| 摘要 |
In a system where multiple inputs are distributed to multiple destinations and vice versa, e.g. on a trading floor or in a call centre, audio data is recorded for archiving in such a way as to improve the viability of the records and to avoid recording redundant information. Sources 110 may be telephone calls destined for a single trader at a workstation or news channels broadcast to a plurality of workstations. The incoming data streams are associated with both source and destination identification fields. The data is supplied over a network 130 to one of several severs 120, each of which communicates with several workstations. The servers include distribution logic to connect the source to the appropriate workstation or workstations. Each data stream, before distribution is tapped and sent to a storage system 140, where it is recorded along with the identifying fields. Response audio from the workstation operators is also recorded. Requests to playback stored audio may be made from the workstations, by using the source and destination identification fields to access the storage memory. |
