Audit/repair of the multi-model denoise work so it actually builds and only
exposes working, self-hosted models.
- Complete the DTLN/DFN3 revert: uninstall @workadventure/noise-suppression
and deepfilternet3-noise-filter (package.json + lockfile), drop the unused
DTLN asset-copy block from vite.config.js (was shipping ~2MB of unused
tflite/wasm), and narrow DenoiseModelId to the bundled models (rnnoise,
speex). Coerce any retired persisted model value back to the default.
- Fix General.tsx CI typecheck failures introduced by the denoise UI: restore
three imports the rewrite deleted (useDateFormatItems, SequenceCardStyle,
useTauriUpdater), add the missing denoise/sound imports, and correct
hallucinated Folds props (Text has no variant/bold; Box uses
alignItems/justifyContent). tsc now passes with 0 errors.
- Harden the vite denoise plugin: required RNNoise/Speex/gate assets and the
shim now fail the build loudly if missing (instead of a silent warn that
shipped a broken ML feature), and the index.html shim injection is verified.
- CI: move the cinny-desktop submodule bump into ci.yml as a `trigger-desktop`
job gated on `needs: build`, and delete the standalone trigger-desktop.yml.
A failing push no longer kicks off the slow Tauri builds in parallel.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implement a flexible, multi-model noise suppression pipeline for Element Call/LiveKit integration:
- ML Engines: Added support for RNNoise, Speex, DTLN, and DeepFilterNet 3 models.
- Pipeline Architecture: Implemented modular audio processing in lotus-denoise.js, supporting 'Series Suppression' (running browser-native NSNet2 before ML) and a hardware-style Noise Gate.
- UI & UX Enhancements:
- Settings UI: Added model comparison chart with CPU/Quality metadata.
- Tuning: Added Live Microphone Meter for calibrating Noise Gate thresholds.
- Reporting: Added LotusToast system to alert users when ML suppression fails or falls back to raw input.
- Robustness & Quality:
- Capture Fidelity: Removed forced 48kHz capture constraints to allow native-rate capture (solving static issues with high-end audio interfaces).
- Performance: Added WASM SIMD detection with transparent fallback.
- Capability Detection: Added browser feature detection to disable unsupported ML modes.
- Build Integration: Updated Vite config to self-host all model WASM/tflite assets in /denoise/ directory.
ML noise suppression produced loud static on real calls. RNNoise requires
mono 48kHz float input; feeding it stereo or wrong-rate data is the classic
cause of that static. Harden the shim:
- request mono (channelCount:1) + 48kHz capture
- run a 48kHz AudioContext and BAIL to the raw mic if the browser won't
give a true 48kHz context (wrong-rate data -> static)
- force the worklet node to explicit mono in/out
- use the non-SIMD rnnoise.wasm (SIMD build artifacts on some GPUs)
- share one AudioContext across captures
Also fix the two CI-blocking eslint errors (unused vars in UrlPreviewCard
and useLocalMessageSearch) and apply repo-wide prettier formatting so
check:eslint and check:prettier pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the boolean call noise-suppression setting with a 3-way control
(Off / Browser-native / ML beta) in Settings -> General -> Calls.
- Off: noiseSuppression=false to Element Call
- Browser-native: EC's built-in WebRTC suppressor (prior default)
- ML (beta): on-device RNNoise (@sapphi-red/web-noise-suppressor)
Element Call captures the mic inside its iframe and publishes to LiveKit,
so the host can't reach that track; LiveKit's Krisp filter is Cloud-only
(we self-host the SFU) and EC's own RNNoise PR #3892 is unmerged. The ML
tier instead injects a same-origin pre-init shim into the vendored EC
index.html (build/lotus-denoise.js, wired by the lotusDenoise vite plugin)
that patches getUserMedia and routes the captured mic through an RNNoise
AudioWorklet before LiveKit sees it -- the same post-capture pipeline as
#3892, with no EC fork/AGPL/rebase burden. Falls back to the raw mic if
setup fails; keeps echoCancellation/AGC on the raw capture.
- settings.ts: callNoiseSuppression -> 'off'|'browser'|'ml' + legacy
boolean migration (true->browser, false->off)
- CallEmbed/useCallEmbed: tier maps to noiseSuppression param and appends
lotusDenoise=ml (native suppressor off in ML mode)
- vite.config.js: copy RNNoise worklet/wasm + shim into the EC bundle and
inject the shim <script> before EC's module entry
- docs: LOTUS_FEATURES.md, LOTUS_TODO.md (P5-30 done)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Point DECORATION_CDN at Lotus Nextcloud WebDAV share instead of external
avatardecoration.com; all 99 APNG files are self-hosted and served via
direct DAV URL (no CORS issue for <img> elements)
- Add onError handler to AvatarDecoration.tsx to silently hide the overlay
if a file is missing or the CDN is unreachable
- Rewrite scripts/syncDecorations.mjs: now sends HTTP HEAD requests to the
live Nextcloud CDN (batches of 16 in parallel) and removes catalog entries
for files that return non-2xx; empty categories are pruned automatically.
Workflow: delete files from Nextcloud → run `npm run sync:decorations` →
commit the updated avatarDecorations.ts. No local files needed.
- Add public/decorations/ to .gitignore; delete the 85 MB local APNG cache
that was downloaded during development (files live on Nextcloud now)
- Add sync:decorations script to package.json
- Update LOTUS_FEATURES.md, LOTUS_TODO.md (P5-13 + P5-14 ✓), README.md
with avatar decoration documentation and catalog sync workflow
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- usePresenceUpdater: replace stale closure with readStatus() called at
invocation time so changing custom status in Profile Settings is never
silently overwritten by subsequent activity events
- CallEmbedProvider: fix m.space.parent state-key lookup by switching
getStateEvent → getStateEvents (plural); space channel voice rooms no
longer trigger the incoming-call ring/animation
- Add useUserNotes hook (io.lotus.user_notes account data, reactive via
useAccountDataCallback, 500-char limit, cross-device sync)
- UserRoomProfile: add UserPrivateNotes textarea with 800ms debounced
auto-save, saving indicator, char counter when <100 chars remain;
shown only when viewing another user's profile
- LOTUS_FEATURES.md: add Private Notes section, Status Revert fix note,
animation improvements subsection, Seasonal Themes section
- LOTUS_BUGS.md: mark presence revert + voice ringing bugs as resolved
- README.md + landing/index.html: document all new June 2026 features
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Sounds (P5-16): browsers block the Web Audio context until a user gesture
starts it, so join/leave sounds — which fire later with no gesture — were
silent. unlockCallSounds() now primes/resumes the shared AudioContext inside
the Join click (centralized in useCallStart so every join path is covered),
making the per-client sounds reliably audible to everyone in the call.
Voice limit (P5-10): the limit is now a hard, cross-client server-side cap
enforced by the voice-limit-guard sidecar (matrix repo) that fronts
lk-jwt-service and refuses LiveKit tokens when a room is full. Updated
LOTUS_FEATURES.md / README.md / LOTUS_TODO.md to reflect that the client
'Channel Full' check is UX only and the server is authoritative.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
P5-10 Voice Channel User Limit:
- New StateEvent.LotusVoiceLimit (io.lotus.voice_limit) with { max_users }
- RoomVoiceLimit admin control in Room Settings > General > Voice
(power-level gated via permissions.stateEvent)
- CallPrescreen reads the limit reactively and disables Join with a
'Channel Full (N/N)' message at capacity; existing members can rejoin
P5-16 Custom Join/Leave Sound Effects:
- useCallJoinLeaveSounds hook wired into CallUtils; detects participant
join/leave via MatrixRTCSession membership changes (sender|deviceId),
filters out self, only fires while joined
- Sounds synthesized in-browser with Web Audio (callSounds.ts) - no
assets bundled; styles Off/Chime/Soft/Retro
- 'Join & Leave Sounds' selector in Settings > Calls (previews on change)
Docs: LOTUS_FEATURES.md, README.md, LOTUS_TODO.md (P5-10/P5-16 marked done)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- LOTUS_FEATURES.md: added sections for Knock-to-Join Notifications for
Admins and AFK Auto-Mute in Voice under their parent headings
- LOTUS_TODO.md: marked P4-3 and P5-11 as [x] completed
- README.md: updated Calls & Voice bullet list with AFK auto-mute entry;
expanded knock admin badge entry with badge-count detail
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- P5-21: Custom @mention highlight color picker in Settings → Appearance.
CSS vars with luminance-computed text color; resets cleanly to theme default.
- P5-22: Font selector (System, Inter, JetBrains Mono, Fira Code) in
Settings → Appearance. Fira Code added to Google Fonts preload.
- P5-27: Gaming/Work/Sleep preset buttons at top of Settings → Notifications.
Each atomically applies a group of notification settings.
- AppearanceEffects component in App.tsx applies CSS vars on settings change.
- LOTUS_BUGS.md: mark presence + manifest icon bugs as resolved.
- LOTUS_TODO.md: mark P3-6, P5-21, P5-22, P5-27 as [x].
- LOTUS_FEATURES.md: document all four completed features.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- LOTUS_FEATURES.md: full reference of every custom Lotus feature with
implementation details, file paths, and API notes
- LOTUS_BUGS.md: audit of confirmed bugs with root causes and fixes
- LOTUS_TODO_REFERENCE.md: technical implementation notes for backlog items
- LOTUS_TODO.md: trim completed audit results section, link to FEATURES doc
- README.md: rewrite to marketing-friendly feature list format; fix
incorrect claim that screenshare auto-reverts view to grid (removed)
Fix: auto-revert spotlight on screenshare was removed (600ms grid-click
caused fullscreen to show avatars); corrected in LOTUS_FEATURES.md and
removed from README.md.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
jared