# Light opening review — 10 September 2026

## Finding

The screenshots showed the front shell with optical components hidden. That makes opaque support walls behind the aperture visible. There was also an actual alignment concern: the visible dot had moved from Y=5 mm to Y=2 mm, while the LED, right-angle light guide and rear lens collector remained at Y=5 mm.

The old clear lens was connected and did not intersect the shell, so this was not proof that the indicator would be completely dark. However, a native Fusion section confirmed opaque holder material in the projected column behind the lowered output and an offset transfer through the lens. A zero lens/shell collision result did not validate that optical path. We have removed the avoidable offset rather than accept an untested shadow/uniformity risk.

## R3.1 correction

| Item | Change |
|---|---|
| Visible light | Unchanged: center X=0, Y=2; aperture Ø4.9 mm; visible lens Ø4.7 mm; same front face and chamfer |
| Hidden lens | New centered Ø5.4 mm flange, Z=-3.5 to -2.15 mm; original offset collector/key tab removed |
| Flange seat | Ø5.6 mm; nominal 0.10 mm radial clearance; flange overlaps Ø4.9 aperture by 0.25 mm radially |
| Front shell LED support | Local cradle rebuilt at Y=2 from the previous cradle dimensions; centered sleeve added; old higher support removed |
| LED, guide, bond film, keeper, liners, two screws | Translate Y=-3 mm together; their individual geometry is unchanged |
| Keeper screw centers | X=-7 mm; Y=-3.4 and +7.4 mm. Original M2 × 5 mm references retained |
| Exterior and other internals | No changes to the case envelope, dot position, PCB, microphone or R3 controls |

The full assembly is still a right-angle optical arrangement: the LED points sideways into a clear guide, which turns the light toward the front face. Seeing a support behind the bend is normal. The corrected exit leg is now aligned with the lens; it does not depend on the old 3 mm sideways transfer at the lens interface.

## Verification

- Autodesk Fusion temporary trial and committed-body checks: 393 candidate body pairs; no new solid intersections above 0.0001 mm³.
- Fourteen relocated solid-body paths plus their parent LED assembly were checked against the original rigid translation. 304 other component paths were checked for unchanged bounds and volume.
- Native receiver probes: Ø4.2 mm from Z=-3.9 to -0.1 mm and Ø3.5 mm from Z=-7 to -3.9 mm, centered X=0/Y=2. Both are filled by optical solids/bond allowance and have zero opaque-shell overlap. The smaller lower probe respects the retained guide's flat side-entry face; it is not a predicted beam diameter.
- The visible Ø4.7 mm lens column is clear of the shell across sampled sections behind the face.
- Saved archive reopened and compared with the export inventory. Exact Boolean comparisons preserve the visible lens, shell exterior and shell outside the local light repair region.
- Independent OpenCascade checks cover STEP validity, counts, bounds and volume; STL bounds and DXF millimeter units are checked separately. See `audit/independent_export_verification.json` for final results.

These are geometry checks, not ray tracing or a measured optical efficiency. The area occupied in a CAD section must not be interpreted as a percentage of brightness lost or gained.

## Optical basis

Use transparent optical polymer for the guide and lens body, with diffusion only where appropriate at the visible face. Clear acrylic/PC, polished guiding surfaces, LED spacing and output diffusion are established light-guide considerations; the selected prototype material and process must be confirmed on the actual part. See the manufacturer's [VCC light-pipe design guide](https://vcc.co/light-pipe-design-guide/). Its general guidance does not qualify this custom geometry or give it a guaranteed efficiency.

The section image uses opaque blue only to identify the clear optical path. It is not a request for blue resin, opaque plastic or glass.