Reverted to default stack allocation for jt9.exe on Windows.
Tidied up FT8 decoder inteface to remove arguments that will probably
never be used.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7768 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
Because Commander queues up CAT commands the timing of Tx audio was
not being correctly synchronized to the rig PTT. This change polls the
rig PTT state after changing PTT and waits for it to go to the
expected state. As it happens, just doing a query does most of the
synchronization as the query gets queued behind the PTT directive
within Commander, with the reply being delayed until the query
completes. This change goes further and checks the PTT state changes
within one second.
Note that users using Commander will should not suffer from ALC
overshoot issues on rigs that suffer such a problem since Tx audio now
only commences after the rig has switched to transmit. Users with
hardware sequencers may need to adjust their "Settings->Advanced->Tx
Delay" parameter, this change does not introduce any new timing hazard
but there may be a slight lengthening of the delay to Tx audio if the
rig reports PTT via CAT as the audio delay may now be timed from the
actual PTT on the rig rather than the assertion of RTS/DTR.
Still outstanding is the slow processing of Commander CAT requests
which can cause considerable delays before Tx audio can start.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7764 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
When using split operating and/or rig mode setting there is a delay
before PTT is asserted while the CAT commands are processed. There
will even be a small delay if CAT is only used for PTT. This change
waits for these commands to be completed before starting the user Tx
delay specified in the settings.
There is a notable exception to this in that DX Lab Suite Commander
does not give us a way of knowing when CAT commands have actually been
completed. This combined with the way that Commander spaces CAT
commands, rather than being paced by the rig's CAT responses, means
that delays of more than 1 second are likely. As most rigs suffer some
sort of ALC overshoot when PTT is asserted with audio already present,
it is strongly recommended not to use the DX Lab Suite Commander as a
rig control proxy if your rig may suffer ALC overshoot and you use an
external PA. Please note that using a typical RF power meter to detect
ALC overshoots is unwise as they are rarely fast enough. It is
probably wise to assume that all rigs suffer ALC overshoot since it is
very difficult to design an ALC loop that avoids it.
This change also increases the range of the "Settings->Advanced->Tx
Delay" spin box to 0s up to 2s.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7756 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
For FT8 mode:
1. Open LU 13 in the proper way.
2. Disable diagnostic LU 51.
3. Add keyboard shortcuts CTRL+Plus, CTRL-Minus, to adjust audio Tx frequency
in even 50 Hz steps.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7749 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
This feature is of limited benefit and causes unexpected behaviour on
rigs with lower than 1Hz resolution. Removed for simplicity.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7743 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. osd codeword selected based on correlation discrepancy instead of Hamming dist (improves sensitivity)
3. trap zero-valued spectrum values caused by large dt.
4. trap all-zero codeword
4. both decoders return number of hard errors
5. osd174 returns correlation discrepancy (rejection threshold still needs to be tuned).
6. both decoders return the codeword with bits in as-transmitted order, for subtraction.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7735 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. ft8d is now a bare-bones multi-decoder.
3. wsjt-x now includes the bare-bones multi-decoder for FT8.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/branches/wsjtx@7729 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
