转发原文
Hi all,
This message is to let you know of some important WSJT-X development
plans. We plan to make a first candidate release of WSJT-X 2.2.0 next
Monday, May 10.
WSJT-X 2.2.0-rc1 will be a beta-quality release candidate providing a
number of new features and capabilities. These include:
- Improvements to the decoders for five modes:
FT4: Corrected bugs that prevented AP decoding and/or multi-pass
decoding in some circumstances. The algorithm for AP
decoding has been improved and extended.
FT8: Decoding is now spread over three intervals. The first
starts at t = 11.8 s into an Rx sequence and typically yields
around 85% of the possible decodes for the sequence. You
therefore see most decodes much earlier than before. A second
processing step starts at 13.5 s, and the final one at 14.7 s.
Overall decoding yield on crowded bands is improved by 10% or
more. (Systems with receive latency greater than 0.2 s will see
smaller improvements, but will still see many decodes earlier
than before.)
JT4: Formatting and display of Averaged and Deep Search decodes
has been cleaned up and made consistent with other modes. JT4
remains the digital mode of choice for EME and other extreme
weak-signal work on microwave bands.
JT65: Many improvements for Averaged and Deep Search decodes and
their display to the user. These improvements are particularly
important for EME on VHF and UHF bands.
WSPR: Significant improvements have been made to the WSPR
decoder's sensitivity, its ability to cope with many signals in
a crowded sub-band, and its rate of undetected false decodes.
We now use up to three decoding passes. Passes 1 and 2 use
noncoherent demodulation of single symbols and allow for
frequency drifts up to ±4 Hz in a transmission. Pass 3 assumes
no drift and does coherent block detection of up to three
symbols. It also applies bit-by-bit normalization of the
single-symbol bit metrics, a technique that has proven helpful
for signals corrupted by artifacts of the subtraction of
stronger signals and also for LF/MF signals heavily cont-aminated
by lightning transients. With these improvements the number of
decodes in a crowded WSPR sub-band typically increases by 10 to
15%.
- New format for "EU VHF Contest" Tx2 and Tx3 messages
When "EU VHF Contest" is selected, the Tx2 and Tx3 messages
(those conveying signal report, serial number, and 6-character
locator) now use hashcodes for both callsigns. This change is
NOT backward compatible with earlier versions of _WSJT-X_, so
all users of EU VHF Contest messages should be sure to upgrade
to versiion 2.2.0.
- Accessibility
Keyboard shortcuts have been added as an aid to accessibility:
Alt+R sets Tx4 message to RR73, Ctrl+R sets it to RRR.
As an aid for partial color-blindness, the "inverted goal posts"
marking Rx frequency on the Wide Graph's frequency scale are now
rendered in a darker shade of green.
- Minor enhancements and bug fixes
"Save None" now writes no .wav files to disk, even temporarily.
An explicit entry for "WW Digi Contest" has been added to
"Special operating activities" on the "Settings | Advanced" tab.
Contest mode FT4 now always uses RR73 for the Tx4 message.
The Status bar now displays the number of decodes found in the
most recent Rx sequence.
Release candidate WSJT-X 2.2.0-rc1 will be available for beta-testing
for one month starting on May 10, 2020. We currently plan a General
Availability (GA) release of WSJT-X 2.2.0 on June 1, 2020.
For those looking even farther ahead: We are well along in the
development of two new modes designed for the LF and MF bands. One
mode is for WSPR-like activity and one for making 2-way QSOs. Both
use Low-density Parity Check (LDPC) codes, 4-GFSK modulation, and
two-minute T/R sequences. The QSO mode reaches threshold SNR
sensitivity around -31 dB on the AWGN channel, and the WSPR-like mode
better than -32 dB.
With best wishes,
-- Joe, K1JT, Steve, K9AN, and Bill, G4WJS
