DV audio processing
As for the audio channel, the user can select either 16 or 12-bit sound mode. The former containing high-quality stereo signals, while the latter offers four separate channels suitable for dubbing. The sound data is sam- pled and packetised, stored and released to the modulator in time with the TDM switch changeover, itself synchronised by the head-drum PG/tacho pulse. The bit rate for the audio bitstream is fixed at 1.25 Mbps and that for the control/housekeeping at 8.716 Mbps. Given that the video bit rate is 24.146 Mbps, the total bitrate for the DV recording is 35.382 Mbps (normally referred to as 36 Mbps).
Signal processing: replay
The first stages in DV replay are signal amplification and equalisation, the latter not to give best bandwidth and noise performance as in analogue tape systems, but to optimise the demodulator’s discrimination between 0 and 1 symbols. Thus furbished, the off-tape signals are gated by the TDM switch to the audio or video stages in synchronism with their readout. The video data is checked and – where possible – errors are repaired in the ECC stage; any data beyond repair is deleted and replaced by good data from the same position in the previous scanning line, a process akin to the action of a dropout compensator in an analogue VCR.
Now the data is shuffled once more, using the same look-up table as was used during the record phase, for its passage to the inverse DCT (I-DCT) where data decompression is carried out. After de-shuffling the picture is built up, frame-by-frame and stored in a SRAM memory chip: it is read from here into the D-A converter whose output provides the recon- stituted video signal. The analogue audio signal is reconstituted likewise, using the inverse processes to those applied during the record phase. Both video and audio can be conveyed from the machine in digital form (IEE 1392/Firewire/i-Link) into other digital equipment (see Chapter 25).