Tnx Mike!
On 5/15/2019 11:36 AM, Black Michael via wsjt-devel wrote:
I'm using an ANAN 100 and Thetis.
One thing I have to do to avoid audio dropouts under CPU load is
increase the buffer latency which affects the dt values.
I think I did tell you guys that I used a modified version of WSJT-X to
subtract 300ms from the audio data and I was able to decode things that
were not decodable without that offset. Simply people with > 0.5
seconds of dt.
Data from the test attached. The beginning of the file (30 minues or
so) should be "normal"...later on in the test I started fudging the time
as I recall to see it's affect so you may see some more odd-ball numbers
in the last 30 minutes as opposed to the first 30. I used my TimeFudge
program to tweak the time to reduce the dt value but didn't take notes
on when I started that.
It appears I started messing with my computer time around then 2100th
sample. I'm pretty sure I didn't message with time during the first
half of the test. I can probably find another Apache owner who
participate if you need more stable data.
Inline image
You can see the effect the buffer latency has on dt values on
hamspots.net. I had to increase my latency this morning as I was doing
some dropout testing...higher than what it was during the FT4 test.
This is closer to the default settings in Thetis of 120ms. I would
imagine I'd have a lot of trouble on FT4 now.
Inline image
Mike
On Wednesday, May 15, 2019, 9:50:00 AM CDT, Joe Taylor
<[email protected]> wrote:
Hi Mike,
I'm looking at some timing issues that are critical for FT4. You're
using a Flex (and I suppose, PowerSDR/Thetis), right? Could you please
send me an extract from your ALL.TXT file for the recent FT4 practice
session? That's May 8, 0000 - 0100 UTC (190509_000000 through
190509_010000). Many thanks!
-- Joe
_______________________________________________
wsjt-devel mailing list
[email protected] <mailto:[email protected]>
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
