Skip to content

:warning: This file is part of the "archive" – it could contain statements or terms which are incompatible with the main body of the specification and the identity and timing model. :warning:

1. Live Synchronisation

Aim: Synchronise TDS 1 with TDS 2

Live Synchronisation

Scenario

  • Two microphones are plugged into an audio mixer
  • The Time Values produced by the microphones relate to the time at which audio hit their transducers
  • The microphones share a timing reference (for example via PTP) and as such can produce TDSs which use Time Context A.
  • When TDS 1 and TDS 2 arrive at the mixer, it can compare their Time Values in order to mix Data Objects (ie audio samples) which were coincident at the capture point of the microphones (*).

Behaviour & Options

The creation of Time Values for TDS 3 could be carried out using multiple methods:

  1. If the mixer does not share a timing reference with the microphones it may create new Time Values at its output using a different Time Context, with no relation to the incoming TDSs. In this case the mixer produces TDSs in Time Context B.

  2. The mixer may share a timing reference with the microphones but choose to create fresh Time Values at its output. The relationship between TDS 3 and the incoming TDS 1 and TDS 2 could as a result be identified via a simple time offset. In this case the mixer produces TDSs in Time Context A.

  3. The mixer may select an incoming TDS as a notional 'master' to take Time Values from and re-use at its output. This avoids the requirement for the mixer to share a timing reference with the rest of the system, but still allows it to maintain a relationship between the input TDS Time Values and the output. In this case there is a zero time offset between TDS 3 and the corresponding audio samples (Data Objects) in TDS 1 and TDS 2. In this case the mixer produces TDSs in Time Context A.

Caveats

  • In practice the audio mixer is likely to carry out more complex processing operations which may include timing offsets. The effect that this processing has on TDSs and their Time Values may be explored in a later example.
  • (*) Audio mixers do not typically automatically compare Time Values at their inputs and perform buffering to synchronise them. Whilst this is unlikely to change, it may be useful for future devices to indicate where mixes have been performed slightly out of sync due to the reported Time Values. This may allow the user to impose manual delays upon the various inputs prior to the mix operation.