Thank you Joe and Bill and the rest of the team for this wonderful new
version!
I am finding that it does the Doppler compensation with Q65-15C
extremely rapidly even with the Max Drift set to 50 and with wide FTol
settings. Its performance is truly amazing!!
With 2.5.0-RC1 we are getting consistent decodes using Q65-15C with
Doppler shift rates of up to 20 Hz/sec, which is all the higher I have
tested thus far as we don't anticipate rates of change greater than this
with 10 GHz aircraft scatter, which is our current interest in this regard.
I have noticed one minor and esoteric issue when using mode Q65-15C,
which I am reporting for completeness, as requested in your announcement
email.
I apologize for the long message but I wanted to give all the details.
In testing I am using settings that I would not use in practice, in
order to test the extremes of the program environment, and I have found
that I can produce a sub-process error that results in program shut-down
by using an odd choice of settings when decoding wave files. The error
occurs at the time that decoding is attempted.
Report is below:
This is of course with WSJT-X version 2.5.0-RC1.
OS is Windows 10 Pro version 20H2.
The critical choice of settings that is required to produce the error
seems to be [1] Max Drift Rate greater than zero, with [2] combinations
of FTol and Rx frequency where FTol=500 or FTol=1000 and Rx frequency
are chosen so that the lower limit of Ftol's frequency range is zero or
negative. As the error message that this set of conditions produces
indicates, this combination of parameters pushes the index of dimension
1 of s1avg in q65.f90 below its lower limit of 1.
For FTol = 1000 any Rx Freq of 1003 or lower produces this error.
For FTol = 500 any Rx Freq of 503 or lower produces this error.
Interestingly, I don't see this when I choose FTol=400 or FTol=300 and
then select the Rx freq so as to push the lower limit of FTol below 0
and then attempt to decode. Because the lower limit of Rx freq is 200, I
of course can't produce this error condition with FTol values of 200 or
less.
And I don't see the error if I first set Rx Freq to a low value such as
200 Hz with FTol 400 Hz or lower and then increase FTol to 500 or 1000
so that the lower limit of FTol will be below zero and then attempt to
decode. But then, by clicking to several different Rx frequencies where
the lower limit of FTol is above zero and attempting decodes (which of
course produce no errors) and then clicking again to an Rx Freq so that
the lower FTol limit is below zero, I can again elicit the subprocess
error with FTol=500 or FTol=1000.
The other settings that are in use when this error occurs are:
Q65-15C
Tx freq 700 Hz
Max Drift 50
Decode Deep
Waterfall frequency limits are 200-3125 Hz.
With Rx Freq = 980 and FTol=1000 the MessageBox error produced shows:
---------------------------------
Subprocess Error
Subprocess failed with exit code 2.
"Details" are:
Running: C:\WSJT\wsjtx\2pt5pt0-rc1\bin\jt9 -s "WSJT-X - VK7MO_Tests" -w
1 -m 3 -e C:\WSJT\wsjtx\2pt5pt0-rc1\bin -a
"C:\Users\73w3s\AppData\Local\WSJT-X - VK7MO_Tests" -t
"C:\Users\73w3s\AppData\Local\Temp\WSJT-X - VK7MO_Tests"
At line 447 of file C:\Users\bill\src\k1jt\wsjtx\lib\qra\q65\q65.f90
Fortran runtime error: Index '-3' of dimension 1 of array 's1avg' below
lower bound of 1
Error termination. Backtrace:
Could not print backtrace: libbacktrace could not find executable to open
#0 0xffffffff
#1 0xffffffff
#2 0xffffffff
#3 0xffffffff
#4 0xffffffff
#5 0xffffffff
#6 0xffffffff
#7 0xffffffff
#8 0xffffffff
#9 0xffffffff
#10 0xffffffff
#11 0xffffffff
#12 0xffffffff
#13 0xffffffff
#14 0xffffffff
---------------------------------
With Rx freq = 1003 and Ftol=1000 details are the same except for the
index value:
Fortran runtime error: Index '0' of dimension 1 of array 's1avg' below
lower bound of 1
With Rx freq = 900 and FTol=1000 this line is:
Fortran runtime error: Index '-15' of dimension 1 of array 's1avg' below
lower bound of 1
WIth Rx freq = 436 and FTol=500 this line is:
Fortran runtime error: Index '-10' of dimension 1 of array 's1avg' below
lower bound of 1
etc.
This is not a complaint and I would not likely use such wide FTol values
in actual operation but I figured that I should report it in case you
had not noticed this behavior and in case you wanted to trap the error
condition.
Of course when the error MessageBox is closed, WSJT-X shuts down as is
appropriate.
Thanks again for this great software! I am really enjoying its amazing
capabilities, and the above issue does not detract at all from my use
and enjoyment of it!
73,
Roger Rehr
W3SZ
On 6/2/2021 05:34 PM, Joe Taylor wrote:
Dear WSJT-X and MAP65 Users,
We hope you will enjoy using this beta release of WSJT-X 2.5.0 and
MAP65 3.0.0, and exercising the new mode Q65. As a beta tester you
should report on your experiences with new features, successful and
otherwise, on one of the relevant WSJT forums. Bugs should be reported
by following instructions found here in the User Guide:
https://www.physics.princeton.edu//pulsar/K1JT/wsjtx-doc/wsjtx-main-2.5.0-rc1.html#_bug_reports
-- 73 from Joe, K1JT; Bill, G4WJS; Steve, K9AN; and Nico, IV3NWV
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
