Tighten IPA source-golf scene voice contract

docs/ipa-vocal-tract-roadmap.md

@@ -748,6 +748,12 @@ movement, verification command, and rollback trigger in `source-change-plan.md`.
 The outer loop runs a focused build before scoring so malformed syntax or
 broken lowering fails before it pollutes the trial store.
 
+`hypotheses.md` is the hypothesis worker's report of record. The external
+agent's final chat transcript (`hypothesis-agent.md`) is only a summary and
+must not be used by the evaluator as a substitute for the 25-lane matrix,
+retrieval receipts, reference matrix, novelty gate, source-edit decision, and
+scene voice audit.
+
 ### Scene Voice Model
 
 IPA parity candidates are now scenes, not isolated waveforms. Every hypothesis
@@ -771,6 +777,11 @@ hide the failure?" It is "which voices are in the scene, which owner synthesizes
 each one, and which witness proves that the helper voice is not impersonating
 speech?"
 
+At least one scene/noise or recording-condition hypothesis family in a
+source-golf round must implement those helpers as real `.aqua` `voice`
+declarations, not prose. The primary `vocal network=...` voice remains present
+unless the family explicitly tests a coherent source/lowering replacement.
+
 Five prompt-optimization smokes on 2026-05-29 sharpened that contract. The best
 agent output came from a fixed receipt packet and a constrained reasoning pass,
 not from letting an external agent wander the store while thinking. Production

state/evidence.jsonl

@@ -245,3 +245,4 @@
 {"date":"2026-05-30","area":"ipa-trial-retrieval-tuning","lesson":"Semantic retrieval smokes showed raw vector similarity was useful but too permissive: class queries could leak through generic weak-articulation language, and mixed-set records could impersonate every class because their target set contains all phonemes. `IpaTrialWorker search` now adds phonetic class-affinity scoring, candidate-specific class ownership for mixed lanes, demotes mixed/generalization to a secondary transfer hint when a real class is present, and only tags dressing when the record itself mentions dressing/FM/AM/envelope/helper evidence.","evidence":"artifacts/parity/ipa-trial-worker-smoke/20260529T2204/retrieval-tuning-20260530-pass1; artifacts/parity/ipa-trial-worker-smoke/20260529T2204/retrieval-tuning-20260530-pass4; tools/IpaTrialWorker/Program.cs"}
 {"date":"2026-05-31","area":"ipa-trial-primitive-timeline-facts","lesson":"IPA trial evidence must carry primitive phase witnesses inside the CultCache record, not only as CSV artifacts or prose claims. `PrimitiveTimelineFactExtractor` turns `ProbeTimelineReport` samples into contact release, source flow, branch admittance, radiation output, path energy, and passivity facts; `IpaTrialWorker` indexes and reports those facts through search/show/dump so hypothesis agents can cite release/radiation/passivity evidence directly.","evidence":"src/AquaSynth.Core/PrimitiveTimelineFactExtractor.cs; src/AquaSynth.Core/IpaTrialResults.cs; tools/IpaTrialWorker/Program.cs; tests/AquaSynth.Dsl.Tests/IpaTrialOrchestratorTests.cs; artifacts/parity/ipa-timeline-fact-smoke/20260531T233217641"}
 {"date":"2026-05-31","area":"ipa-trial-source-golf-scene-voices","lesson":"IPA hypothesis agents need authority to golf the DSL/Faust machine when repeated evidence points at a missing owner, and every parity patch should model the reference as a scene with named voices. Source-edit-enabled rounds may now change parser/model/lowering/Faust/tests/docs with a source-change plan and pre-score build; candidate reports must separate primary articulation, tract noise, release/transient, background-room, and recording-condition voices. General synth primitives are allowed for scene/background modeling, but helper voices cannot impersonate articulation without lower-layer evidence.","evidence":"tools/run-ipa-trial-loop.ps1; docs/ipa-vocal-tract-roadmap.md; docs/audio-doctrine.md"}
+{"date":"2026-05-31","area":"ipa-trial-source-golf-smoke","lesson":"The first loose source-edit-enabled IPA round completed without tracked source edits and did not improve the aggregate metrics, which is valid negative evidence rather than a failure of authority. It exposed two harness contracts to tighten: evaluators must judge `hypotheses.md` as the report of record instead of the external agent's final `hypothesis-agent.md` chat summary, and scene/noise families must implement at least one actual helper/background `.aqua` `voice` declaration instead of satisfying scene modeling with prose.","evidence":"artifacts/parity/ipa-trial-loops/20260531T001831690; tools/run-ipa-trial-loop.ps1; docs/ipa-vocal-tract-roadmap.md; state/scratch.md"}

state/memory.json

@@ -92,7 +92,7 @@
       "Surge XT for modern hybrid synth modulation",
       "Cardinal/VCV-style patches for modular cable/rate semantics"
     ],
-    "rule": "For speech, add schema fields and learned-control lanes only when the parity harness has a loss surface that can exercise them. For reference synths, add DSL abstractions only when rebuilds prove the missing invariant. Parsing alone is inventory, not evidence of expressivity. Pink Trombone should be decomposed into reusable low-level Aqua primitives before audio golf, not copied as a bespoke clone or inflated into PT-shaped modules. After the 2026-05-29 proprioception correction, collapse graph-vocal lowering into a small set of instrumented physical primitives before more utterance golf. Gesture motion belongs in ControlSurface/ControlSpline timeline ownership, not primitive-local smoothing; cut any generated contact/port spline that invents motion not supported by primitive reference timelines. Public clients author gestures through ControlSurfaceCatalog plus ControlSplineTimeline and stream surface control dictionaries; higher-level gesture templates and phoneme_gesture are only spline emitters. IPA descriptors seed which anatomical surfaces move, while local datasets golf numeric spline defaults and patch voice code. Training reports must separate gesture_score, clean_vocal_score, and full_parity_score so full-patch recording-condition helpers cannot impersonate articulation. Frozen IPA gesture rounds own candidate scripts, primitive timelines, gesture metrics, manifests, evidence receipts, metric summaries, candidate clusters, and science briefs; `IpaTrialResult` CultCache records own rendered trial outcomes plus extracted primitive timeline facts for contact release, source flow, branch admittance, radiation output, path energy, and passivity. External Codex IPA exploration must query the `.cc` store through the worker's semantic `search`/`show` organ, backed by Qdrant/Ollama vector retrieval when available; Qdrant is a derived index over CultCache and local artifacts, not a second truth. Search reports expose class focus/match and use phonetic class-affinity plus timeline-fact text so agents do not treat generic weak-articulation language as cross-class evidence. Source-edit-enabled IPA rounds may golf DSL syntax, parser/model records, Faust lowering, worker evidence surfaces, and tests when repeated evidence identifies a missing owner; every source edit needs an owner map, invariant, expected metric movement, verification, and rollback trigger. IPA parity candidates should model scene voices explicitly: primary articulation, tract noise, release/transient, background-room, and recording-condition. General synth primitives are allowed for helper scene voices, but they cannot count as articulation without gesture, primitive timeline, and clean-vocal movement. Crude full-store dumps are audit fallback, not agent memory. Statistical science workers and audio parity stages consume those bundles without rewriting gesture evidence. If gesture scores rise but rendered audio is silent, fix the primitive source/tract owner before adding full-patch dressing."
+    "rule": "For speech, add schema fields and learned-control lanes only when the parity harness has a loss surface that can exercise them. For reference synths, add DSL abstractions only when rebuilds prove the missing invariant. Parsing alone is inventory, not evidence of expressivity. Pink Trombone should be decomposed into reusable low-level Aqua primitives before audio golf, not copied as a bespoke clone or inflated into PT-shaped modules. After the 2026-05-29 proprioception correction, collapse graph-vocal lowering into a small set of instrumented physical primitives before more utterance golf. Gesture motion belongs in ControlSurface/ControlSpline timeline ownership, not primitive-local smoothing; cut any generated contact/port spline that invents motion not supported by primitive reference timelines. Public clients author gestures through ControlSurfaceCatalog plus ControlSplineTimeline and stream surface control dictionaries; higher-level gesture templates and phoneme_gesture are only spline emitters. IPA descriptors seed which anatomical surfaces move, while local datasets golf numeric spline defaults and patch voice code. Training reports must separate gesture_score, clean_vocal_score, and full_parity_score so full-patch recording-condition helpers cannot impersonate articulation. Frozen IPA gesture rounds own candidate scripts, primitive timelines, gesture metrics, manifests, evidence receipts, metric summaries, candidate clusters, and science briefs; `IpaTrialResult` CultCache records own rendered trial outcomes plus extracted primitive timeline facts for contact release, source flow, branch admittance, radiation output, path energy, and passivity. External Codex IPA exploration must query the `.cc` store through the worker's semantic `search`/`show` organ, backed by Qdrant/Ollama vector retrieval when available; Qdrant is a derived index over CultCache and local artifacts, not a second truth. Search reports expose class focus/match and use phonetic class-affinity plus timeline-fact text so agents do not treat generic weak-articulation language as cross-class evidence. Source-edit-enabled IPA rounds may golf DSL syntax, parser/model records, Faust lowering, worker evidence surfaces, and tests when repeated evidence identifies a missing owner; every source edit needs an owner map, invariant, expected metric movement, verification, and rollback trigger. `hypotheses.md` is the hypothesis worker report of record; `hypothesis-agent.md` is only the external agent's final chat summary. IPA parity candidates should model scene voices explicitly: primary articulation, tract noise, release/transient, background-room, and recording-condition. At least one scene/noise or recording-condition family in a source-golf round must use real auxiliary `.aqua` `voice` declarations when claiming helper/background scene modeling. General synth primitives are allowed for helper scene voices, but they cannot count as articulation without gesture, primitive timeline, and clean-vocal movement. Crude full-store dumps are audit fallback, not agent memory. Statistical science workers and audio parity stages consume those bundles without rewriting gesture evidence. If gesture scores rise but rendered audio is silent, fix the primitive source/tract owner before adding full-patch dressing."
   },
   "ipa_vocal_tract_path": {
     "status": "This is now the active project center. Roadmap and research notes align the curriculum around structured utterance metadata -> learned utterance embedding -> learned synth automation -> vocal tract/envelope/LFO/filter/spectral controls -> rendered audio loss. The first Core contracts still defend the boundary: PhoneticIntent is the Weksa/language-owned handoff, ArticulatoryPlan is the AquaSynth realization layer, VocalTractMorphology declares anatomy capabilities, ArticulatoryConstraintReport is the host-visible veto surface, and VocalTractPlanResult is the tiny end-to-end proof artifact for host branching. IPA is an authoring input only: tokenization and feature mapping lower into language phonology profiles, timed articulatory gestures, morphology-specific tract/excitation curves, and Faust DSP. The staged support plan starts with tokenization and vowels, then core pulmonic consonants, expanded human IPA, airstream/phonation, and Weksa/alien morphologies. Optional eSpeak NG tests provide the first generated speech workout when ESPEAK_NG or espeak-ng is installed; eSpeak is intelligibility/timing/log-mel pressure, not anatomy truth. AquaSynth owns the learned utterance embedding encoder and synth driver: speech text embeddings plus prosody/emphasis hints plus Ghostlight/Epiphany-shaped character state vectors compress into utterance embeddings, and phonetic features plus that embedding feed a packed-buffer native C# backprop/Adam mapper that predicts tract, AM/FM, LFO, filter, and mel spectral-envelope controls. SpeechBackpropagationPipeline chains the two models: synth-driver target loss returns gradients with respect to its input, slices the utterance-embedding gradient, and updates the utterance encoder from downstream speech loss. The generated ground-truth training fixture renders eSpeak syllables, extracts log-mel evidence, and trains that chained pipeline toward eSpeak-seeded mel/control targets. CultMesh speech training now has payload manifests, peer/lease admission receipts, worker assignment, result collection, and central checkpoint application; workers do not own model updates. Real remote artifact chunk transfer and process supervision remain CultMesh infrastructure surfaces, not AquaSynth workaround territory. IpaGestureExperiment writes frozen descriptor/variant rounds with candidate scripts, primitive timelines, gesture metrics, manifests, and evidence JSONL, then analyzes them into metric summaries, score-band clusters, and a science brief. IpaTrialOrchestrator now writes local five-seed IPA render trials, accepts externally authored candidate scripts through RunCandidateScriptsAsync, writes WAV/report artifacts, evaluator reports, extracted primitive timeline facts, and `ipa-trial-results.cc` CultCache records. `tools/IpaTrialWorker` is the public worker tool for seed/score/search/show/index over the store; search now auto-upserts Qdrant/Ollama evidence chunks including timeline-fact text and returns vector-backed reports, while `tools/run-ipa-trial-loop.ps1` dispatches external Codex hypothesis/evaluator workers around five five-phoneme target sets. The latest seed run improved voiced vowels/nasals after primitive `SourcePort` lowering gained an oscillatory carrier; plosive closure/release remains the next weak owner.",
@@ -122,8 +122,8 @@
   },
   "last_verified": {
     "command": "powershell -NoProfile -ExecutionPolicy Bypass -File tools\\run-ipa-trial-loop.ps1 -Rounds 0 -SkipLocalTrialRun; dotnet build tests\\AquaSynth.Dsl.Tests\\AquaSynth.Dsl.Tests.csproj --no-restore --disable-build-servers -p:UseSharedCompilation=false -p:BuildInParallel=false -v minimal",
-    "result": "IPA trial loop prompts now allow source-edit-enabled DSL/Faust/worker golf with a source-change plan and pre-score build, and require a scene voice model using general AquaSynth primitives for helper/background voices while preserving articulation evidence boundaries.",
-    "commit": "pending ipa source golf scene voices"
+    "result": "IPA trial loop prompts now allow source-edit-enabled DSL/Faust/worker golf with a source-change plan and pre-score build, treat `hypotheses.md` as the report of record, and require at least one real auxiliary `.aqua` helper voice when a scene/noise family claims background or recording-condition modeling.",
+    "commit": "pending ipa source golf scene voice contract"
   }
 }
 

state/scratch.md

@@ -2,6 +2,23 @@
 
 Current slice:
 
+- Loose one-round IPA source-golf run completed under
+  `artifacts/parity/ipa-trial-loops/20260531T001831690` with 25 candidates,
+  source edits enabled, and focused pre-score build passing. The external agent
+  did not edit tracked source and chose `Source Edit Decision: none`, which is
+  acceptable evidence: the round found value-sweep pressure, not a repeated
+  DSL/Faust owner failure. Metrics did not improve: seed logMean `0.42306004`
+  vs loose logMean `0.41700276`, artMean stayed flat around `0.181`,
+  rmsMean dropped from `0.48316268` to `0.445383`, releaseMean dropped from
+  `0.06991132` to `0.06348304`, and radiationMean rose slightly from
+  `0.27946476` to `0.28285928`. The useful lesson was harness coherence: the
+  evaluator incorrectly judged `hypothesis-agent.md` instead of the actual
+  full report `hypotheses.md`, and the scene-voice prompt allowed prose-only
+  background modeling. `tools/run-ipa-trial-loop.ps1` now names
+  `hypotheses.md` as the report of record and requires at least one
+  scene/noise or recording-condition family to add actual auxiliary `.aqua`
+  `voice` declarations while preserving the primary articulatory vocal network.
+
 - Expanded the IPA trial-loop agent contract so future rounds can golf DSL
   syntax, lowering, and Faust machinery when the evidence says the owner is
   wrong. `tools/run-ipa-trial-loop.ps1` now gives source-edit-enabled

state/spine.yaml

@@ -146,7 +146,7 @@ live_invariants:
   - IPA training scoring contract is documented. Reports must keep `gesture_score`, `clean_vocal_score`, and `full_parity_score` separate. Gesture score measures descriptor-to-surface coverage, motion direction, spline contour/timing, primitive timeline consequences, and optional external articulation landmarks. Full parity patches may use FM/AM/modulators/envelopes/filters, auxiliary animated voices, and recording-condition helpers, but those helpers cannot count as articulation evidence unless gesture and clean-vocal scores move coherently.
   - `IpaGestureExperiment.WriteRound` is the first frozen batch harness for IPA hypothesis work. It writes `manifest.yaml`, candidate `.aqua` scripts, primitive `ProbeTimelineReport` CSVs, gesture-layer `metrics.csv`, and per-candidate `evidence.jsonl` under a caller-provided artifact root. `IpaGestureExperiment.AnalyzeRound` writes `analysis/metric-summary.csv`, `analysis/candidate-clusters.csv`, and `analysis/science-brief.md` for the science-worker handoff. This harness owns frozen descriptor/spline/primitive evidence only; clean vocal identity, full spectrogram parity, optimizer checkpoints, and worker orchestration stay downstream. Focused verification: `PatchScriptTests` pass 93/93.
   - `IpaTrialOrchestrator.RunAsync` is the first local render/scoring worker for IPA trials. It generates candidate patches through `IpaGestureExperiment`, renders them through Faust, compares against local PT fixture references with `AudioAnalyzer`, writes WAV/report artifacts, and stores typed `IpaTrialResult` records in `ipa-trial-results.cc`. The five-seed trial run at `tests/AquaSynth.Dsl.Tests/bin/Debug/net10.0/artifacts/parity/ipa-trials/20260529T214127955/five-seed-trials` showed voiced vowels/nasals became audible only after primitive `SourcePort` lowering gained a true oscillatory voiced carrier; sibilants are pressure, plosives remain weak. Verification: opt-in IPA trial test passed, and focused `PatchScriptTests|IpaTrialResultsRoundTripThroughCultCache` pass 94/94.
-  - External IPA exploration now has a worker/tool boundary. `tools/IpaTrialWorker` owns `seed`, `score`, `search`, `show`, and audit-only `dump` commands over the typed `.cc` trial-result store; `score` renders external Codex-authored `.aqua` candidates through `IpaTrialOrchestrator.RunCandidateScriptsAsync`, while `search` is the semantic retrieval organ for hypothesis agents. `IpaTrialResult` records now embed extracted primitive timeline facts from `ProbeTimelineReport` for contact closure/reservoir/release, source load/flow, branch admittance/exchanged flow, radiation boundary/output, path area/delay/energy, and passivity; `search`, `show`, `dump`, and summary CSV output surface release/radiation/passivity witnesses directly from the `.cc` record. `search` follows VoidBot's Qdrant shape: `.cc` records plus timeline facts and referenced local artifacts become evidence chunks, local Ollama `qwen3-embedding:0.6b` produces 1024-dim vectors, Qdrant collection `aquasynth_ipa_trial_results` stores stable SHA-256 UUID points with trial/candidate payload metadata, and the report surfaces vector scores, excerpts, tags, metric facts, timeline facts, contrast candidates, `class_focus`, per-hit `class_match`, and `show` commands. Qdrant is a derived index; CultCache remains truth, and lexical/metric search is fallback when the vector stack is unavailable. Retrieval tuning added phonetic class-affinity scoring and candidate-specific mixed-lane ownership so `mix-p` can support stop seams without every mixed candidate matching stop evidence. `tools/run-ipa-trial-loop.ps1` orchestrates external `codex exec` hypothesis/evaluator workers around that store. Source-edit-enabled rounds now allow agents to golf DSL syntax, parser/model records, Faust lowering, worker evidence surfaces, and tests when evidence points at a repeated owner failure; agents must write `source-change-plan.md`, and the loop runs a focused build before scoring. Candidate reports must include a `Scene Voice Model` naming primary-articulatory, tract-noise, release/transient, background-room, and recording-condition voices, with general AquaSynth synth primitives allowed for helper scene voices but not accepted as articulation evidence unless lower-layer witnesses move. The active trial shape is five target sets of five phoneme lanes each: vowels, nasals/approximants, fricatives, stops, and mixed generalization. Candidate names are exactly `<targetId>__<family>__<hypothesis-name>.aqua`, with lowercase kebab-case family ids. Three dry external Codex smoke passes refined hypothesizer/evaluator prompts into exact 25-lane output, retrieval receipts, claim audits, evidence-quality ledgers/verdicts, and fail-closed evaluator compliance. A later five-pass prompt optimization showed the best science comes from fixed receipt packets plus constrained reasoning; production prompts now require contrast pairs, `show` metric values, primitive timeline facts/excerpts or explicit `missing`, class-consistent reference matrices, scene voice models, `X owns Y so Z remains true` owner sentences, at least five micro-sweep perturbations before a seam is actionable, and a novelty gate rejecting loudness/dressing-only families as loss-landscape novelty. Verification: focused builds/tests pass, old-store show compatibility passes, and a fresh one-candidate smoke under `artifacts/parity/ipa-timeline-fact-smoke/20260531T233217641` proves release/radiation/passivity facts persist and retrieve through Qdrant/Ollama.
+  - External IPA exploration now has a worker/tool boundary. `tools/IpaTrialWorker` owns `seed`, `score`, `search`, `show`, and audit-only `dump` commands over the typed `.cc` trial-result store; `score` renders external Codex-authored `.aqua` candidates through `IpaTrialOrchestrator.RunCandidateScriptsAsync`, while `search` is the semantic retrieval organ for hypothesis agents. `IpaTrialResult` records now embed extracted primitive timeline facts from `ProbeTimelineReport` for contact closure/reservoir/release, source load/flow, branch admittance/exchanged flow, radiation boundary/output, path area/delay/energy, and passivity; `search`, `show`, `dump`, and summary CSV output surface release/radiation/passivity witnesses directly from the `.cc` record. `search` follows VoidBot's Qdrant shape: `.cc` records plus timeline facts and referenced local artifacts become evidence chunks, local Ollama `qwen3-embedding:0.6b` produces 1024-dim vectors, Qdrant collection `aquasynth_ipa_trial_results` stores stable SHA-256 UUID points with trial/candidate payload metadata, and the report surfaces vector scores, excerpts, tags, metric facts, timeline facts, contrast candidates, `class_focus`, per-hit `class_match`, and `show` commands. Qdrant is a derived index; CultCache remains truth, and lexical/metric search is fallback when the vector stack is unavailable. Retrieval tuning added phonetic class-affinity scoring and candidate-specific mixed-lane ownership so `mix-p` can support stop seams without every mixed candidate matching stop evidence. `tools/run-ipa-trial-loop.ps1` orchestrates external `codex exec` hypothesis/evaluator workers around that store. Source-edit-enabled rounds now allow agents to golf DSL syntax, parser/model records, Faust lowering, worker evidence surfaces, and tests when evidence points at a repeated owner failure; agents must write `source-change-plan.md`, and the loop runs a focused build before scoring. `hypotheses.md` is the hypothesis worker report of record; `hypothesis-agent.md` is only the external agent's final chat summary. Candidate reports must include a `Scene Voice Model` naming primary-articulatory, tract-noise, release/transient, background-room, and recording-condition voices, with general AquaSynth synth primitives allowed for helper scene voices but not accepted as articulation evidence unless lower-layer witnesses move. At least one scene/noise or recording-condition family in a source-golf round must implement real auxiliary `.aqua` `voice` declarations instead of satisfying scene modeling with prose. The active trial shape is five target sets of five phoneme lanes each: vowels, nasals/approximants, fricatives, stops, and mixed generalization. Candidate names are exactly `<targetId>__<family>__<hypothesis-name>.aqua`, with lowercase kebab-case family ids. Three dry external Codex smoke passes refined hypothesizer/evaluator prompts into exact 25-lane output, retrieval receipts, claim audits, evidence-quality ledgers/verdicts, and fail-closed evaluator compliance. A later five-pass prompt optimization showed the best science comes from fixed receipt packets plus constrained reasoning; production prompts now require contrast pairs, `show` metric values, primitive timeline facts/excerpts or explicit `missing`, class-consistent reference matrices, scene voice models, `X owns Y so Z remains true` owner sentences, at least five micro-sweep perturbations before a seam is actionable, and a novelty gate rejecting loudness/dressing-only families as loss-landscape novelty. Verification: focused builds/tests pass, old-store show compatibility passes, and a fresh one-candidate smoke under `artifacts/parity/ipa-timeline-fact-smoke/20260531T233217641` proves release/radiation/passivity facts persist and retrieve through Qdrant/Ollama.
   - TurbulenceJet now includes a release-timed burst-noise component in addition to the deterministic pressure pulse, but a stronger trial barely moved papa; the source is not the main remaining articulation bottleneck. Graph radiation now mixes 30% raw boundary flow with 70% high-passed flow after local admittance weighting, down from the over-bright 15/85 blend. Latest radiation/source artifact under `artifacts/parity/pink-trombone-utterance-logmel/20260527T174236049`: mama/papa/thrombosis cosine 0.6578/0.7692/0.5585, rmsRatio 3.0006/2.6153/2.3488, articulation 0.1437/0.1173/0.1191. Static `closure-release` under `artifacts/parity/pink-trombone-logmel/20260527T174139661` has rmsRatio 0.9900 and speechBandRatio 0.0842. This is still not speaking; next pressure is true tract resonance/passive scattering and pressure storage, not more burst gain.
   - Same-path two-port area discontinuities now emit `graph_area_energy_in_*` and `graph_area_energy_out_*` probes alongside the three-port branch `graph_connection_energy_*` probes. Use these to inspect passivity and formant-body failures at the scatter sites rather than judging only final output loudness.
   - `FaustExportOptions.DebugProbeUi` wraps graph probe signals in Faust `vbargraph` UI controls under `/debug/...` without changing audio output arity. Native streaming Faust patches now separate writable controls from read-only probe zones and can read/snapshot probe values after `ProcessBlock`; use this for acoustic graph passivity/radiation inspection, leave it off for normal parity renders.

tools/run-ipa-trial-loop.ps1

@@ -346,6 +346,8 @@ Scene voice model:
 - Every candidate patch represents a scene, not a single naked waveform. The report must name the scene voices and say how each is synthesized.
 - Required scene voice roles: `primary-articulatory`, `tract-noise`, `release/transient`, `background-room`, and `recording-condition`.
 - Use ordinary AquaSynth primitives for helper voices where useful: `voice` layers, noise, FM, AM/tremolo, envelopes, filters, formants, additive/spectral voices, macros, and modulation. These helper voices are allowed in full parity.
+- At least one hypothesis family must implement real helper/background scene voices in the `.aqua` scripts using ordinary `voice` declarations, such as filtered `wave=noise`, FM partials, AM/tremolo beds, or formant-colored capture layers. Do not satisfy the scene voice requirement with prose alone.
+- When adding helper voices, keep gain low and name the role in the report. The primary articulatory `vocal network=...` voice must remain present unless the family explicitly tests a source-edit replacement.
 - Helper/background voices must not count as articulation evidence unless gesture, primitive timeline, and clean vocal metrics move coherently.
 - For each family, say which voices are active, which controls they expose, and which metric/timeline witness would prove the helper is not merely dressing.
 
@@ -365,7 +367,7 @@ Candidate design contract:
 - A seam claim must use an owner sentence: `X owns Y so Z remains true`.
 - A seam is actionable only if the next candidate family proposes at least five micro-sweep perturbations across that seam, spread across target lanes when possible.
 - Novelty gate: a family must move a new owner, control axis, timing contour, or primitive relationship. Raising loudness, burst gain, or dressing alone is not novel.
-- Scene novelty can be valid only when it changes a named voice role, routing/lowering primitive, modulation relationship, or background-noise synthesis model and predicts a non-articulation witness such as noise-floor, spectral tilt, RMS stability, or room/recording match.
+- Scene novelty can be valid only when it changes a named voice role, routing/lowering primitive, modulation relationship, or background-noise synthesis model and predicts a non-articulation witness such as noise-floor, spectral tilt, RMS stability, or room/recording match. A scene-noise or recording-condition family that does not add an actual helper voice is not novel.
 - Include a class-consistent reference matrix that maps each target id to the exact reference trial or fixture id used for comparison; mixed-set evidence cannot silently stand in for single-phoneme evidence.
 
 Required `$roundDir/hypotheses.md` shape:
@@ -380,7 +382,7 @@ Required `$roundDir/hypotheses.md` shape:
 - `Novelty Gate`: one row per family saying which new owner/control axis/timing contour/primitive relationship is being explored; write `dressing-only` if the family mainly changes loudness/FM/AM/noise/envelopes.
 - `Candidate Matrix`: exactly 25 rows with target id, filename, family, expected metric movement, and risk.
 - `Source Edit Decision`: `none` unless source edits are allowed and justified by evidence; if source was edited, link `$roundDir/source-change-plan.md`, list files changed, tests/builds run or pending, and rollback trigger.
-- `Acceptance Checklist`: 25 files, one per target id, filename schema, required report sections, >=3 search receipts, >=3 show receipts.
+- `Acceptance Checklist`: 25 files, one per target id, filename schema, required report sections, >=3 search receipts, >=3 show receipts, and at least one family with actual auxiliary `voice` helper/background declarations in its candidate scripts.
 
 Authority boundary:
 - Source edits allowed: $AllowCodexSourceEdits
@@ -489,7 +491,7 @@ Before judging:
    - weak regressions / failed articulation;
    - promising transfer / generalization, including DSL lowering/Faust/scene-voice source-edit evidence when source edits were allowed.
 2. Use `show` on at least one promising and one weak candidate from the current round.
-3. If the hypothesis worker skipped the required 25-lane matrix or retrieval receipts, mark orchestration/prompt compliance as failed even if some audio metrics improved.
+3. If the hypothesis worker skipped the required 25-lane matrix or retrieval receipts in `$roundDir/hypotheses.md`, mark orchestration/prompt compliance as failed even if some audio metrics improved.
 4. If source edits were allowed, inspect `$roundDir/source-change-plan.md` when present and verify the source edit is a coherent owner move, not a metric hack.
 
 Use the worker as your search organ:
@@ -511,8 +513,9 @@ Verify compliance from disk as well as the hypothesis report:
 - validate family ids are lowercase kebab-case.
 - derive canonical family ids only from the filename family segment; ignore prose family names when they disagree.
 - validate that each actionable seam has a contrast pair, a primitive timeline excerpt or explicit `missing`, an owner sentence, and at least five planned micro-sweep perturbations.
-- validate that the hypothesis report includes a class-consistent reference matrix, scene voice model, and novelty gate.
+- validate that `$roundDir/hypotheses.md` includes a class-consistent reference matrix, scene voice model, and novelty gate. `hypothesis-agent.md` is only the external agent's final chat summary and must not be treated as the full report.
 - validate that every family identifies primary-articulatory, tract-noise, release/transient, background-room, and recording-condition roles, even if some are explicitly silent.
+- validate that at least one family actually adds ordinary helper/background `voice` declarations in its `.aqua` scripts for scene noise or recording-condition modeling.
 - validate that helper/background voices are evaluated separately from articulation evidence.
 - validate source edit compliance if source edits were allowed: coherent owner map, edited file list, invariant, expected metric movement, verification command, and rollback trigger.
 - If any required section, receipt class, per-target validation, filename validation, family extraction, reference matrix, scene voice model, contrast pair, owner sentence, or novelty gate is incomplete, set `Round Compliance: failed`.
@@ -527,7 +530,7 @@ $roundDir/evaluation.md
 Required `$roundDir/evaluation.md` shape:
 - `Retrieval Receipts`: exactly five receipts: three search outputs plus one weak show output and one promising show output. Cite no other evidence files in this section. Only cite claims backed by these file paths; write `unknown` when evidence is absent.
 - `Commands Run`: exact search/show commands and output paths used to support claims.
-- `Round Compliance`: passed/failed, candidate count, target coverage, naming validity, family extraction, reference matrix, contrast-pair coverage, owner-sentence coverage, micro-sweep coverage, novelty gate, and whether the report had the required matrix.
+- `Round Compliance`: passed/failed, candidate count, target coverage, naming validity, family extraction, reference matrix, contrast-pair coverage, owner-sentence coverage, micro-sweep coverage, novelty gate, auxiliary scene voice presence, and whether `hypotheses.md` had the required matrix.
 - `Family Verdicts`: one row per canonical filename family with improved/flat/regressed/unknown metrics. Treat |delta| < 0.01 as flat; if deltas are not present in cited show artifacts, write `unknown` and reason=`no delta in receipts`.
 - `Evidence Quality Verdicts`: one row per family with specificity, comparability, falsifiability, reuse value, primitive timeline support, and weakest missing evidence. Penalize vague receipts, search-summary-only claims, missing contrast pairs, missing reference ids, and missing timeline excerpts even when metrics improve.
 - `Seam Audit`: one row per named seam with owner sentence, contrast pair, timeline excerpt, planned perturbation count, and verdict `actionable` or `not actionable`; fewer than five perturbations means `not actionable`.