ResonaMatrix
Phase 4 Hybrid Neural VAD & Connected DB
FPS:
60
Audio DSP:
5.1%
BLAKE3 Blocks:
0
ASR Tokens in DB:
0
DB Sync:
ACTIVE
๐ Master Dashboard
๐ก๏ธ Liveness & Deepfakes
๐ฃ๏ธ Formants & Biomarkers
๐ฅ Speaker Diarization
๐ ASR & Token Alignment
๐ฎ Blind RIR Profiler
๐ข ISO 22955 Workplace
๐ CAD Blueprint & BOM
๐ฏ 360ยฐ Beamforming
๐ Permanent DB & BLAKE3
Start Live Microphone
dBA
dBC
dBZ
Wiener Denoiser: OFF
Run Pre-Flight Audit
๐งช Test Corpus:
๐ข Natural Human (Jitter 0.65% + F1/F2 Formants)
๐ด AI Deepfake (Jitter 0.00% + >8kHz Noise)
๐ฃ Modulated Voice (Out-of-Bounds IPA)
๐ก Physically Masked (N95 HF Tilt)
๐ด Biological Cough / Throat Clear
๐ก Biological Laughter (5Hz Pulse)
โถ Inject Vector
VAD: Idle
Cal:
+94.0 dB
๐ก
Phase 4 Active: Hybrid Neural VAD is streaming directly into the IndexedDB BLAKE3 tamper-evident ledger.
Sound Level
--.-
dBA Fast
Voice Level (L
speech
)
--.-
Gated RMS
Noise Floor (L
noise
)
--.-
NC -- Rating
Speech SNR
-- dB
SNR
voice
Passive STI (STI
p
)
--
ANSI SII:
--%
Neural VAD Prob
0.00
P(Speech) Recurrent
1/1 Octave Spectrum & NC Rating Curves
NC --
ISO 12913 Circumplex Map (Soundscape)
Evaluating...
Authenticity Score
100 / 100
NATURAL HUMAN
Micro-Jitter Floor
0.45 %
> 0.12% Human OK
Replay THD (2f0, 3f0)
0.8 %
Direct Air Mouth
Mask Low-Pass Tilt
-6.2 dB/oct
Unmasked Phonation
Active Articulatory Unit
Silent / Idle
Phonetic Granularity
Biological Event
None
Laughter/Cough Gater
Micro-Jitter Scope (S23)
Threshold < 0.12%
IPA Vowel Convex Polygon Validator (S24)
Human Vocal Tract Bounds
F1 (Jaw Height)
-- Hz
Vowel Height
F2 (Tongue Front)
-- Hz
Vowel Frontness
Local Jitter
-- %
< 1.0% Normal
Local Shimmer
-- %
< 3.5% Normal
Harmonics-to-Noise
-- dB
> 18 dB Healthy
Vocal Fatigue (VFI)
-- / 100
Cord Strain
IPA Vowel Formant Space (F1 vs F2)
Active: --
14th-Order LPC Resonance Envelope
0 - 5000 Hz
Speaker A (Cyan)
-- %
0.0s
Speaker B (Purple)
-- %
0.0s
Speaker C (Yellow)
-- %
0.0s
Speaker D (Pink)
-- %
0.0s
Turn Velocity
0 /min
Handover Rate
Cross-Talk Overlap
0
Collision Gater
30-Second Multi-Lane Diarization Timeline
Active: Listening...
Live Diarized Transcript Stream & DB Sync
Export JSON
Transcription engine active. Every phrase is committed to IndexedDB with physical acoustic telemetry.
Token-Level Acoustic Clarity Scope
STI_word Stream
Acoustic Root-Cause Diagnoser
Speech signal is clean. High acoustic SNR ensures zero transmission masking.
Passive RT60 (Blind)
-- s
Ensemble Averaged
Passive EDT
-- s
Initial Decay Slope
Passive Clarity (C50)
-- dB
Direct / Reverb Ratio
Decay Samples
0 pcs
Phoneme Offsets
Convergence Confidence
0 %
Stability Filter
Natural Phoneme Decay Tail & Schroeder Fit
Phoneme RIR
Passive RT60 Convergence Time-Series
History
Conversational Distraction (CDI)
-- %
Focus Loss Score
ISO 22955 Compliance
-- %
Typology Target
Distraction Radius (rD)
-- m
STI > 0.50 Threshold
Privacy Radius (rP)
-- m
STI < 0.20 Threshold
2D Workplace Acoustic Floor & Distraction Heatmap
ISO 22955 Layout
Absorption Needed (ΔA)
-- mยฒ
Inverse Sabine
50mm Wall Panels
-- pcs
NRC 0.90
Ceiling Clouds
-- pcs
NRC 0.95
Corner Bass Traps
4 pcs
< 150 Hz Modes
Projected STI Gain
+0.22
Post-Treatment
2D CAD Architectural Treatment Blueprint
Ray Paths
Material Bill of Materials (BOM)
Export BOM
Material
Qty
Area
50mm Panels
14 pcs
10.1 mยฒ
Ceiling Clouds
6 pcs
4.3 mยฒ
Bass Traps
4 pcs
Corners
360ยฐ Polar Spatial Acoustic Radar (GCC-PHAT AoA)
AoA: 0ยฐ
๐ก๏ธ Audit & Verify Entire Ledger
Download Signed .CoC.JSON
Download Forensic CSV
Purge Database
Audit Status: Ready
Multi-Store BLAKE3 Telemetry & Transcript Ledger (Last 50 Records)
IndexedDB Persisted
Block
UTC Time
BLAKE3 Hash
SPL
SNR
STIp
P(VAD)
Auth
No blocks recorded in BLAKE3 ledger yet.
๐ Pre-Flight Health Audit
BLAKE3 Secured
Auditing...
Proceed