Add 4 new audio visualization modes powered by MusicAnalyzer:
- pulse_on_beat: BPM-synced pulsing with smooth beat phase
- energy_gradient: bass/mid/treble mapped to scrolling gradient
- spectrum_bands: three VU zones for frequency bands
- strobe_on_drop: state-driven strobe on detected musical drops
MusicAnalyzer provides BPM estimation (median IBI), beat phase tracking,
asymmetric energy envelope, 3-band frequency splitting, and drop
detection state machine (idle/buildup/drop/recovery).
Add auto_gain audio filter for automatic level normalization via rolling
peak tracking with configurable target level and response time.
Deprecate auto_gain on Audio Value Source (use the filter instead).
Mathematical wave generator that produces per-LED colors from
configurable waveform layers (sine, triangle, sawtooth, square) with
superposition, mapped through a gradient palette. Supports sync clocks,
bindable speed, and up to 8 wave layers.
- Storage model with wave validation and apply_update
- Numpy-vectorized stream with gradient LUT color mapping
- API schemas (create/update/response) and route registration
- Frontend editor with dynamic wave layer rows, gradient picker,
speed widget, and IconSelect waveform selectors
- i18n in en/ru/zh
Fix test_list_filters test (filter_id field name mismatch).
Add tests for audio filters, template store, and source store.
All 678 tests pass, ruff clean, tsc clean, esbuild clean.
No dead code remaining from old source types.
Add AudioFilterPipeline for chained filter execution on AudioAnalysis.
Wire filter pipelines into AudioColorStripStream, AudioValueStream,
and WebSocket test endpoint. Add hot-update support via
ProcessorManager.refresh_audio_filter_pipelines(). Thread
AudioProcessingTemplateStore through dependency injection hierarchy.
Replace MultichannelAudioSource with CaptureAudioSource, add
ProcessedAudioSource (audio_source_id + audio_processing_template_id),
remove MonoAudioSource and BandExtractAudioSource entirely.
Update store resolution to walk processed chains collecting template IDs.
Update all API schemas, routes, and frontend references.
Add the foundation for audio processing filters, mirroring the existing
picture filter/postprocessing template system:
- AudioFilter base class, AudioFilterRegistry, AudioFilterOptionDef
- AudioProcessingTemplate dataclass + SQLite-backed store
- audio_filter_template meta-filter with recursive resolution
- Full REST API: CRUD templates + filter registry discovery
- Dependency injection wired in dependencies.py and main.py
MQTT broker connections are now managed as entities (like HA sources)
instead of a single global config. Each MQTTSource gets its own
runtime with auto-reconnect, ref-counted via MQTTManager.
Backend:
- MQTTSource dataclass + MQTTSourceStore (SQLite)
- MQTTRuntime (per-broker connection, refactored from MQTTService)
- MQTTManager (ref-counted pool, same pattern as HAManager)
- CRUD API at /api/v1/mqtt/sources + test + status endpoints
- MQTTRule gains mqtt_source_id field for source selection
- Automation engine acquires/releases MQTT runtimes automatically
- Legacy MQTTService kept for backward compat during transition
Frontend:
- MQTT source cards in Streams > Integrations tab
- Create/edit modal with test button
- Dashboard integration cards with health-dot indicators
- Removed MQTT tab from settings modal
Receive real-time events from games (CS2, Dota 2, LoL, etc.) and drive
LED effects through the existing color strip and value source pipelines.
Core:
- GameEventBus (thread-safe pub/sub) with standardized 23-type event vocabulary
- GameAdapter ABC + AdapterRegistry + MappingAdapter (YAML-driven)
- Built-in adapters: CS2 GSI, Dota 2 GSI, LoL Live Client, Generic Webhook
- Community YAML adapters: Minecraft, Valorant, Rocket League
- GameEventColorStripStream with 5 effects (flash/pulse/sweep/color_shift/breathing)
- GameEventValueSource with EMA smoothing and timeout
- 4 built-in effect presets (FPS Combat, MOBA Health, Racing, Generic Alert)
- Auto-setup for Valve GSI games (Steam path detection, cfg file writing)
- Demo capture engine exposed to non-demo mode
Frontend:
- Game tab in Streams tree navigation with integration cards
- Game integration editor modal with adapter picker, config fields, event mappings
- game_event source type in CSS and ValueSource editors
- Setup instructions overlay (markdown rendered)
- Live event monitor and connection test
API:
- Full CRUD for game integrations
- Event ingestion endpoint (adapter-level auth)
- Adapter metadata, presets, auto-setup, status/diagnostics endpoints
New value source that monitors host hardware via psutil/pynvml:
cpu_load, cpu_temp, gpu_load, gpu_temp, ram_usage, disk_usage,
network_rx, network_tx, battery_level, fan_speed.
Each metric normalizes to 0.0-1.0 with configurable ranges, poll
interval, EMA smoothing, and sensor_label for multi-sensor systems.
Conditional editor fields show/hide based on selected metric.
Also fixes: WS test crash when raw_value streams lack _min_ha attr,
toast timer overlap on rapid calls, SW cache bump to v34.
Add navigateToCard crosslinks for ha_entity (→ HA source), gradient_map
(→ input value source + gradient entity), and css_extract (→ color strip).
Gradient map test now charts the input interpolation factor instead of
output luminance, making the 0–1 chart meaningful.
WS now sends color RGB data for color-type streams. Test modal renders
a color history swatch strip below the chart, colors the chart line and
fill area with the current output color, and shows rgb() in the stats.
Works for static_color, animated_color, and adaptive_time_color sources.
- WS sends raw_value and raw_range for HA entity streams
- Canvas draws right-side Y-axis labels: configured min/max range
and current raw HA value (green, positioned at correct Y)
- Replace plain <select> for scene behavior with IconSelect (moon/sun)
New value source types:
- ha_entity: reads numeric values from HA entity state/attribute, normalizes
via min/max range, applies EMA smoothing. EntitySelect for HA connection
and entity selection with live entity list fetching.
- gradient_map: maps a float value source (0-1) through a gradient entity.
EntitySelect for both input source and gradient with inline previews.
- css_extract: extracts single color by averaging LED range from a color
strip source. EntitySelect for source selection.
Value source type picker:
- Filter tabs (All / Numeric / Color) above the icon grid
- showTypePicker extended with filterTabs + onFilterChange support
Palette selectors converted to EntitySelect:
- Effect palette, gradient preset, and audio palette selectors now use
command-palette style EntitySelect with gradient strip previews
Tab indicator fixes:
- Icon now updates on tab switch (was passing no args to updateTabIndicator)
- Visible with any background effect active, not just Noise Field
- Noise Field is the default background effect for new users
Dashboard section collapse fix:
- Split header into clickable toggle (chevron+label) and non-clickable
actions area — buttons no longer trigger collapse/expand
Discriminated union fix (422 errors):
- source_type/target_type now always included in update payloads for:
CSS editor, LED target, HA light target, simple calibration,
advanced calibration
BetterCam (DXGI) only supports one capture session per display, so creating
a second engine for the KC test produced no frames when a target was already
running. Now uses LiveStreamManager.acquire() which ref-counts and reuses
the running capture.
Also removed double postprocessing — ProcessedLiveStream already applies
filters, so re-applying them in the KC test halved the resolution (960x400
→ 240x100).
Introduce BindableFloat abstraction that allows any numeric property to be
either a static value or dynamically driven by a ValueSource. Backward-compatible
serialization: plain float when unbound, {value, source_id} dict when bound.
Backend:
- storage/bindable.py — BindableFloat dataclass + bfloat() helper
- 25+ scalar properties converted across all entity types
- Runtime VS acquisition in ColorStripStreamManager for CSS bindings
- All stream hot loops use self.resolve() for live values
- KeyColorsColorStripStream now inherits ColorStripStream
Frontend:
- BindableScalarWidget (slider + VS picker toggle) for all editors
- TypeScript BindableFloat type + helpers
- Graph editor edges for all bindable properties
- Audio source channel IconSelect grid
Fixes: daylight longitude, candlelight wind_strength/candle_type from_dict
- Add brightness_value_source_id to HALightOutputTarget model, to_dict,
from_dict, update_fields, register_with_manager, API response
- Wire value stream in HALightTargetProcessor: acquire/release on
start/stop, multiply brightness in _update_lights loop
- Fix transition=0 not saving (parseFloat("0") || 0.5 was falsy)
- Fix dashboard showing "Key Colors" for HA targets — now "Home Assistant"
- Fix dashboard FPS showing 0/2 — HA targets show target/target
- Add CSS source subtitle to HA target dashboard cards
Key Colors refactor:
- New `key_colors` CSS source type with inline rectangles
- KeyColorsColorStripStream: extracts N colors from screen regions
- CSS editor: EntitySelect for picture source, IconSelect for color mode
- Configure Regions button on card opens pattern canvas editor
- Live WS preview at 5 FPS with rectangle overlay + color swatches
- Removed KC target type, pattern template entity, and related API routes
- Removed KC/pattern template sections from Targets tab
HA light target improvements:
- Update rate, transition, mappings, brightness VS now editable via PUT
- Card crosslinks for HA source, CSS source, brightness VS
- HA connection status icon, text metrics (Hz, uptime)
- Brightness value source selector in editor
New output target type `ha_light` that sends averaged LED colors to HA
light entities via WebSocket service calls (light.turn_on/turn_off):
Backend:
- HARuntime.call_service(): fire-and-forget WS service calls
- HALightOutputTarget: data model with light mappings, update rate, transition
- HALightTargetProcessor: processing loop with delta detection, rate limiting
- ProcessorManager.add_ha_light_target(): registration
- API schemas/routes updated for ha_light target type
Frontend:
- HA Light Targets section in Targets tab tree nav
- Modal editor: HA source picker, CSS source picker, light entity mappings
- Target cards with start/stop/clone/edit actions
- i18n keys for all new UI strings
Add full Home Assistant integration via WebSocket API:
- HARuntime: persistent WebSocket client with auth, auto-reconnect, entity state cache
- HAManager: ref-counted runtime pool (like WeatherManager)
- HomeAssistantCondition: new automation trigger type matching entity states
- REST API: CRUD for HA sources + /test, /entities, /status endpoints
- /api/v1/system/integrations-status: combined MQTT + HA dashboard indicators
- Frontend: HA Sources tab in Streams, condition type in automation editor
- Modal editor with host, token, SSL, entity filters
- websockets>=13.0 dependency added
- Dismissible donation/open-source banner after 3+ sessions (30-day snooze)
- New About tab in Settings: version, repo link, license info
- Centralize project URLs (REPO_URL, DONATE_URL) in __init__.py, served via /health
- Center settings tab bar, reduce tab padding for 6-tab fit
- External URL save button: icon button instead of full-width text button
- Remove redundant settings footer close button
- Footer "Source Code" link replaced with "About" opening settings
- i18n keys for en/ru/zh
- Replace URL-based image_source/url fields with image_asset_id/video_asset_id
on StaticImagePictureSource and VideoCaptureSource (clean break, no migration)
- Resolve asset IDs to file paths at runtime via AssetStore.get_file_path()
- Add EntitySelect asset pickers for image/video in stream editor modal
- Add notification sound configuration (global sound + per-app overrides)
- Unify per-app color and sound overrides into single "Per-App Overrides" section
- Persist notification history between server restarts
- Add asset management system (upload, edit, delete, soft-delete)
- Replace emoji buttons with SVG icons throughout UI
- Various backend improvements: SQLite stores, auth, backup, MQTT, webhooks
API Input:
- Add interpolation mode (linear/nearest/none) for LED count mismatch
between incoming data and device LED count
- New IconSelect in editor, i18n for en/ru/zh
- Mark crossfade as won't-do (client owns temporal transitions)
- Mark last-write-wins as already implemented
LED Preview:
- Fix zone-mode preview parsing composite wire format (0xFE header
bytes were rendered as color data, garbling multi-zone previews)
- Fix _restoreLedPreviewState to handle zone-mode panels
FPS Charts:
- Seed target card charts from server metrics-history on first load
- Add fetchMetricsHistory() with 5s TTL cache shared across
dashboard, targets, perf-charts, and graph-editor
- Fix chart padding: pass maxSamples per caller (120 for dashboard,
30 for target cards) instead of hardcoded 120
- Fix dashboard chart empty on tab switch (always fetch server history)
- Left-pad with nulls for consistent chart width across targets
Dashboard:
- Fix metrics row alignment (grid layout with fixed column widths)
- Fix FPS label overflow into uptime column
- Create utils/image_codec.py with cv2-based image helpers
- Replace PIL usage across all routes, filters, and engines with cv2
- Move Pillow from core deps to [tray] optional in pyproject.toml
- Extract shared build logic into build-common.sh (detect_version, cleanup, etc.)
- Strip unused NumPy/PIL/zeroconf/debug files in build scripts
Replace 22 individual JSON store files with a single SQLite database
(data/ledgrab.db). All entity stores now use BaseSqliteStore backed by
SQLite with WAL mode, write-through caching, and thread-safe access.
- Add Database class with SQLite backup/restore API
- Add BaseSqliteStore as drop-in replacement for BaseJsonStore
- Convert all 16 entity stores to SQLite
- Move global settings (MQTT, external URL, auto-backup) to SQLite
settings table
- Replace JSON backup/restore with SQLite snapshot backups (.db files)
- Remove partial export/import feature (backend + frontend)
- Update demo seed to write directly to SQLite
- Add "Backup Now" button to settings UI
- Remove StorageConfig file path fields (single database_file remains)
New audio source type that filters a parent source to a specific frequency
band (bass 20-250Hz, mid 250-4kHz, treble 4k-20kHz, or custom range).
Supports chaining with frequency range intersection and cycle detection.
Band filtering applied in both CSS audio streams and test WebSocket.
New standalone WeatherSource entity with pluggable provider architecture
(Open-Meteo v1, free, no API key). Full CRUD, test endpoint, browser
geolocation, IconSelect provider picker, CardSection with test/clone/edit.
WeatherColorStripStream maps WMO weather codes to ambient color palettes
with temperature hue shifting and thunderstorm flash effects. Ref-counted
WeatherManager polls API and caches data per source.
CSS editor integration: weather type with EntitySelect source picker,
speed and temperature influence sliders. Backup/restore support.
i18n for en/ru/zh.
Composite layers now support optional start/end LED range (toggleable)
and reverse flag, making composite a superset of mapped source. Layers
are collapsible with animated expand/collapse and consistent 0.85rem
font sizing. Delete button restyled as ghost icon button.
Also includes minor dashboard CSS overflow fixes.
- Add /api/v1/system/shutdown endpoint that triggers clean uvicorn exit
- Persist all 15 stores to disk during shutdown via _save_all_stores()
- Add force parameter to BaseJsonStore._save() to bypass restore freeze
- Restart script now requests graceful shutdown via API (15s timeout),
falls back to force-kill only if server doesn't exit in time
- Broadcast server_restarting event over WebSocket before shutdown
- Frontend shows "Server restarting..." overlay instantly on WS event,
replacing the old dynamically-created overlay from settings.ts
- Add server_ref module to share uvicorn Server + TrayManager refs
- Add i18n keys for restart overlay (en/ru/zh)
Add gradient_id field to color strip sources for referencing reusable
gradient entities. Improve audio stream processing and effect stream
with new parameters.
New notification effects:
- Chase: light bounces across strip with Gaussian glow tail
- Gradient flash: bright center fades to edges with exponential decay
Queue priority: notifications with color_override get high priority and
interrupt the current effect.
Also fixes transient preview for notification sources — adds WebSocket
"fire" command so inline preview works without a saved source, plus
auto-fires on preview open so the effect is visible immediately.
Effects: add 7 new procedural effects (rain, comet, bouncing ball, fireworks,
sparkle rain, lava lamp, wave interference) and custom palette support via
user-defined [[pos,R,G,B],...] stops.
Gradient: add easing functions (linear, ease_in_out, step, cubic) for stop
interpolation, plus noise_perturb and hue_rotate animation types.
Daylight: add longitude field and NOAA solar equations for accurate
sunrise/sunset based on latitude, longitude, and day of year.
Candlelight: add wind simulation (correlated gusts), candle type presets
(taper/votive/bonfire), and wax drip effect with localized brightness dips.
Also fixes editor preview to include all new fields for inline LED test.
Auth is now optional: when `auth.api_keys` is empty, all endpoints are
open (no login screen, no Bearer tokens). Health endpoint reports
`auth_required` so the frontend knows which mode to use.
Backup/restore fixes:
- Auto-backup uses atomic writes (was `write_text`, risked corruption)
- Startup backup skipped if recent backup exists (<5 min cooldown),
preventing rapid restarts from rotating out good backups
- Restore rejects all-empty backups to prevent accidental data wipes
- Store saves frozen after restore to prevent stale in-memory data
from overwriting freshly-restored files before restart completes
- Missing stores during restore logged as warnings
- STORE_MAP completeness verified at startup against StorageConfig
- Notification history: replace text buttons with icon buttons, use modal-footer for proper positioning
- CSS test: reject 0-LED picture sources with clear error message, show WS close reason in UI
- Calibration: distribute LEDs by aspect ratio (16:9 default) instead of evenly across edges
- Value source schedule: replace native time input with custom HH:MM picker matching automation style
- Remove "No ... yet" empty state labels from all CardSection instances
- Replace winsdk (~35MB) with winrt packages (~2.5MB) for OS notification
listener. API is identical, 93% size reduction.
- Replace wmi (~3-5MB) with ctypes for monitor names (EnumDisplayDevicesW)
and camera names (SetupAPI). Zero external dependency.
- Migrate cv2.resize/imencode/LUT to Pillow/numpy in 5 files (filters,
preview helpers, kc_target_processor). OpenCV only needed for camera
and video stream now.
- Fix DefWindowProcW ctypes overflow on 64-bit Python (pre-existing bug
in platform_detector display power listener).
- Fix openLightbox import in streams-capture-templates.ts (was using
broken window cast instead of direct import).
- Add mandatory data migration policy to CLAUDE.md after silent data
loss incident from storage file rename without migration.
Server: send initial frame immediately after metadata for api_input
sources so the client gets the current buffer (fallback color if
inactive) instead of never receiving a frame when push_generation
stays at 0.
Frontend: clear all test modal canvases on open to prevent stale
pixels from previous sessions. Reset FPS timestamps after metadata
so the initial bootstrap frame isn't counted in the FPS chart.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- KC test WS now acquires from LiveStreamManager instead of creating its
own DXGI duplicator, eliminating capture contention with running LED targets
- Tree-nav refactored to compact dropdown on click with outside-click dismiss
(closes on click outside the trigger+panel, not just outside the container)
- KC target card badge (e.g. "Daylight Cycle") no longer wastes empty space
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Settings modal split into 3 tabs: General, Backup, MQTT
- Log viewer moved to full-screen overlay with compact toolbar
- External URL setting: API endpoints + UI for configuring server domain
used in webhook/WS URLs instead of auto-detected local IP
- Sources tab tree restructured: Picture Source (Screen Capture/Static/
Processed sub-groups), Color Strip, Audio, Utility
- TreeNav extended to support nested groups (3-level tree)
- Audio tab split into Sources and Templates sub-tabs
- Fix audio template test: device picker now filters by engine type
(was showing WASAPI indices for sounddevice templates)
- Audio template test device picker disabled during active test
- Rename "Input Source" to "Source" in CSS test preview (en/ru/zh)
- Fix i18n: log filter/level items deferred to avoid stale t() calls
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New features:
- Override composite blend mode: per-pixel alpha from brightness
(black=transparent, bright=opaque). Ideal for API input over effects.
- API input test preview FPS chart uses shared createFpsSparkline
(same look as target card charts)
Fixes:
- Fix api_input source not surviving server restart: from_dict was
still passing removed led_count field to constructor
- Fix composite layer brightness/processing selectors not aligned:
labels get fixed width, selects fill remaining space
- Fix CSPT input selector showing in non-CSPT CSS test mode
- Fix test modal LED/FPS controls showing for api_input sources
- Server only sends test WS frames when api_input push_generation changes
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
API Input CSS rework:
- Remove led_count field from ApiInputColorStripSource (always auto-sizes)
- Add segment-based payload: solid, per_pixel, gradient modes
- Segments applied in order (last wins on overlap), auto-grow buffer
- Backward compatible: legacy {"colors": [...]} still works
- Pydantic validation: mode-specific field requirements
Test preview:
- Enable test preview button on api_input cards
- Hide LED/FPS controls for api_input (sender controls those)
- Show input source selector for all CSS tests (preselected)
- FPS sparkline chart using shared createFpsSparkline (same as target cards)
- Server only sends frames when push_generation changes (no idle frames)
HAOS integration:
- New light.py: ApiInputLight entity per api_input source (RGB + brightness)
- turn_on pushes solid segment, turn_off pushes fallback color
- Register wled_screen_controller.set_leds service for arbitrary segments
- New services.yaml with field definitions
- Coordinator: push_colors() and push_segments() methods
- Platform.LIGHT added to platforms list
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Backend: send border_width in WS metadata and frame_dims (width, height)
as a separate JSON message on first JPEG frame.
Frontend: render semi-transparent green overlay rectangles on each active
edge showing the sampling region depth, plus a small px label. Overlays
are proportionally sized based on border_width relative to frame dimensions.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
alexei.dolgolyov