I repeated these tests using the fastest possible autoscan interval + the fastest possible scan settings, per your suggestion, and now the latency between the TTL trigger from the EVT pin (going directly into the EEG system) and the TTL trigger from the Trill via the Bela is ~1.8ms, so implementing the regular (fastest possible) scan interval does appear to help overall latencies.
Okay, so, to confirm, this latency (between the TTL trigger from the EVT pin and the TTL trigger from the Trill via the Bela) can be assumed to reflect A+B+C+D in your message above?
And therefore to ascertain the full latency between contact on the Trill and when the TTL message from the Trill via the Bela, the only 2 additional pieces of information needed are:
1) The Trill scan time (which is known, i.e. 1.71 ms with the fastest scan settings/lowest resolution)
2) The time between making a contact on the Trill and the initiation of a scan? Is this time known?
Is this correct?
Thanks again, and hopefully this is the final question here.