Historical task source. The Eve task board holds current task state.
Phase 133: Composer DAW SOTA Expansion#
Closes the remaining professional music-production sovereignty gaps beyond Phase 44.3 (Composer DAW core) and Phase 132 (audio runtime + plugin hosting). Delivers notation engraving (Dorico / Sibelius class), Ableton Live session view, iZotope RX-class restoration, Melodyne-class polyphonic pitch, Ozone/Neutron-class mastering, VCV Rack-class modular, Max/MSP and SuperCollider-class patching and live coding, advanced synthesis (additive, physical modeling, modal, spectral, neural), scoring-to-picture, ADR/Foley workflows, DJ decks, and a sovereign sample-library content strategy.
Once complete, every professional music-production workflow that currently requires Logic Pro + Dorico + Melodyne + iZotope + VCV Rack + Ableton + Kontakt
- Final Cut Pro becomes doable entirely within the Oshun stack.
133.1 @neith/notation — Dorico-Class Notation Engraving#
Full music-notation authoring and engraving tool, not just MusicXML parsing.
133.1.1 Score Model#
- 133.1.1.1 Implement hierarchical score model (flow → system → staff → measure → voice → note)
- 133.1.1.2 Create time-signature hierarchy with additive/irregular meters
- 133.1.1.3 Implement key-signature handling with quarter-tone accidentals
- 133.1.1.4 Create instrument transposition and concert-pitch toggle
- 133.1.1.5 Implement clef types (treble, bass, C clefs, percussion, tab, neutral)
- 133.1.1.6 Create ossia and cue-sized note handling
- 133.1.1.7 Implement divisi and a2 / a3 auto-layout
- 133.1.1.8 Create grace notes (appoggiatura, acciaccatura)
- 133.1.1.9 Implement tuplets (including nested and irregular)
- 133.1.1.10 Create tremolo (measured and unmeasured)
- 133.1.1.11 Implement chord symbols with Roman numeral analysis
- 133.1.1.12 Create figured bass notation
- 133.1.1.13 Implement microtonal accidentals (24-EDO, 31-EDO, arbitrary)
- 133.1.1.14 Create tablature for fretted strings with fingering
- 133.1.1.15 Implement percussion notation with unpitched staves
133.1.2 Engraving Engine#
- 133.1.2.1 Implement SMuFL font loading (Bravura, Leland, Petaluma)
- 133.1.2.2 Create beam grouping with beat-group rules per meter
- 133.1.2.3 Implement slur and tie curve generation (Bezier with collision)
- 133.1.2.4 Create stem direction rules (voice, pitch, context)
- 133.1.2.5 Implement automatic rest consolidation
- 133.1.2.6 Create optimal spacing with note-value proportionality
- 133.1.2.7 Implement collision detection and avoidance for all elements
- 133.1.2.8 Create cross-staff beaming
- 133.1.2.9 Implement optical alignment (hanging lyrics, chord symbols)
- 133.1.2.10 Create horizontal justification across systems
- 133.1.2.11 Implement vertical justification across staves
- 133.1.2.12 Create system breaks with automatic casting-off
- 133.1.2.13 Implement page breaks with frontmatter/endmatter
- 133.1.2.14 Create music-frame / text-frame layout model
- 133.1.2.15 Implement rehearsal marks, cue marks, and section barlines
133.1.3 Dynamics, Articulations, Ornaments#
- 133.1.3.1 Standard dynamics (ppp–fff, sfz, fp, niente, cresc., dim.)
- 133.1.3.2 Hairpins with auto-length and break handling
- 133.1.3.3 Articulations (staccato, accent, tenuto, marcato, stress)
- 133.1.3.4 Bowings (up/down, natural harmonics, col legno, sul tasto)
- 133.1.3.5 Wind articulations (flutter, overblown, multiphonics)
- 133.1.3.6 Ornaments (trill, mordent, turn, slide, glissando)
- 133.1.3.7 Custom symbols via SMuFL extension
- 133.1.3.8 Arpeggiation symbols with direction
- 133.1.3.9 Pedal markings (piano, continuous)
- 133.1.3.10 Fermata, breath marks, caesura
133.1.4 Lyrics, Text, Chord Symbols#
- 133.1.4.1 Multi-verse lyric model with syllable hyphenation
- 133.1.4.2 Melisma extensions and centered-between-notes layout
- 133.1.4.3 Multi-language lyric support (RTL, CJK, IPA)
- 133.1.4.4 Tempo text with metronome (text + symbolic)
- 133.1.4.5 Expressive text (espr., dolce, leggiero)
- 133.1.4.6 Chord symbol engine with jazz/pop nomenclature
- 133.1.4.7 Chord-diagram grids (guitar, ukulele, banjo, mandolin)
- 133.1.4.8 Nashville-number notation
- 133.1.4.9 Solfege / movable-do annotation
- 133.1.4.10 Annotations with anchor and leader line
133.1.5 Parts, Scores, Layouts#
- 133.1.5.1 Full-score layout with instrument ordering presets
- 133.1.5.2 Parts extraction with rest-consolidation and cue-in
- 133.1.5.3 Lead-sheet layout
- 133.1.5.4 Piano-reduction generation
- 133.1.5.5 Vocal-score generation with piano reduction
- 133.1.5.6 Alternate layout variants (A4/Letter/A3, portrait/landscape)
- 133.1.5.7 Two-staff vs grand-staff keyboard layout
- 133.1.5.8 Drum set layout with custom staff positions
- 133.1.5.9 Guitar tab + notation paired staff
- 133.1.5.10 Multi-metric (polymetric) score layout
133.1.6 Playback & Export#
- 133.1.6.1 NotePerformer-style expressive playback engine
- 133.1.6.2 Articulation-to-keyswitch / velocity-layer mapping
- 133.1.6.3 Trill/ornament realization with humanization
- 133.1.6.4 Swing playback for jazz notation
- 133.1.6.5 Tempo curves with ritardando/accelerando
- 133.1.6.6 Dynamic envelope from hairpins
- 133.1.6.7 MusicXML round-trip (import + export, Dorico-compatible)
- 133.1.6.8 MEI (Music Encoding Initiative) export
- 133.1.6.9 MNX (emerging) export
- 133.1.6.10 LilyPond export (for typesetting interchange)
- 133.1.6.11 PDF/A export with embedded fonts
- 133.1.6.12 SVG export (per-page and per-system)
- 133.1.6.13 PNG / EPS export for publication
- 133.1.6.14 Braille music notation export (BMML)
- 133.1.6.15 Audio render (offline) with articulation map
133.2 @neith/session-view — Ableton-Class Clip Launcher & Session View#
Non-linear clip-launching surface for live electronic music production.
133.2.1 Session Grid#
- 133.2.1.1 Implement scenes × tracks grid model
- 133.2.1.2 Create clip slot with audio or MIDI content
- 133.2.1.3 Implement clip launching with global quantize (1/4, 1/2, 1, bar, 2bars)
- 133.2.1.4 Create scene launching (launch all clips in row)
- 133.2.1.5 Implement clip recording into empty slot
- 133.2.1.6 Create clip duplication with variants
- 133.2.1.7 Implement clip envelope per parameter (per-clip automation)
- 133.2.1.8 Create clip properties (loop start/end, gain, transpose)
- 133.2.1.9 Implement warp modes (Beats, Tones, Texture, Re-Pitch, Complex)
- 133.2.1.10 Create transient detection with warp-marker editor
- 133.2.1.11 Implement follow actions (next, previous, random, other, stop)
- 133.2.1.12 Create clip linking / group launch
- 133.2.1.13 Implement legato mode between clips (phase-preserve)
- 133.2.1.14 Create session-to-arrangement capture
- 133.2.1.15 Implement arrangement-to-session consolidate
133.2.2 Performance Features#
- 133.2.2.1 Live looper with overdub and reverse
- 133.2.2.2 Real-time time-stretch on input
- 133.2.2.3 Ableton Link integration (cross-device tempo sync)
- 133.2.2.4 Push/Launchpad surface-first workflow
- 133.2.2.5 Scene marker follow with color and name
- 133.2.2.6 Clip trigger from MIDI / OSC / UDP
- 133.2.2.7 Trigger recording (capture what you just played)
- 133.2.2.8 Session-view racks (drum, instrument, audio-effect)
- 133.2.2.9 Macro knobs with multi-parameter mapping
- 133.2.2.10 Crossfader A/B routing
133.3 @neith/restoration — iZotope RX-Class Dialogue & Audio Restoration#
Offline and real-time restoration suite for post-production and archival work.
133.3.1 Noise & Artifact Removal#
- 133.3.1.1 Spectral de-noiser with learn-from-sample profile
- 133.3.1.2 De-click / crackle (LP/78 restoration)
- 133.3.1.3 De-hum (50/60 Hz + harmonics, adaptive)
- 133.3.1.4 De-reverb (dereverberation via NMF and neural models)
- 133.3.1.5 De-wind (low-frequency gust suppression)
- 133.3.1.6 De-ess (sibilance, wideband and split-band)
- 133.3.1.7 Mouth-click removal with phase-aware inpainting
- 133.3.1.8 Breath control / removal
- 133.3.1.9 Plosive (P-pop) suppressor
- 133.3.1.10 Guitar amp hum / single-coil buzz removal
- 133.3.1.11 Tape hiss reduction with calibrated reference curves
- 133.3.1.12 Broadband noise neural network model
- 133.3.1.13 Intermittent-noise detector (chair creak, door)
- 133.3.1.14 Spectral repair inpainting (paint-out → regenerate)
- 133.3.1.15 Ambience match / room-tone generator
- 133.3.1.16 Multi-mic phase-align analyzer with cross-correlation and sub-sample delay estimation
- 133.3.1.17 Automatic polarity-flip suggester for paired-mic sources with confidence score
- 133.3.1.18 Comb-filter detector with frequency-domain cancellation heatmap for dialogue overlap
133.3.2 Speech & Dialogue#
- 133.3.2.1 Dialogue isolate (voice from mixed source)
- 133.3.2.2 Dialogue contour / level matcher (cross-cut consistency)
- 133.3.2.3 Auto-edit (remove gaps / overlaps) for dialogue
- 133.3.2.4 Mouth-movement to phoneme alignment for ADR sync
- 133.3.2.5 Voice activity detection with endpoint tagging
- 133.3.2.6 Deep filter for low-SNR dialogue recovery
- 133.3.2.7 Codec-artifact inverse for legacy AAC/MP3 source
- 133.3.2.8 Clipping repair with waveform reconstruction
- 133.3.2.9 Phase-reversed-mic noise cancellation
- 133.3.2.10 Audio-to-text alignment for caption generation
133.3.3 Spectral Editing#
- 133.3.3.1 Interactive time-frequency canvas
- 133.3.3.2 Magic wand frequency-region selection
- 133.3.3.3 Spectral brush with frequency-aware falloff
- 133.3.3.4 Inpainting via spectral diffusion model
- 133.3.3.5 Harmonic selection with fundamental tracking
- 133.3.3.6 Content-aware gap filling
- 133.3.3.7 Spectral copy/paste with phase preservation
- 133.3.3.8 Layer mode for non-destructive edits
- 133.3.3.9 FFT window/overlap user controls
- 133.3.3.10 GPU-accelerated rendering
133.3.4 Diffusion & Multi-Speaker SOTA#
Closes gap vs iZotope RX 12 (Dialogue Isolate, Streaming De-rustle, Repair Assistant v4), Accentize dxRevive Pro, Adobe Enhance Speech v2, NVIDIA Maxine.
- 133.3.4.1 Diffusion-based Dialogue Isolate for multi-speaker stem separation
- 133.3.4.2 Streaming De-rustle real-time clothing/handling-noise removal (<20 ms)
- 133.3.4.3 Multi-speaker-aware Repair Assistant with per-speaker noise profiles
- 133.3.4.4 Narrowband (8/16 kHz) to full-band (48 kHz) bandwidth extension via DNN
- 133.3.4.5 Background-strength slider with music-bed retention (not total vocal isolate)
133.4 @neith/pitch — Melodyne-Class Polyphonic Pitch & Time#
Polyphonic note-level pitch and time editing (DNA Direct Note Access class).
- 133.4.1 Polyphonic note detection (multi-pitch estimation)
- 133.4.2 Formant-preserving pitch shift
- 133.4.3 Per-note pitch correction with tolerance curves
- 133.4.4 Per-note timing edit
- 133.4.5 Per-note amplitude edit
- 133.4.6 Per-note vibrato / tremolo sculpt
- 133.4.7 Per-note formant edit
- 133.4.8 Scale / mode quantize with custom scales
- 133.4.9 Macro pitch-drift correction
- 133.4.10 Chord analysis with inversion labels
- 133.4.11 Audio-to-MIDI extraction (polyphonic)
- 133.4.12 Cross-track harmonic analysis
- 133.4.13 Neural transcription (piano, guitar, ensemble)
- 133.4.14 Pitch contour editor (bezier over notes)
- 133.4.15 ARA2 integration for DAW round-trip
133.5 @neith/mastering — Ozone-Class Mastering Suite & Neutron-Class Mix Assistant#
AI-assisted and reference-based mastering and mix analysis.
133.5.1 Mastering Modules#
- 133.5.1.1 Dynamic EQ with band solo and sidechain
- 133.5.1.2 Linear-phase EQ with per-band latency toggle
- 133.5.1.3 Multiband compressor (4/8 bands) with look-ahead
- 133.5.1.4 Brickwall limiter with ISP oversampling and IRC modes
- 133.5.1.5 Exciter / harmonic enhancer (tube, tape, warm, retro)
- 133.5.1.6 Imager (multiband stereo width + elliptical EQ)
- 133.5.1.7 Maximizer with multiple saturation models
- 133.5.1.8 Vintage-tape emulation with wow/flutter/hiss
- 133.5.1.9 Vinyl emulation with RIAA curve and groove noise
- 133.5.1.10 De-esser and sibilance control for mastering
133.5.2 Assistant & Reference#
- 133.5.2.1 Master assistant (genre-aware target curves)
- 133.5.2.2 Reference track compare with tonal balance overlay
- 133.5.2.3 Automatic match-EQ against reference
- 133.5.2.4 Loudness / dynamics match
- 133.5.2.5 Mix assistant (track-aware chain suggestion)
- 133.5.2.6 Vocal / drum / bass / guitar mix balance coach
- 133.5.2.7 Masking meter per instrument group
- 133.5.2.8 Low-end mono-compatibility meter
- 133.5.2.9 Tonal-balance control (static and adaptive targets)
- 133.5.2.10 Target-loudness verifier for streaming platforms
133.5.3 Stem-Focus & Diffusion Mastering SOTA#
Closes gap vs Ozone 12 Stem Focus, Neutron 5 Unmask 2.0, Melodyne 6 Leveling Macro + Chord Track, LANDR Studio AI stem mastering.
- 133.5.3.1 Stem-Focus mastering: separate master into 4 stems, process per-stem, recombine with target loudness
- 133.5.3.2 Diffusion-based master-chain proposer from reference track
- 133.5.3.3 Cross-track Unmask (multi-track frequency-unmasking recommender)
- 133.5.3.4 RMS/LUFS-driven Leveling Macro with automated gain-ride curves
- 133.5.3.5 Streaming-platform LUFS preset library (Spotify/Apple/YouTube/TIDAL/Amazon HD/Club/Broadcast/Podcast/Atmos LKFS)
133.6 @neith/modular — VCV Rack-Class Modular Synth Environment#
Sovereign modular environment with Eurorack-pattern modules.
133.6.1 Rack Core#
- 133.6.1.1 Rack canvas with HP (hp) width modeling
- 133.6.1.2 Module browser with favorites/recents
- 133.6.1.3 Patch cable rendering with physics slack
- 133.6.1.4 Polyphonic cables (up to 16 voices)
- 133.6.1.5 Module engine with plugin API (Rust trait + WASM)
- 133.6.1.6 CV at ±10 V standard with clip/overdrive
- 133.6.1.7 Gate/trigger signal level (0/5 V)
- 133.6.1.8 Sample-accurate polyphonic engine
- 133.6.1.9 Patch storage with preset variants
- 133.6.1.10 Integration as plugin inside Composer DAW (AU/VST3/CLAP)
133.6.2 Module Library (First-Party Sovereign Modules)#
- 133.6.2.1 VCO (saw, pulse, triangle, sine, with PWM)
- 133.6.2.2 Wavetable oscillator with morph
- 133.6.2.3 FM/PM operator
- 133.6.2.4 Granular oscillator
- 133.6.2.5 Additive oscillator (up to 1024 partials)
- 133.6.2.6 Noise source (white, pink, blue, violet)
- 133.6.2.7 VCF (4-pole LP, 2-pole MS-20, SVF, Moog ladder)
- 133.6.2.8 VCA (linear and exponential)
- 133.6.2.9 ADSR and ADSR+ envelope
- 133.6.2.10 LFO (multi-shape, BPM sync)
- 133.6.2.11 Sample & hold, track & hold
- 133.6.2.12 Sequencer (8/16/32 step)
- 133.6.2.13 Euclidean sequencer
- 133.6.2.14 Turing machine (random-walk sequencer)
- 133.6.2.15 Clock divider / multiplier
- 133.6.2.16 Mixer (linear and matrix)
- 133.6.2.17 Reverb / delay utility modules
- 133.6.2.18 Scope / spectrogram module
- 133.6.2.19 CV recorder
- 133.6.2.20 MIDI/CV interface
133.6.4 Frontier Modular & Live-Code SOTA#
Closes gap vs VCV Rack 2.6 (CLAP host, ARM native), Cherry Audio Voltage Modular 2.5, Softube Modular 3, Bitwig 5.2 Grid, Max 9 RNBO++, Pd Vanilla 0.55, SuperCollider 3.14, Sonic Pi 5.0, TidalCycles 1.9 / Strudel.
- 133.6.4.1 VCV Rack 2.6-class CLAP hosting inside rack + ARM-native inference path
- 133.6.4.2 Bitwig 5.2 Grid poly-grid module parity (per-voice modulation, vectorized audio graph)
- 133.6.4.3 Max 9 RNBO++ compile-to-native + jit.gl.pbr audio-reactive GPU visual export
- 133.6.4.4 Strudel (browser TidalCycles) live-coding kernel embedded in rack with shared clock
- 133.6.4.5 SuperCollider 3.14 + Sonic Pi 5.0 runtime bridge with network sclang control
133.7 @neith/patch-lab — Max/MSP / Pure Data-Class Visual Patching#
Visual dataflow patching for experimental audio/video/control systems.
- 133.7.1 Patcher canvas with box/cord model
- 133.7.2 Object library (math, logic, timing, message, list)
- 133.7.3 MSP-style DSP blocks (signal boxes vs control boxes)
- 133.7.4 Jitter-style video / matrix blocks for AV work
- 133.7.5 Abstraction / subpatch hierarchy
- 133.7.6 Polyphonic voice management (poly~ equivalent)
- 133.7.7 Bang / int / float / list / symbol / signal types
- 133.7.8 Snapshots and preset interpolation
- 133.7.9 Pattr / state binding system
- 133.7.10 Gen-style code-compilation to optimized DSP (like Gen~)
- 133.7.11 WASM export for web runtime
- 133.7.12 Embeddable as plugin in Composer
- 133.7.13 Live-editable while audio runs
- 133.7.14 Pd compatibility importer (best-effort)
- 133.7.15 External object SDK
133.8 @neith/live-code — SuperCollider / TidalCycles / Sonic Pi-Class Live Coding#
Text-based live-coding environment for algorithmic music.
- 133.8.1 Embedded Lua / Rust / TypeScript live-coding runtime
- 133.8.2 Pattern DSL (Tidal-style, cyclic pattern notation)
- 133.8.3 Strudel / FoxDot-compatible pattern library
- 133.8.4 SuperCollider SCLang-compatible mode
- 133.8.5 Csound orchestra/score integration
- 133.8.6 Faust DSP language JIT compile to SIMD / WASM
- 133.8.7 ChucK-style strongly-timed concurrency model
- 133.8.8 Live re-evaluation with crossfaded transition
- 133.8.9 Syntax highlighting and error overlay during performance
- 133.8.10 Snippet library and sharing
- 133.8.11 Session recorder (code + audio)
- 133.8.12 Multiplayer co-coding (CRDT)
- 133.8.13 MIDI clock / Ableton Link sync
- 133.8.14 Integration with modular (133.6) for hybrid patches
- 133.8.15 Visual feedback waveform overlaid on code
133.9 @neith/synth-advanced — Advanced Synthesis Engines#
Beyond the subtractive/FM/wavetable/granular already in 44.3.4.
- 133.9.1 Additive synthesis engine (up to 1024 partials with per-partial envelopes)
- 133.9.2 Spectral synthesis with partial editing
- 133.9.3 Modal synthesis (resonator banks)
- 133.9.4 Physical modeling: plucked string (Karplus-Strong extended)
- 133.9.5 Physical modeling: bowed string
- 133.9.6 Physical modeling: woodwind (tube + reed)
- 133.9.7 Physical modeling: brass (lip + bore)
- 133.9.8 Physical modeling: piano (hammer-string-soundboard with damper pedal)
- 133.9.9 Physical modeling: drum membrane (finite element)
- 133.9.10 Physical modeling: bell / metallophone
- 133.9.11 Neural synth (DDSP-style differentiable)
- 133.9.12 RAVE-style autoencoder synthesis
- 133.9.13 Scanned synthesis
- 133.9.14 Waveguide mesh for 2D/3D acoustic spaces
- 133.9.15 Vector synthesis (4-corner blending)
133.10 @neith/scoring — Scoring-to-Picture & Film Workflow#
Film/TV/game scoring tools beyond generic DAW + video sync.
- 133.10.1 Video window with frame-accurate scrub
- 133.10.2 Spotting notes timeline (director cues linked to picture)
- 133.10.3 Streamers (moving vertical lines over video)
- 133.10.4 Punches (flash frames)
- 133.10.5 Click-track generator with tempo map
- 133.10.6 Hit-point marker with tempo / meter snapping
- 133.10.7 Cue list with durations and start/end timecode
- 133.10.8 Cue sheet auto-generation (PRO formats)
- 133.10.9 Conductor-score variant layout
- 133.10.10 Streamers / punches delivered as export layer
- 133.10.11 Dolby Atmos renderer direct feed
- 133.10.12 Stems export per cue with naming convention
- 133.10.13 OMF / AAF export for picture round-trip
- 133.10.14 Score marker sync to OTIO for editorial
- 133.10.15 Music editor (cue edit between tempo and picture changes)
133.11 @neith/adr-foley — ADR, Foley, and Production Sound#
Specialised post-production dialogue and foley rooms.
133.11.1 ADR#
- 133.11.1.1 ADR cue list with line text, actor, character
- 133.11.1.2 Streamers/beeps configuration per cue
- 133.11.1.3 Auto-pull of original production track
- 133.11.1.4 Take management with slate + take number
- 133.11.1.5 Loop playback around cue with configurable pre/post roll
- 133.11.1.6 Remote ADR over network (low-latency NDI + JamKazam-class)
- 133.11.1.7 Matching EQ / reverb to production track
- 133.11.1.8 Auto-alignment of ADR to production track (VocAlign-class)
- 133.11.1.9 Actor talent tracking and logging
- 133.11.1.10 ADR session report export
133.11.2 Foley#
- 133.11.2.1 Foley session template with footstep/movement/prop lanes
- 133.11.2.2 Floor mic / boom mic multi-track routing
- 133.11.2.3 Foley library browser with auto-audition
- 133.11.2.4 Sync-clip-to-event alignment
- 133.11.2.5 Mic-perspective matching per scene
- 133.11.2.6 Walla / crowd recording
- 133.11.2.7 Prop library metadata (surface, material, weight)
- 133.11.2.8 Foley pass: footsteps
- 133.11.2.9 Foley pass: cloth
- 133.11.2.10 Foley pass: props
133.12 @neith/dj — DJ Decks & Stem Separation#
Serato / Traktor / Rekordbox-class performance decks.
- 133.12.1 Two/four-deck UI with waveform overview
- 133.12.2 Beat-grid detection and editor
- 133.12.3 Key detection (Camelot wheel) and harmonic mixing assist
- 133.12.4 Master tempo (key-lock) time-stretch
- 133.12.5 Hot cue bank with color labels
- 133.12.6 Loop (manual, auto 1/4–32 bar)
- 133.12.7 Slicer mode
- 133.12.8 Crossfader with curve selection
- 133.12.9 Real-time stem separation (drums / bass / vocals / melody) — Demucs/Spleeter sovereign model
- 133.12.10 Per-stem mute/solo during performance
- 133.12.11 Effects rack (flanger, gate, echo, filter)
- 133.12.12 Sampler pad with drop samples
- 133.12.13 Library with crates, playlists, smart lists
- 133.12.14 Rekordbox / Serato / Traktor import compatibility
- 133.12.15 External controller support (CDJ, XDJ, S-series, DDJ)
- 133.12.16 Neural Mix 3-class real-time 6-stem (vocals/drums/bass/harmonic/melodic/FX) <10 ms on-device
- 133.12.17 Automix AI with crowd-energy curve targeting and beat-locked transitions
- 133.12.18 Per-stem FX chain (EQ/filter/echo/reverb applied only to one stem)
- 133.12.19 Apple Vision Pro / spatial UI performance surface
- 133.12.20 CloudDirectPlay streaming from Dropbox / Google Drive / S3 with offline cache
133.13 @neith/sample-content — Sovereign Sample Library Strategy#
Curated, owned-or-licensed sample content to remove dependency on Spitfire, Vienna, East West, NI Komplete.
133.13.1 Capture Pipeline#
- 133.13.1.1 Recording session spec template (mic positions, round-robins, articulations)
- 133.13.1.2 Scripted capture runner (tempo/articulation grid)
- 133.13.1.3 Auto-labelling of captured takes (pitch, dynamic, articulation)
- 133.13.1.4 Neural alignment of round-robin takes
- 133.13.1.5 Loop-point detection for sustained notes
- 133.13.1.6 Release-tail extraction with gate envelope
- 133.13.1.7 Tuning correction per sample to exact semitone
- 133.13.1.8 Velocity-layer normalization
- 133.13.1.9 Sample compression (FLAC, SFZ opus mode)
- 133.13.1.10 Sample provenance / rights ledger
133.13.2 Library Definition & Packaging#
- 133.13.2.1 Open sample-library format spec (oshun-sfz+)
- 133.13.2.2 SFZ import/export compatibility
- 133.13.2.3 DecentSampler import compatibility
- 133.13.2.4 Kontakt NKI read-only importer (for user-owned libraries)
- 133.13.2.5 EXS24 / Ableton Sampler importer
- 133.13.2.6 Library signing and tamper detection
- 133.13.2.7 Library licensing and entitlement system
- 133.13.2.8 Library download / delta-update client
- 133.13.2.9 Library CDN distribution
- 133.13.2.10 Library marketplace UI
133.13.3 Core Sovereign Libraries (Content)#
- 133.13.3.1 Orchestral strings (full sections + solo)
- 133.13.3.2 Orchestral woodwinds
- 133.13.3.3 Orchestral brass
- 133.13.3.4 Orchestral percussion (pitched + unpitched)
- 133.13.3.5 Concert piano (multi-mic, multi-dynamic)
- 133.13.3.6 World percussion (dunun, djembe, tabla, frame drums, shekere)
- 133.13.3.7 World plucked (kora, oud, sitar, mbira, koto, guzheng)
- 133.13.3.8 Vocal ensemble library (SATB, gospel, chamber)
- 133.13.3.9 Drum-kit / hybrid kit library
- 133.13.3.10 Electronic production library (synth multisamples, drum one-shots, FX)
133.13.4 AI Sample Generation & Browser SOTA#
Closes gap vs Splice Create 2 + AI Stacks + Bridge, Output Arcade 4 + Co-Producer + Pack Generator, Loopcloud 9, Samplab 3 Texture Transform, Sononym 3 neural browser, Audialab Emergent Drums 3.
- 133.13.4.1 Text-to-loop generation (prompt + key/BPM lock), key/BPM-auto-matched to active session
- 133.13.4.2 AI Stacks: layered loop-bed generator keyed/BPM-matched to current project
- 133.13.4.3 Pack Generator: cohesive 20-loop packs with consistent palette/key/BPM
- 133.13.4.4 Neural multi-modal sample browser (audio + text-prompt similarity search, clustering)
- 133.13.4.5 Texture Transform polyphonic pitch/time + timbre morphing between samples
133.14 @neith/comping — Comping, Track Stacks, Folder Tracks#
Large-session management tools missing from Phase 44.3.
- 133.14.1 Take-folder data model with parent clip
- 133.14.2 Quick-swipe comping UI
- 133.14.3 Playlist (alternate takes) per track
- 133.14.4 Automatic best-take AI suggestion
- 133.14.5 Track stack (summing + VCA)
- 133.14.6 Folder track with collapse and solo/mute inheritance
- 133.14.7 Group-edit mode for multi-track editing
- 133.14.8 Coalesce / flatten comped result
- 133.14.9 Comp from multiple source tracks (multi-mic comp)
- 133.14.10 Comped region with per-take crossfade
133.15 @neith/realtime-assist — Real-Time AI Mixing / Mastering Assist#
Integrates with Phase 47 (Nous) to offer on-device AI collaboration.
- 133.15.1 Real-time mix critique (balance, EQ clashes)
- 133.15.2 Suggest-and-apply EQ with undo
- 133.15.3 Automatic gain-staging
- 133.15.4 Automatic panning suggestion
- 133.15.5 Auto-chain builder (genre-aware)
- 133.15.6 Style match (apply reference production style)
- 133.15.7 AI vocal-tuning with taste slider
- 133.15.8 AI de-masking (reduce inter-track masking automatically)
- 133.15.9 AI reverb-sends calibration
- 133.15.10 Narration of changes for learning/reproducibility
133.16 @neith/foley — Generative Foley & Video-Conditioned SFX#
Generative Foley suite for film/TV/game post-production, replacing or augmenting library-based Foley workflows. Closes gap vs ElevenLabs SFX v2, Stable Audio 2.5 SFX, Meta MovieGen Audio, Google Veo 3 audio, OpenAI Sora 2 audio, FoleyCrafter 2, Foley Flex, Sony Foley AI, Ubisoft La Forge Granular Foley, EA SEED Audio Palette.
- 133.16.1 Text-to-SFX generator (30s clips, loopable, reference-audio conditioning, stem export)
- 133.16.2 Video-conditioned SFX synthesis with joint audio-video diffusion (MovieGen / Veo 3 / Sora 2 class)
- 133.16.3 Onset-aware Foley conditioning with temporal CLAP alignment (FoleyCrafter 2 / Foley Flex class)
- 133.16.4 Frame-accurate Foley library retrieval + generative blending (Sony Foley AI class) with timecode sync
- 133.16.5 Physics-parameter-conditioned neural impact synthesis (mass, material, velocity) — Ubisoft Scarlet class
- 133.16.6 Real-time streaming Foley replacement (<200ms Moshi-Audio class) for live ADR sessions
- 133.16.7 V2V Foley workflow: drop-in video → synchronized impact/footstep/cloth/ambience stems
- 133.16.8 Granular-Foley diffusion footstep synthesis with material-aware variation (Ubisoft La Forge AES 2025)
- 133.16.9 RVQ-token SFX library + small LM for real-time variation (EA SEED Audio Palette class)
- 133.16.10 48kHz studio-tier output with Dolby Atmos bed + object export integration
Phase 133 Summary — Composer DAW SOTA Expansion#
| Library | Section | Tasks |
|---|---|---|
| @neith/notation | 133.1 | 75 |
| @neith/session-view | 133.2 | 25 |
| @neith/restoration | 133.3 | 43 |
| @neith/pitch | 133.4 | 15 |
| @neith/mastering | 133.5 | 25 |
| @neith/modular | 133.6 | 35 |
| @neith/patch-lab | 133.7 | 15 |
| @neith/live-code | 133.8 | 15 |
| @neith/synth-advanced | 133.9 | 15 |
| @neith/scoring | 133.10 | 15 |
| @neith/adr-foley | 133.11 | 20 |
| @neith/dj | 133.12 | 20 |
| @neith/sample-content | 133.13 | 35 |
| @neith/comping | 133.14 | 10 |
| @neith/realtime-assist | 133.15 | 10 |
| @neith/foley | 133.16 | 10 |
| Phase 133 Total | 383 |