On 20/02/2020 22:01, Martin Davies G0HDB wrote:
I'm running WSJT-X v2.1.2 0068f9 on a PC with Win 7 (Pro).
Back in late January I posted a message, #5852, on ws...@groups.io
about a 15-second discrepancy I'd observed between the times of FT8
decodes displayed in the Band activity window and the times recorded
in the ALL.TXT file. The full path to my message is:
https://groups.io/g/WSJTX/message/5852
Sadly I never received a response to my message so I haven't yet
received any explanation of why the displayed and recorded times of
decodes might be different.
I've recently had reason to spend a bit of time doing some analysis of
some of the decodes recorded in my ALL.TXT file and have come across
another oddity. My latest finding is that my ALL.TXT file shows a
list of 'Rx' decodes and a single 'Tx' entry for my callsign all
having the same recorded time; surely this shouldn't happen because I
can't both receive and transmit within a single 15-second period?
Here's an extract from my ALL.TXT file showing what I've described above:
200220_001130 5.357 Rx FT8 -1 0.1 1863 F6EAZ SP2EWQ RR73
200220_001130 5.357 Rx FT8 -15 0.1 2101 CQ KS4OT EM83
200220_001130 5.357 Rx FT8 -7 0.1 2299 CQ A45XR LL93
200220_001145 5.357 Tx FT8 0 0.0 735 WB5XX G0HDB R-19
200220_001200 5.357 Rx FT8 -19 0.2 612 E77D EA5IVR IM97
200220_001200 5.357 Rx FT8 -11 -0.1 664 CQ DX K4EA EM74
200220_001200 5.357 Rx FT8 -16 0.1 855 AE4WG FM1HN RR73
200220_001200 5.357 Rx FT8 -12 0.3 1033 OO7B W4JSI R-22
200220_001200 5.357 Rx FT8 -4 0.4 1501 CO8LY IZ5MOQ JN54
200220_001200 5.357 Rx FT8 0 0.1 1863 F6EAZ SP2EWQ RR73
200220_001215 5.357 Rx FT8 -18 0.1 2100 CQ KS4OT EM83
200220_001215 5.357 Rx FT8 -9 0.2 2299 CQ A45XR LL93
200220_001215 5.357 Tx FT8 0 0.0 735 WB5XX G0HDB R-16
200220_001230 5.357 Rx FT8 -11 -0.1 664 EA5CMW K4EA -14
200220_001230 5.357 Rx FT8 -15 0.1 854 OZ4VV FM1HN R-07
200220_001230 5.357 Rx FT8 -14 0.3 1034 OO7B W4JSI 73
200220_001230 5.357 Rx FT8 -4 0.4 1501 CO8LY IZ5MOQ JN54
200220_001245 5.357 Rx FT8 -1 0.1 1863 F6EAZ SP2EWQ RR73
I've added the colouring to highlight my transmissions.
As can be seen, my two transmissions to WB5XX are separated by 30
seconds as would be expected, but the two 'Rx' entries for KS4OT's and
A45XR's CQs immediately preceding my second transmission at
200220_001215 are shown as having the same time as my transmission
whereas I suspect they should have been recorded with the time of the
preceding 15-second period, ie 200220_0011200, which would have been
the expected 30 seconds after their previous CQ transmissions.
Also, the final line in the above extract of ALL.TXT, ie. the RR73
message sent from SP2EWQ to F6EAZ, should presumably have had a
recorded time of 200220_001230 and not the time of 200220_001245 that
has been written into the ALL.TXT file.
There are other examples of this inconsistency between the recorded
times for Rx decodes and outgoing transmissions further down in my
ALL.TXT file for earlier today (Thurs 2020-Feb-20); there may be many
other similar discrepancies but I haven't delved any further into my
ALL.TXT file.
It seems to me that for some reason the times for some Rx decodes
within a 15-second period are being written incorrectly into the
ALL.TXT file and are being recorded as having the time of the
following 15-second period, despite the times shown in the Band
activity window potentially being correct (as I'd reported in January).
I will be grateful if the development team could take a look at this
issue, which is making any 'forensic' analysis I or others might wish
or need to do of the FT8 decodes and transmissions recorded in the
ALL.TXT file all but impossible to do with any sort of accuracy or
rigour. The presence of the discrepancies and inaccuracies in the
times recorded in the ALL.TXT file is a considerable inconvenience and
surely shouldn't be happening...!
--
73, Martin G0HDB
Hi Martin,
I suspect the ALL.TXT rows you are referring to were decoded after the
following period started. The algorithm for allocating to the Rx period
is simplistic and probably unchanged since times when changeover periods
were sufficient to complete all decoding before the end of the period.
With FT8 and probably more so with FT4 on busy bands that is no longer
the case.
73
Bill
G4WJS.
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
