Hi Steve,
Actually I can determine the problem occurred between SVN 4038 and 4089.
I do know, when changing USB sound device types, the audio levels are all
different. Also, as far as I know, in the FreeDV API there are no calls into
PortAudio.
The top level program hands audio to the FreeDV API which in turn hands back
audio,
un-modemed and un-codec2ed. (if there are such words)
On a "broken' system, occasionally it will do a full decode. So, was it a fluke
in timing?
So so much time is wasted with small coding mistakes. I caused one myself, just
the
simple adding an extra statement after an "else" and not enclosing the two with
{}.
And yes, I could do a "diff -r <dir1> <dir2>" but the actual programmer might
well remember
what he did at that time.
Actually, this is one reason I LOVE ARM SBC boards with a SATA disk port. All
my dev work
over months was on the disk, not the SD card which was over-written when I
changed from
a Banana Pi (original) to a Banana Pi M2 Berry after the lightning strike.
On another note, "C" not being my "bread and butter", I see "callbacks" are
used in
"freebeacon" to hand back the text data. and there is another callback in
freedv_api.c so
I may be able to get the Codec2 raw un-modemed data, through FEC, to write to a
file.
Is this worth pursuing?
Thanks Steve, I'm encouraged when the arduous work of a Systems Integrator is
noticed.
Alan VK2ZIW
On Wed, 14 Aug 2019 11:00:18 -0500, Steve wrote
> Here how to do a diff:
>
> svn diff -r 4010:4121 >diff.txt
>
> Then look through the diff.txt to locate software changes over the years.
>
> _______________________________________________
> Freetel-codec2 mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/freetel-codec2
---------------------------------------------------
Alan Beard
OpenWebMail 2.53
_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
