Roadmap: Laser/DMT-claim OCR research tool hardening

[ProhibitedTV]

Goal

Solidify LaserAnalysisAI / LaserLab into a conservative research tool for analyzing laser footage where people claim to observe code-like structure in diffracted laser light. The project should not claim what any signal means. It should test whether footage contains repeatable structured detections above matched controls.

Research posture

• Analyze already-recorded optical footage and controlled fixture captures.
• Keep the language neutral: detections, candidates, controls, null results, reproducibility.
• Avoid instructions related to drug procurement, dosing, ingestion, or human-subject experimentation.
• Prioritize blinding, matched controls, provenance, repeatability, and false-positive control.

Target media model

LaserLab's primary research input should be wall-projection laser experiment footage: a real laser projected onto a wall or flat surface, optionally through an intermediate optical/diffraction material, captured with a fixed camera and compared against matched controls. Public optical videos and Wikimedia explanation clips are useful as calibration/regression/demo fixtures, but should not be treated as the core research footage model.

Audience and product identity

LaserLab should welcome DMT psychonauts who are bringing reports and footage, while remaining credible to science-minded reviewers. The interface can be dark, edgy, and cyberpunk, but the evidence posture must stay conservative: blinded review, matched controls, repeatability, null results, and audit trails.

Roadmap themes

1. Experiment protocol and safety/ethics boundaries.
2. Media ingestion, provenance, and frame extraction.
3. Optical preprocessing, ROI detection, and calibration.
4. OCR and glyph/structure detection ensemble.
5. Synthetic fixtures and matched negative controls.
6. Statistical validation, effect sizes, and false discovery control.
7. Human review UI and reproducible reporting.
8. CI/golden fixtures, packaging, and documentation.
9. Dark cyberpunk community-science UI polish.
10. Wall-projection capture model and capture-quality guidance.
11. Research-design metadata, laser safety, and optional subjective-report context.
12. Community falsification matrix and signal-detection protocol modules.
13. Diffracted cross-laser hardware profiles, wavelength variants, and optional symbol-corpus interoperability.
14. Optional Veilbreak API/MCP/schema interoperability while preserving local-first analysis.

Child tickets

• Phase 0: Define research protocol, scope, and safety boundaries #4 — Phase 0: Define research protocol, scope, and safety boundaries
• Phase 1: Media ingestion manifest with provenance and capture metadata #5 — Phase 1: Media ingestion manifest with provenance and capture metadata
• Phase 1: Deterministic frame extraction and sampling profiles #6 — Phase 1: Deterministic frame extraction and sampling profiles
• Phase 2: Optical preprocessing sweep for laser footage #7 — Phase 2: Optical preprocessing sweep for laser footage
• Phase 2: ROI candidate detector for code-like regions in beams #8 — Phase 2: ROI candidate detector for code-like regions in beams
• Phase 3: OCR engine adapter layer and ensemble results schema #9 — Phase 3: OCR engine adapter layer and ensemble results schema
• Phase 3: Non-OCR structure detectors for glyph-like patterns #10 — Phase 3: Non-OCR structure detectors for glyph-like patterns
• Phase 4: Synthetic fixture generator for positives, negatives, and artifacts #11 — Phase 4: Synthetic fixture generator for positives, negatives, and artifacts
• Phase 4: Matched controls and blinding workflow #12 — Phase 4: Matched controls and blinding workflow
• Phase 5: Statistical validation, null models, and false-discovery control #13 — Phase 5: Statistical validation, null models, and false-discovery control
• Phase 6: Blinded candidate review dashboard #14 — Phase 6: Blinded candidate review dashboard
• Phase 6: Evidence report export with reproducible audit trail #15 — Phase 6: Evidence report export with reproducible audit trail
• Phase 7: Public fixture and footage catalog with licensing notes #16 — Phase 7: Public fixture and footage catalog with licensing notes
• Phase 7: Golden-run regression tests and CI artifacts #17 — Phase 7: Golden-run regression tests and CI artifacts
• Phase 8: Researcher workflow docs and Windows release polish #18 — Phase 8: Researcher workflow docs and Windows release polish
• UI/UX: Dark cyberpunk community-science theme and simple information design #21 — UI/UX: Dark cyberpunk community-science theme and simple information design
• Capture Model: Prioritize wall-projection laser experiment footage #22 — Capture Model: Prioritize wall-projection laser experiment footage
• Research Design: Wall-projection protocol metadata, safety, and subjective-report context #23 — Research Design: Wall-projection protocol metadata, safety, and subjective-report context
• Research Design: Community falsification matrix and signal-detection protocol modules #24 — Research Design: Community falsification matrix and signal-detection protocol modules
• Hardware Profile: Diffracted cross lasers, wavelength variants, and symbol-corpus linkage #25 — Hardware Profile: Diffracted cross lasers, wavelength variants, and symbol-corpus linkage
• Interoperability: Veilbreak API/MCP data bridge and schema mapping #26 — Interoperability: Veilbreak API/MCP data bridge and schema mapping

Definition of done

• A user can create a blinded experiment with laser and control captures.
• The app can sample frames, generate preprocessing variants, identify candidate ROIs, run OCR/structure detectors, and compare against controls.
• Reports clearly distinguish no signal, artifact, anomaly, above-control candidate, and repeatable candidate.
• Every run emits reproducible JSON/CSV/HTML outputs with enough metadata to audit the result.
• The dashboard is simple, informative, dark-mode by default, visually distinctive, and aligned with the psychonaut + science community audience without overclaiming.
• Documentation, UI setup, manifests, and reports distinguish primary wall-projection experimental footage from calibration/regression/demo fixture footage.
• Research-design metadata captures optical setup variables and optional subjective-report context without turning subjective reports into detector evidence or giving substance-use instructions.
• Community falsification ideas are implemented as optional, clearly labeled protocol modules with tested/not-tested confounds, chance baselines, and conservative interpretation.
• Hardware metadata supports diffracted cross-laser pattern type, wavelength/color variants, cross-region ROI registration, and optional symbol-corpus export/import without assuming external dataset access.
• Optional Veilbreak-style references/imports preserve attribution, privacy status, source hashes, and separation between external community metadata and local detector evidence.

Implementation order

1. Phase 0: Define research protocol, scope, and safety boundaries #4, Phase 1: Media ingestion manifest with provenance and capture metadata #5, Phase 1: Deterministic frame extraction and sampling profiles #6 — lock down protocol, provenance, and deterministic media handling.
2. Phase 2: Optical preprocessing sweep for laser footage #7, Phase 2: ROI candidate detector for code-like regions in beams #8, Phase 3: OCR engine adapter layer and ensemble results schema #9, Phase 3: Non-OCR structure detectors for glyph-like patterns #10 — build analysis primitives.
3. Phase 4: Synthetic fixture generator for positives, negatives, and artifacts #11, Phase 4: Matched controls and blinding workflow #12, Phase 5: Statistical validation, null models, and false-discovery control #13 — prove the system can beat controls and reject false positives.
4. Phase 6: Blinded candidate review dashboard #14, Phase 6: Evidence report export with reproducible audit trail #15 — make review/export usable and auditable.
5. Phase 7: Public fixture and footage catalog with licensing notes #16, Phase 7: Golden-run regression tests and CI artifacts #17, Phase 8: Researcher workflow docs and Windows release polish #18 — harden fixtures, CI, and release/docs.
6. UI/UX: Dark cyberpunk community-science theme and simple information design #21 — apply the dark cyberpunk UI identity while preserving simplicity, readability, and conservative interpretation language.
7. Capture Model: Prioritize wall-projection laser experiment footage #22 — refine the capture model so LaserLab optimizes for real wall-projection experiment footage rather than generic laser explainer media.
8. Research Design: Wall-projection protocol metadata, safety, and subjective-report context #23 — add structured research-design metadata, laser safety/ethics boundaries, and optional subjective-report context that remains separate from detector evidence.
9. Research Design: Community falsification matrix and signal-detection protocol modules #24 — turn community falsification ideas into optional protocol modules such as grid-coordinate transcription, SDT/noise controls, gaze stability, parallax, binocular checks, and symbol-transcription statistics.
10. Hardware Profile: Diffracted cross lasers, wavelength variants, and symbol-corpus linkage #25 — add diffracted cross-laser hardware profiles, wavelength presets, cross-pattern ROI analysis, and optional symbol-corpus interoperability fields.
11. Interoperability: Veilbreak API/MCP data bridge and schema mapping #26 — add optional Veilbreak API/MCP/schema mapping so LaserLab can complement existing community data infrastructure without creating a network dependency or mixing external reports into detector evidence.