Skip to content

Hermetic Calibration Baseline — Live Failure State

Notebook type: live failure / presentation evidence Date: 2026-07-26 (captures) · entry prepared 2026-08-01 · ontology pass note 2026-08-03 Engine commit (entry baseline): 6e69d792 Doctrine: Display ≠ probe. Outcome-plane semantics outrank RGB. Magenta on the plate is not a ProbeOutcomeCode.


Question

Can these live captures establish a Hermetic Normal Calibration baseline (closed box → HitGeometry → valid normals → multi-face NormalRGB seams)?


Experiment identity

Field Value
Intent Establish Hermetic Normal Calibration baseline
Actual result Live failure state relative to that intent
Evidence class Presentation only (two live screenshots)
Sealed outcome plane for these frames Not attached to this notebook entry
SceneIds involved (host) Gallery / portal-facing (captures); Hermetic intended for calibration ladder below

Presentation evidence (paths):

  1. screenshots/2026-07-26 Canonical Portal Scene with GRIN Film Camera Diagnostics normals.png
  2. screenshots/2026-07-26 Canonical Portal Scene with GRIN Film Camera Diagnostics normals_B.png

Chrome visible in captures includes roughly: Film LIVE · Preview 80×45 · Opacity ~40% · Field ~0.95 · exploratory legend including legacy “Magenta = no hit” wording (treat as legacy presentation chrome, not validated semantics).


DIRECT OBSERVATIONS

From the two supplied live screenshots as pixels and UI chrome only:

# Observation Caption role
1 Gallery viewport shows visible scene geometry (walls, portal surface, lettering / room structure depending on frame). world viewport
2 Gallery film plate is predominantly magenta, with a narrow gray (or non-magenta) region along part of the plate—not a multi-face orientation partition. Observation Plate + legacy presentation
3 Hermetic viewport (when shown / compared in the live session narrative) is nearly white—enclosure structure is not clearly readable as a multi-wall calibration cage in that presentation. world viewport
4 Hermetic film plate is predominantly magenta. Observation Plate + legacy presentation
5 No clean multi-face NormalRGB partition is visible on either film plate. display mapping (failed teaching read)

Also directly visible as presentation constraints: plate at 80×45 upscaled; ~40% opacity softens plate authority against the 3D view; portal glyphs can appear warped in Gallery (transport presentation active)—still not an outcome-plane class.


INTERPRETATION

Live failure state: presentation shows magenta-dominant plates and a near-white Hermetic viewport, so the session cannot be used as a successful Hermetic normal-calibration baseline until the experiment ladder below is executed with sealed evidence.

This is a qualification failure of the live baseline, not a proof of transport absence (warped portal lettering in Gallery still suggests active transport presentation).


NOT ESTABLISHED

Question Why not established
Whether rays hit Hermetic geometry Requires cathedral.probe.outcome with HitGeometry counts; RGB magenta does not decide this
Whether normals exist / are valid Requires trustworthy per-sample normals + typically HitGeometry
Whether camera contexts match Requires context key / lifecycle record per frame
Whether magenta means no hit Forbidden inference; legacy legend is not the outcome plane
Whether field curvature is responsible for magenta plates Requires Straight vs field-on under Complete snapshots + histograms

CORRECTION NOTES

Prior / risky reading Correction
HUD “Magenta = no hit” Retired as doctrine; presentation chrome only
Magenta plate ⇒ MaxStepsExhausted Requires sealed outcome plane
Field ~0.95 caused failure Unproven without histogram delta
NormalRGB “proves” faces Display mapping only

Post-36b003bd (exclude observer from film transport): later sessions may show increased structured room hits after self-intersection correction—still require sealed histograms; this notebook’s captures predate or do not attach that evidence.


Experiment ladder (required next)

Execute in order. Do not skip to NormalRGB interpretation. Do not interpret magenta as an outcome code at any step.

1. Six-axis straight-ray enclosure test

Goal Confirm the observer is inside a closed collision/visual enclosure (Hermetic primary).
Pass signal Consistent interior hits on all six axes under straight policy.
Fail signal Escape to void, exterior wall face, or no enclosure contact.

2. Completed FieldStrength = 0 snapshot

Goal Probe-complete SNAPSHOT with Straight transport.
Pass signal Lifecycle Complete; Unprocessed = 0; histogram sums to total.
Note Complete ≠ “pretty NormalRGB.”

3. Semantic outcome histogram

Channel cathedral.probe.outcome
Report Full code counts
Forbidden Inferring histogram from plate magenta

4. Valid-normal census

Among HitGeometry only. If unavailable: record unavailable; do not claim geometric-normal success.

5. NormalRGB encoding comparison

Compare Display Mode NormalRGB only on HitGeometry (and valid normals if known). Prefer 160×90 and 100% opacity for evidence stills.

6. Field-on repeat

Same pose; only field strength changes. Report semantic delta—not “magenta because field.”


UNRESOLVED QUESTIONS

  • Sealed multi-face NormalRGB success under Hermetic still not yet qualified.
  • Whether raw geometry.normal ships before UI hatch overlays.
  • How much of the 2026-07-26 magenta dominance was self-intersection vs fallback vs true max-steps.

CLAIM BOUNDARY

Allowed from these two screenshots Disallowed
Direct observation list above Magenta ⇒ no hit / max-steps
“Baseline not qualified for Hermetic Normal Calibration” Rays miss Hermetic geometry (unproven)
Need for experiment ladder 1–6 Field curvature caused magenta (unproven)
Display ≠ probe reminder Context match Gallery↔Hermetic (unproven)

Relation to other notes


One-line conclusion

Live presentation shows magenta-dominant Observation Plates and an unreadable Hermetic viewport; until the six-step ladder produces a Complete outcome plane and (if available) a valid-normal census, this session is a failed Hermetic normal-calibration baseline, not a semantic transport diagnosis.

Display ≠ probe. Magenta is not an outcome code.