Reino,
Many thanks. I like to have all the background. So I cleared the concern
with a fresh TX5S QSO last night. FT8 blasted through all the CW
contest. I still got the missing decode in the RX Frequency window, though.
--73, Glenn, AF8C
On 1/27/2024 1:29 AM, Reino Talarmo via wsjt-devel wrote:
Yes, that improves decoding sensitivity quite a lot and
is useful once you have started a QSO. You may study the
a priori principle in the user guide 12.1. AP Decoding
and in more details in the reference 33,
https://wsjt.sourceforge.io/FT4_FT8_QEX.pdf, especially
Figure 7, Table 6 and Figure 8.
Of course you may not use it in FT8, just disable in
Decode menu the Enable AP.
On the other hand also without the AP you still can get
false decodes as the message error detection is not 100
% reliable. There are in the message 77 information bits
and 14 error detection bits and that provides good
enough message error detection after forward error
correction. Perhaps FT8 is too good compared to RTTY and
CW and operators rely on what's written on screen, hi!
I have some 24/7 measurements over two years and I have
collected more than 80 000 callsigns and there are e.g.
108 callsigns starting with '0', 193 callsigns starting
with '1', the latter contains some real ones. That
continues to all figures and numbers, but it is harder
to tell fake ones; say order of 4000 total e.g. 5%. We
should remember that the fake ones are received only
once almost for certain and the real ones 100's of
times. Most of those false decodes are not due to AP,
perhaps half of the recording time I have not used AP as
it takes more processor resources.
It is an operator's fun to use his brains and detect
false decodes although those may be with red background!
73, Reino OH3mA
-----Original Message-----
From: Glenn Williams via wsjt-devel [mailto:wsjt-
de...@lists.sourceforge.net]
Sent: Saturday, January 27, 2024 1:12 AM
To: wsjt-devel@lists.sourceforge.net
Cc: Glenn Williams <a...@alumni.caltech.edu>
Subject: Re: [wsjt-devel] wsjt-devel Digest, Vol 119,
Issue
45
So,
Do you mean that a priori information is used to help
decode? That's a bias. Shouldn't decodes stand on
their
own?
-73, Glenn, AF8C
On 1/26/2024 12:32 PM, Reino Talarmo via wsjt-devel
wrote:
Hi Glenn, congrats about a nice false decode. The
'a3'
tells that the decoder tried as the known
information
your call, DX
call, and rest is decoded. The '?'
indicates that the probability of correct message is
not
too high.
Also the contents of the decoded part is suspicious
as
the report is
just 'R' and the locator 'RA92' is in the middle of
Antarctica.
73, Reino OH3mA
-----Original Message-----
From: Glenn Williams via wsjt-devel [mailto:wsjt-
de...@lists.sourceforge.net]
Sent: Friday, January 26, 2024 6:43 PM
To: wsjt-devel@lists.sourceforge.net
Cc: Glenn Williams <a...@alumni.caltech.edu>
Subject: Re: [wsjt-devel] wsjt-devel Digest, Vol
119,
Issue
45
I am reporting here another issue with my TX5S QSO
and the FT8
windows that happened last night. (Previous report
is under
"Subject:
[wsjt-devel] skips RX frequency window (bug?)
wsjtx-
2.7.0-rc3".) I am also NOT confirmed in ClubLog!
My screenshots with iPhone are in
https://www.hamfaqs.com/tx5s
and I don't believe it was RFI because I was ONLY
receiving at the
time!
--73, Glenn, AF8C
On 1/26/2024 11:28 AM, Uwe, DG2YCB via wsjt-devel
wrote:
Hi Fred and all,
It's really strange. Seems to be working well
here.
Look
at the Rx
Frequency windows of the three instances. (all
connected via virtual
audio, so nothing went over HF). To the left the
Fox
station, and to
the right the two hounds. I tried to reproduce
your
QSO situation when
the Fox was also in QSO with JH1BAM. Do you see
any
message not being
displayed there?
73 de DG2YCB,
Uwe
________________________________________
German Amateur Radio Station DG2YCB
Dr. Uwe Risse
eMail: dg2...@gmx.de
Info: www.qrz.com/db/DG2YCB
--
This email has been checked for viruses by Avast antivirus software.
www.avast.com
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
