On 05/12/2018 00:12, Al Pawlowski wrote:
I wonder if I should add in, or subtract, another 0.1s for the tx
delay I have set.
Hi Al,
the Tx delay does not change the timing of the signal. The Tx signal
synchronized with the PC clock at the start of a Tx period (interpolated
forwards if Tx starts late), from then on to the end of the Tx period
synthesized samples are delivered to the sound stream buffers on demand.
So the sound card clock (which may not even be in the PC e.g. a rig with
a built in audio codec or an SDR) does the timing after the start. The
digital latency is the sum of buffer length times sample period and
processing delays. Process priority should only have an effect on its
ability to keep up with the sound stream's need for samples. Not enough
priority might cause an under-run and consequent drop out, from then on
the rest of the samples will be delayed by the drop out length. We don't
correct for drop outs as we don't know about them and anyway a drop out
is far more disruptive to decoding than a small extra delay so it would
be too late even if knew it had happened.
73
Bill
G4WJS.
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
