The WSPR 7-bit power field contains the raw dBm value (0-60) with no
offset. The decoder was subtracting 64, turning valid power values into
negative numbers that always failed the range check, causing
unpack_message to return None for every real signal. Also fix callsign
trimming to strip leading spaces from space-padded callsigns.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The deinterleave function had its indices swapped: it wrote
out[j] = symbols[p] instead of out[p] = symbols[j]. This fed
completely scrambled data to the Fano decoder, making convergence
impossible. Matched against the reference implementation in
raptor/lib/wsprd/wsprd_utils.c which does tmp[p] = sym[j].
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The OSD-lite decoder was the source of FT2 false positives. It tries
685 CRC-14 checks across 5 passes (1 + 16 + 120 per pass), giving a
~4% chance of accepting random noise as a valid decode.
The reference implementation (decode174_91) verifies OSD results
against the received signal; the trx-rs OSD-lite only checked CRC.
Add ft2_count_hard_errors_vs_llr() which counts how many of the 174
coded bits in an OSD candidate disagree with the received hard
decisions. A legitimate correction disagrees in very few positions;
a false CRC match on noise disagrees in ~40-50 parity positions.
Reject OSD results with more than 36 hard errors.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The ft8_wrapper.c references ft4_encode and ft8_encode from encode.c,
but encode.c was not included in build.rs, causing linker errors.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The WSPR decoder was producing almost no valid decodes despite audible
signals. Three root causes:
- No sync vector: the 162-bit WSPR sync pattern was not used, so signal
detection relied on raw peak power which is unreliable.
- Coarse frequency search: 4 Hz steps with 1.465 Hz tone spacing could
miss signals entirely. Now uses 2 Hz coarse + 0.25 Hz fine refinement.
- Fixed timing: assumed signal starts exactly 1s into the slot. Now
searches +/-2s in 0.5s steps to handle real-world timing jitter.
Also evaluates up to 8 frequency/timing candidates per slot and reports
the actual measured timing offset in dt_s.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The previous SNR formula (cand->score * 0.5 - 29.0) used the adjacent
tone bin as a noise reference. On a crowded FT8 band that bin is often
occupied by another station, inflating the apparent noise floor by
10-15 dB and capping reported SNR at around -10 dB even for strong
signals.
Replace with ftx_post_decode_snr(): re-encode the decoded message to
obtain the exact per-symbol tone sequence, compare each signal bin
against the minimum of the remaining (noise-only) bins, average over
all valid symbols, and apply the WSJT-X 2500 Hz bandwidth correction
dynamically per protocol. This produces accurate SNR estimates for both
FT8 and FT4 regardless of band occupancy.
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
Replace the always-None stub with a full decode pipeline:
- Bit-reversal deinterleave of 4-FSK symbols (data_bit = symbol >> 1)
- Fano sequential decoder for K=32, rate-1/2 convolutional code
(polynomials 0xF2D05351 / 0xE4613C47, 100k-cycle budget)
- Payload unpack: 28-bit callsign (mixed-radix N1), 15-bit Maidenhead
grid (M1 formula), 7-bit power code (dBm + 64)
- Validity checks on callsign, grid, and power range
- Round-trip unit test for K1JT FN20 37
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
The OSD-3 (triples) path over 5 LLR passes was doing ~11,600 CRC checks
per candidate. With a 14-bit CRC this gives ~0.7 expected false positives
per candidate — far too high.
Remove OSD-3 entirely. Cap max_candidates at 16 for OSD-1/OSD-2, giving
136 CRC checks per pass (680 total). Gate OSD-lite behind a check that
LDPC reached within 6 parity errors of converging, so it only fires when
the LLRs are already trustworthy. Combined false-positive rate drops to
~0.04 per near-miss candidate.
Also remove the now-unused ft2_osd_decode and ft2_codeword_distance
functions.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Increase BP/SP iteration count from 30 to 50 to match WSJT-X reference
and give belief propagation more opportunities to converge near-threshold
candidates.
Replace the parity-based OSD-1/OSD-2 fallback (which required LDPC to
have nearly converged) with ft2_osd_lite_decode applied to all five LLR
combination passes. The CRC-based decoder works directly from raw LLRs
without depending on LDPC convergence, searching the 24 least-reliable
systematic bits for up to three bit errors via OSD-3.
Also increase max_candidates in ft2_osd_lite_decode from 12 to 24 for
broader coverage of likely error positions.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Diagnostic logging showed the FT2 BP/SP decoders consistently reach
1-8 residual parity errors rather than zero — the LLRs are correct
in direction but LDPC belief propagation stalls just short of
convergence.
Add ft2_osd_decode() implementing Ordered Statistics Decoding orders
1 and 2: after the five-pass BP/SP loop fails, sort the 174 codeword
bits by |LLR| ascending and trial-flip single bits (OSD-1, always)
or all pairs of the 50 least-reliable bits (OSD-2, when the remaining
error count is <= 4). Each trial costs one O(83) parity check;
worst-case overhead is ~1300 checks per candidate, negligible next to
the 5 x 30-iteration BP/SP passes already performed.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Add err=N/N/N/N/N to the FT2 window diagnostic log line, showing the
minimum number of unsatisfied parity equations across all candidates
for each of the five LLR passes. This makes it possible to distinguish
between a signal-quality-limited failure (small error count) and a
systematic decoder bug (large error count), which is the key unknown
in diagnosing the current FT2 LDPC non-convergence.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Normalize FT2 log-likelihoods before LDPC and fall back to\nthe standard waterfall candidate decoder when the raw FT2\npath produces no decodes.\n\nCo-authored-by: OpenAI Codex <codex@openai.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Use an FT2-specific raw sample window and candidate acquisition path.
Keep the build clean by removing the stale monitor warning.
Co-authored-by: OpenAI Codex <codex@openai.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
Use FT2-specific analysis settings and a wider receive span.
Switch server-side FT2 decoding to a rolling async window.
Widen FT2 candidate timing search in the vendored decoder.
Co-authored-by: OpenAI Codex <codex@openai.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
Implement a distinct FT2 protocol path in the decoder stack and align\nits timing with the confirmed FT2 framing used by Decodium.\n\nCo-authored-by: OpenAI Codex <codex@openai.com>
Signed-off-by: Stanislaw Grams <stanislawgrams@gmail.com>
Skip decoded candidates where ftx_message_decode() returns a non-OK
status instead of forwarding a synthetic error string.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
The score from ft8_lib is an averaged uint8 difference between Costas
sync tones and their neighbours (each unit = 0.5 dB). The previous
score * 0.5 gave the signal-above-adjacent-noise in dB relative to a
single 3.125 Hz waterfall bin, yielding values of +5 to +50 dB —
all wrong.
Subtract 10*log10(2500/3.125) ≈ 29 dB to normalise to the 2500 Hz
reference bandwidth used by WSJT-X and expected by PSKReporter:
snr = score * 0.5 - 29.0
This maps score 10 (minimum decodable) → -24 dB and score 60 → +1 dB,
matching typical WSJT-X SNR report ranges.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Add [workspace.package] version = "0.1.0" to the root Cargo.toml and
switch all 21 member crates to version.workspace = true so the entire
workspace is versioned from a single place.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Adds a second APRS demodulator path tuned for the HF APRS standard
(300 baud Bell 103-style AFSK, mark=1600 Hz / space=1800 Hz), active
on RigMode::DIG. Shares AX.25 framing, APRS parsing, APRS-IS uplink,
and frontend display with the existing VHF stack.
- trx-aprs: parameterise Demodulator::new(); add AprsDecoder::new_hf()
- trx-core: HfAprs variant in DecodedMessage; hf_aprs_decode_enabled /
hf_aprs_decode_reset_seq in RigState/RigSnapshot; SetHfAprsDecodeEnabled
and ResetHfAprsDecoder commands; handlers.rs fallback arm updated
- trx-protocol: client command variants + bidirectional mapping; test
fixture updated
- trx-server: run_hf_aprs_decoder() task (activates on DIG mode);
hf_aprs history in DecoderHistories; rig_task command dispatch;
aprsfi uplink forwards HfAprs via OR-pattern
- trx-frontend: hf_aprs_history in FrontendRuntimeContext
- trx-frontend-http: prune/record/snapshot/clear helpers; SSE history
replay; toggle_hf_aprs_decode + clear_hf_aprs_decode endpoints;
/hf-aprs.js endpoint; HF APRS tab in web UI
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Replace the free-running phase counter in slice_pi4_qpsk_symbols with
linear interpolation to sample the IQ stream at exact symbol epochs.
Add estimate_differential_cfo() that uses the 4th-power method to
cancel pi/4-QPSK modulation phase, yielding a per-burst CFO estimate
that is removed before differential decoding.
At the ~1.25 samples/symbol IQ rate produced by the current decimation
pipeline, a closed-loop Gardner or Mueller-Müller TED requires at least
2 SPS and cannot be applied; the open-loop linear interpolation is the
best achievable without restructuring the IQ tap.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
- Reduce analysis window from 50ms to 10ms so the decoder can detect
dits at 25+ WPM (at 25 WPM a dit is 48ms, shorter than the old window)
- Fix dot/dash classification threshold from 2.0× to 1.5× unit_ms;
ITU Morse dah = 3× dit, so the midpoint boundary is 1.5×
- Replace O(n) on_durations.remove(0) with drain() to trim the window
- Remove pointless emit_text() wrapper; callers now call emit_event()
directly
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Keep manual CW tone setting range at 100-10000 Hz, but limit auto-tone
scanning to 300-1200 Hz to avoid locking onto high-frequency noise.
Retain Nyquist-safe upper clamping.
Co-authored-by: Codex <codex@openai.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Expand CW decoder tone scan/control range to 100-10000 Hz and cap the
upper bound by sample-rate Nyquist to keep detection stable.
Co-authored-by: Codex <codex@openai.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Restore empty events on tone edges so the frontend signal indicator updates again.
Add synthetic-tone tests covering transitions and basic decoding.
Co-authored-by: OpenAI Codex <codex@openai.com>
Signed-off-by: Stan Grams <sjg@haxx.space>
Keep geo sanity checks, but treat most marginal VDES decodes as low confidence instead of rejecting them outright.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
Score parsed VDES payloads and fall back to unsynced output for obviously weak or invalid decodes, including invalid geo boxes.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
Keep the current VDES thresholds but restore a maximum burst duration so continuously open detections still produce frames.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
Restore the more permissive intermediate VDES burst thresholds while keeping the current detector reset and close behavior.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
Publish decoded VDES positions into the map and revert the VDES burst detector to its original gating thresholds.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
Raise the VDES burst thresholds and minimum burst length to cut false positives after the recent sensitivity increase.
Co-authored-by: Stan Grams <sjg@haxx.space>
Signed-off-by: Stan Grams <sjg@haxx.space>
sjg
