Mimir

Realtime cameras, microphones, chirplets, and field evidence organized into one coherent stream machine.

Bioacoustic Timeline Watermark

8 min read

Bioacoustic Timeline Watermark

Purpose

The active timing signal is no longer supposed to be a row of tidy lab beeps. That shape proved the math, but it is the wrong animal for a room.

Mimir now treats active acoustic sync as a low-gain bioacoustic language: short birdsong-like motifs whose identity survives through contour, log-frequency position, spectral envelope, rhythm, and speaker-specific timbre. The receiver should not ask only “which bin fired?” It should ask:

which motif contour did I hear?
which self-identifying song word did that contour encode?
which speaker vocabulary emitted it?
which time and frequency anchors did that one call expose?
what delay and clock fit make the heard anchors coherent?
what did this room/mic path do to the spectrum?

Current Machine

MimirBioacousticTimeline is the current active runtime watermark.

  • 128 self-identifying word positions, each with a left-speaker and right-speaker variant.
  • Each word contains four short syllables.
  • Each syllable carries a log-frequency glide plus formant-like partials.
  • Motif rhythm, contour, syllable duration, root, and speaker tint vary by word.
  • No de Bruijn grammar is used in the active bioacoustic path. One heard song word identifies its canonical event directly inside the current operating horizon.
  • The decoder emits motif candidates, direct song-word anchors, a source clock fit, and per-band response evidence.
  • A song word is not one scalar chirp. It is a contour packet: multiple syllable onsets, bends, formants, rhythm offsets, payload ornaments, and speaker-colored spectral features all become anchors in log-mel space. One successful call can pin a cluster of time/frequency constraints at once.
  • The first anchor-rich canary packet explicitly shapes timing chips, formant pivots, harmonic-envelope notches, and payload ornaments. Calibration receipts now persist those intra-call anchors so path learning can discover which features survive each output/mic pair.
  • MimirComplexContourMatchedFilterBank is the first sonar-shaped receiver cut: it turns known contour anchors into complex matched-filter responses, keeps multiple lobes per anchor, and hands them to MimirDirectPathTracker. The tracker takes the current path hypothesis as authority, chooses the coherent direct-path cluster inside that gate, and records later clusters as reflection taps instead of letting them become alternate timelines.
  • Runtime emission sends the left vocabulary to the left speaker and the right vocabulary to the right speaker, so microphones can learn each side of the room separately.
  • The runtime active sync evidence kind is bioacoustic.

The old MimirChirpBinTimeline remains as a calibration/reference artifact with ASIO capture tooling. It is useful for comparing acoustic paths and existing calibration captures, but it is no longer the expressive runtime watermark target.

Ownership

MimirBioacousticTimeline owns:

  • song-word codebook shape;
  • deterministic phrase schedule;
  • PCM rendering for active emission;
  • streaming motif scoring;
  • direct contour-anchor decoding;
  • source clock fitting;
  • response-band evidence.

MimirAudioSynchronizationAnalyzer owns:

  • comparing loopback/reference and candidate receiver decodes;
  • choosing passive evidence before active evidence in hybrid;
  • reporting bioacoustic delay when active evidence is used;
  • fractional delay refinement from waveform evidence.

Fensalir audio only emits the rendered watermark. It does not interpret it. Faust/native DSP will eventually own the fractional-delay/SRO actuator.

Why This Shape

Birdsong survives ugly channels because the information is redundant:

  • contour survives when absolute pitch shifts;
  • rhythm survives when frequency response is bad;
  • formant envelope survives when one band dies;
  • speaker-specific variants expose room asymmetry;
  • repeated local structure gives the receiver more than one way to recover time response, and local payload identity.

That is the physical path Mimir needs. A phone mic, camera mic, Scarlett input, or network receiver should be able to recognize canonical time from a damaged short phrase, not from a perfect isolated tone.

Proofs

Synthetic checks:

dotnet run --project .\src\Mimir.BufferSmoke\Mimir.BufferSmoke.csproj -- --bioacoustic-self-test
dotnet run --project .\src\Mimir.BufferSmoke\Mimir.BufferSmoke.csproj -- --standalone-bioacoustic-self-test --sample-rate 48000 --delay-samples 1269.5
dotnet run --project .\src\Mimir.BufferSmoke\Mimir.BufferSmoke.csproj -- --chirp-only-sync-self-test --sample-rate 48000
dotnet run --project .\src\Mimir.BufferSmoke\Mimir.BufferSmoke.csproj -- --complex-contour-tracker-self-test --sample-rate 192000 --delay-samples 693.5 --reflection-delay-samples 116

Expected state:

  • --bioacoustic-self-test recovers direct word anchors across the canonical phrase.
  • --standalone-bioacoustic-self-test recovers delayed canonical source time from codebook/schedule state alone.
  • --chirp-only-sync-self-test reports evidence=bioacoustic and recovers fractional delay below printed microsecond precision.
  • --bioacoustic-cepstral-smoke round-trips the active call through a degraded mel-cepstral representation, then decodes word identity with an augmented projection-hash MFCC index instead of brute-forcing every word waveform.
  • --bioacoustic-train runs a hypothesis panel over decoder shapes and writes CultCache receipts plus pre-warp, post-warp, and reconstructed-from-detections WAV artifacts. See Bioacoustic Training Harness.
  • --complex-contour-tracker-self-test proves the first direct-path tracker shape with a louder delayed reflection present. The current 192 kHz synthetic smoke reports about 5.249 us error, so this is a useful tracker surface, not the final microsecond claim.

Render a raw Float32 preview:

dotnet run --project .\src\Mimir.BufferSmoke\Mimir.BufferSmoke.csproj -- --render-bioacoustic-f32 --output artifacts/asio/bioacoustic-f32.raw --seconds 8

Next Cut

The current physical receipt path uses the canary packet song as the first serious contour-anchor witness. The next coherent cut is to promote that shape into the live streaming log-mel receiver:

audio window
-> log-mel / constant-Q evidence surface
-> contour packet proposals
-> direct song-word identity over a tiny candidate set
-> intra-call time/frequency anchor extraction
-> global delay, clock, and path hypothesis
-> path response model

Do not add a pile of threshold rules to make one room capture look good. The right extension is learned path weighting over contour parts: which syllable onsets, bends, formants, payload ornaments, partials, and time gaps survive each output/mic path.

The latest training receipt says identity can survive the current mel-cepstral degradation panel, but timing collapses under warped domains. That points the next cut at a global delay/clock/path hypothesis over detected words, not a larger brute-force dictionary.

That cut has started in the harness. --bioacoustic-train persists a global clock hypothesis per result, and --calibrate-contestant-asio-f32 now persists the first physical packet-song response model. The current Scarlett canary receipt clears 10x realtime and proves that a single song word can own local identity; the remaining work is to make each syllable/formant contour emit enough stable anchors that physical mic residuals collapse from tens of microseconds toward the one-microsecond target. The first anchor-rich receipt improves loopback but not the physical mics, which means anchor observability is now present and anchor survivability is the next design problem. The complex contour tracker now runs against stored Scarlett ASIO captures with seeded path hypotheses: stored/fresh shotgun estimates land within about 5.860 us and 7.247 us of the existing path fit, while stored/fresh cardioid estimates land within about 15.817 us and 8.963 us. Confidence remains low because the current matched-filter bank exposes reflection taps but does not yet fit a full phase/group-delay channel surface. The tracker now emits per-band delay and phase residual observations plus a MimirDirectPathChannelModel correction surface. That model is deliberately not treated as live truth from one window; it exists so calibration/tracking can learn stable path corrections across time instead of overfitting the current reflection field. --complex-contour-replay-panel persists this measurement surface to calibration/bioacoustic/complex-contour-replay-panel.json for regression and path-learning work.

--learn-complex-contour-channel-model now turns that replay receipt into a path-scoped complex contour channel model at calibration/bioacoustic/complex-contour-channel-model.json. The model records stable per-band delay and phase corrections plus observed reflection tap spacing. --complex-contour-asio-f32 --channel-model ... can apply it explicitly, and --evaluate-complex-contour-channel-model writes calibration/bioacoustic/complex-contour-channel-model-evaluation.json.

The current model is useful but not yet trusted as silent runtime authority. Across the four stored/fresh Scarlett replay cases it improves 3/4 absolute path-seed errors and lowers mean direct-cluster residuals overall. The best fresh cardioid case reaches about 0.202 us from the seeded path fit, but the stored cardioid path still worsens from about 15.817 us to 17.476 us. That means the learned phase/group-delay surface is real evidence, while the Perfect Machine still needs more captures and a stronger direct-path hypothesis before calibration can be applied automatically.

The contour machine is no longer artifact-only. MimirSynchronizationHub can now run a live complex-contour lane over rolling Float32 audio buffers when enableComplexContourRuntime is true. In that mode Mimir emits the configured bioacousticWitnessProfileId through the same canary-packet renderer used by the contour tracker, loads complexContourChannelModelPath, reads ASIO-style rolling buffers directly, and publishes complex-contour reports alongside the existing audio sync reports. --complex-contour-runtime-self-test proves the surface synthetically at 192 kHz: a 693.5 sample delay is recovered at about 0.219 us error from runtime-shaped rolling buffers.

The first fresh live Scarlett run after this wiring is preserved at calibration/bioacoustic/complex-contour-live-20260525-063229.md. It rendered a new canary-packet-trill witness, played it through Focusrite ASIO, recorded all four 192 kHz inputs, and analyzed the capture with the current complex contour channel model. Loopback 1 to Loopback 2 measured -0.014 us, proving the interface clock path is effectively sample-locked. Physical mic paths survived the room: the shotgun path landed -3.984 us from its prior seed and the cardioid landed +1.547 us. After correcting the direct-path confidence scorer, loopback confidence is 0.971, cardioid confidence is 0.488, and the weaker/disconnected-looking path is 0.335. This is fresh meatspace evidence through the actual DAC/speaker/room/mic/ADC path, though final microsecond sync still needs independently measured path truth instead of comparison to prior seeds.

Correction after signal inspection: the apparent asio-ch0 cardioid estimate from that receipt is invalid. Its log-mel panel and channel statistics show a flat low-level signal with essentially zero correlation to loopback (0.000032). The tracker had enough seeded timing pressure to fit noise near the expected delay. Treat that receipt as a clean loopback/interface proof plus a weak/non-authoritative input witness, not a physical cardioid proof.

Backlinks