• Joined on 2026-03-13
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 20:02:12 +03:00
bf8c12603c readme: sync to 22i — P5 status, floor blend*ln10/20, Step 7 refuted, detector front-end priority, new tooling
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 19:51:57 +03:00
d43bbfe66b 22i: RT_LVL_CAP inert (model post-IIR levels ~1.06 << cap) — gap relocated to detector front-end amplitude chain
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 19:46:24 +03:00
94383b15d3 22h: floor law mapped — floor_dB(sens)=20log10(ln10/20)-(sens-6)/3, depth shifts too; exact anchor sens6=ln10/20
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 19:39:22 +03:00
ef7d3f49c7 22g: NO-LUT test isolates gap to detector front-end; unified sidechain-clamp hypothesis (knee at 0dBFS, identical renders above)
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 19:10:51 +03:00
41b0d3c250 22f: floor discrimination via param bridge — dry/wet crossfade confirmed, floors=k*ln10/20, detector-cap hypothesis leads
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 18:59:05 +03:00
7febe744e9 plan: sync to 22e — Step 7 refuted, floor blend*ln10/20 solved, new priority Step 9 (floor mechanism)
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 18:54:48 +03:00
3970c1e309 22e: floor algebra solved — gain_floor=20log10(blend*ln10/20), matches -20.72dB within 0.006dB; lead FUN_180529c60
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 18:40:15 +03:00
9c14a74318 docs: 22d — hard gain floor -20.72dB mapped (t-clamp live confirmation); cold/hot discrepancy open
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 18:30:55 +03:00
f5ef81667c bench: official param bridge (setparam/dump_params.lua) + rpp_setparam format tool; 22c — XML params decorative, depth=+-18dB, trim=+-24dB, reduction curves measured
7fac8b2a4f step7: live BandConfig capture infra + FUN_180563a60 decode — GUI-only, audio path refutes premise (22b)
c222e054ca docs: Phase B recovery — 4 offline detector hypotheses refuted; scripts into scripts/
Compare 3 commits »
Matiq pushed to main at Matiq/soothe2-re 2026-08-22 10:59:29 +03:00
71644ff3e6 fix(render48k): chunk-scan WAV loader — hdr[40] is not data size when bext/junk present
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 19:29:22 +03:00
3e5085e06d phase A: reduction law found (affine in dB), canon unchanged; keep RT_DUMP tooling
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 13:54:07 +03:00
702b74e297 docs: sync BITEXACT_PLAN/roadmap/AGENTS to 2026-08-21 state
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 13:12:55 +03:00
b2cb9235be feat(scripts): structural-path full-corpus harness + first honest 62-case numbers
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 12:55:23 +03:00
a9d5581ad8 docs: NOTES — IIR3 bidirectional decode, combine thunk semantics, Step 2 re-scope
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 12:54:48 +03:00
1a616e1b7b feat(dsp): IIR3 bidirectional per decomp (fwd+bwd x2), combine semantics decoded
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 12:45:37 +03:00
9ffe60e66e feat(dsp): wire IIR3 x2 (kIIR_A3/B3) post-warp into structural chain
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 12:42:16 +03:00
7a108c529a docs: NOTES_LEVEL update — LUT level-domain discovery, sweep tooling hazard
535b150f75 feat(dsp): move dB-LUT to level domain (pre-IIR/exp2) in structural chain
9b7e9d3099 feat(dsp): add dB-domain LUT (FUN_180563a60) to structural chain on 48k grid
6924e539e0 feat(dsp): render48k pipeline + structural chain on 48000/4096 grid
7659eb0362 feat(dsp): SpectralProcessor sample_rate param
Compare 6 commits »
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 01:06:07 +03:00
fa71240d92 test: add validation harness (scripts/corpus.py + bridge baseline) to prevent Phase B regression
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 01:00:44 +03:00
f17ee78061 docs: BITEXACT_PLAN — 8-step path to byte-parity; fix repo docs to reflect not-yet-bit-exact status
Matiq pushed to main at Matiq/soothe2-re 2026-08-21 00:52:44 +03:00
e990c21e94 fix: framed_test parse every comma band (was first-arg only); comb tests were single-band