A select few (ie, not many turned up) listened to an interesting talk delivered by Pierre Watts
(I shall assume that any inaccuracies will be remedied by whoever was there, including Pierre himself)Pierre explained the concept, very popular with suppliers wishing to distribute audio via an Ethernet link, where errors could develop along the signal chain, with the signal going through multiple switches, being subject to interference from other network traffic. Various attempts have been made to address this problem, one of the successful ones being to add a time signal to the data stream that could be read at the destination to ensure the signal was recompiled in an accurate fashion
Ethernet is extremely popular since the basic infrastructure is readily available, the hardware is easy to set up and it is inexpensive. Ethernet was not originally designed to carry audio signals and these work-arounds are needed to correct these deficiencies
Pierre then touched on the design of Class D amplification, which, he said, only became a meaningful alternative to traditional Class A and Class A-B after about 2000 when improvements in filter design were able to remove the inherent problems of this design. A lot of expensive research has been done by manufacturers like Devialet who are supported by the French Govt to improve Class D. One huge benefit of Class D is that the amplifier will be above 85% efficient, thereby not wasting resources (a better "green" concept)
After this interesting but intense scientific discourse, we retired to the dining room and partook of dohnuts, samoosas, some red wine and managed to finish off a bottle of Old Brown Sherry.
After all, its winter in the Cape and you gotta keep yourself warm
