assumes that it is in sync and starts decoding the audio blocks.
Since the CRC1 word is read at the beginning of the frame and
it covers the rst 5/8 of the frame, the result of a CRC1 check
may be available after only 5/8 of the frame has been received.
The decoder can check for CRC2 once the entire frame is read.
If a CRC1 or CRC2 mismatch is detected, the decoder resets
itself and starts to search for the next synchronization word. A
CRC error sets the Audio CRC Error or Illegal Bit Error Interrupt
bit in Register 4 (
page 4-9), asserts INTRn to the host if the
interrupt is not masked, and mutes the audio output.
Step 2.
BSI Field Unpacking
Some of the BSI data may be copied from the Audio ES
Channel Buffer to dedicated registers for the host, some may
be copied to specic working memory locations, and some may
simply be left in the Audio ES Channel Buffer with pointers to
them saved in another location for use when the information is
required.
Step 3.
Exponent Decoding
The audio data in the Dolby Digital bitstream is the compressed
frequency coefcients. These coefcients are represented in
the form of an exponent and a mantissa. The exponent, which
could range from 0 to 24 bits, indicates the number of leading
zeroes in the binary representation of a frequency coefcient.
The exponent must be decoded rst to collect enough
information to decode the mantissa. In the Dolby Digital
bitstream, there are exponents for all independent channels, all
coupled channels, and for the coupling and low frequency
effects channels.