Phase coherent mixing in SQ
Posted by Keith Johnson, Last modified by Keith Johnson on 30 August 2020 04:22 PM

Bus processing

On SQ, the bus processing latency is compensated so that no matter the internal routing, all signals are phase aligned on mix outputs.

For example:

  • Input to Group to Mains is aligned with Input to Mains
  • Group to Aux to Matrix is aligned with Group to Matrix
  • Input to LR to Matrix is aligned with Input to Aux to Matrix (With Aux Channel Source set to 'Post Delay')

In the analogue days this would have been irrelevant as processing latency was basically zero, but with most digital consoles, routing paths as the ones above will result in comb filtering producing cancellation at certain frequencies.

In the SQ, there are two alignments - the LR mix, Groups, Auxes and the Matrices.
Input direct outputs and tie lines have a slightly lower latency which depends on sources and tap off points, whilst routing through matrices will add 12 samples (125us).

The compensation allows for the signals on the mix bus to be always phase aligned down to the sample - even in cases of parallel processing - resulting in a clear and accurate sum. And all at under 0.7ms of overall system latency from analogue in to analogue out.

Channel, insert and FX processing

All channel processing in SQ is also compensated i.e. adding channel processing including DEEP preamp, compressor and GEQ models or switching AMM in/out will not add any extra latency or affect coherency.

External inserts, and inserted FX (whether active or bypassed) are NOT delay compensated. For example, inserting a De-esser or DynEQ from the FX Rack adds 24 samples on input / bus processing (250us). In some critical applications, you might want to manually compensate for this latency by adding channel delay to correlated, unprocessed channels.