Hi David,
You had lots of good ideas, so I started digging into the problem a bit
more.
I tried several different machines and sound devices of different
types. The sending machine played a WAV file created by freedv_tx on an
x86 machine, and the receiving machine was running freedv 1.3 on x86.
The codec I used for the first pass was the 3689 revision.
I only found one combination of hardware that worked 100%. Strangely
enough, that combination used a RasPi playing the WAV file through its
"native" sound device, which is well-known for being a minimal hardware
hack. The other combinations using real sound devices all had the
creeping errors issue I described earlier.
I never saw the sawtooth drift symptom you described. The timing line
drifted smoothly and slowly between ~ -0.1 and +0.1. The frequency line
was flat, even during bursts of bit-errors.
I also didn't see the PortAudio{1|2} counters move during any tests,
including the failures. The mailing list won't let me send pictures,
but I'll send you a screen capture out-of-list. The counters behaved
the same with both failures and successes.
I'm rebuilding with 3888 to see if there are any differences, but
downloading the tree has been taking a while.
Still tinkering...
73,
Matt
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
Freetel-codec2@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
