Freddy-Fazbear-Digital-Eyes/hardware/fit.md

# Freddy mask eye fit — preliminary review (2026-09-27)

The mask is stationary in a skeleton's lap on the covered porch. These observations are for a removable behind-the-opening screen mount; no mount geometry is approved for printing yet.

## Evidence received

- [Front photo with annotations](../reference_images/freddy-mask-eye-measurements.png): 2.48 in (62.99 mm) is marked vertically across the photo-left opening and horizontally across the photo-right opening. The irregular rims mean this is **not** a caliper measurement of the narrowest opening in both axes of both eyes.
- [Rear photo](../reference_images/freddy-mask-eye-rear-view.png): the eye rims sit in a curved, faceted shell. Brow and muzzle/bridge geometry occupy space close to the openings; a center seam and glue are visible. No flat eye-board mounting land or verified connector route is evident from this view alone.
- [Rear dry-fit photo](../reference_images/IMG_6528.jpeg) and [side dry-fit photo](../reference_images/IMG_6529.jpeg): one unpowered Waveshare board has been held behind an eye opening. Its bare body appears to clear enough to occupy this hand-held position, and its rear mounting standoffs are exposed. The board's inner edge is close to the glued center bridge. The side photo shows a setback/angle relationship to the printed rim, but includes no scale, so depth and glass-to-rim clearance cannot be quantified.
- [Connected rear photo](../reference_images/IMG_6530.jpeg): in a held position, the `USB` power cable and 4-pin UART header plug are both inserted. The UART wires leave near the glued center bridge; the separate UART USB-C port appears unplugged. This confirms plug insertion is possible on the photographed eye, not that a fixed mount leaves sufficient bend radius, strain relief, or clearance on both sides.
- [Front close-up 1](../reference_images/IMG_6532.jpeg) and [front close-up 2](../reference_images/IMG_6533.jpeg): the dark, unpowered module sits behind the opening. A narrow bright sliver appears at the photo-right edge, but the unlit view does not distinguish the active screen from its surround and the deeper socket. With the screen off and no scale, these views alone do not establish active-pixel visibility, final eye appearance, or exact alignment.
- [Powered close-up](../reference_images/IMG_6534.jpeg), [wider front view](../reference_images/IMG_6535.jpeg), and [another held orientation](../reference_images/IMG_6536.jpeg): the rendered blue iris/white eye is clearly visible through one opening. These views clarify that a substantial dark portion of the socket remains visible around the illuminated display, especially above and to one side; the earlier bright edge was not sufficient to judge the overall fit. The board's apparent position and rotation change between shots (the on-screen label rotates), so none establishes a fixed center or final orientation. In the last view a red light is visible near the photo-right edge of the socket; check whether that onboard indicator would be visible to visitors in the installed position. The screen is not visibly clipped in these views, but the photos do not show both eyes installed together or measured clearances.
- [Shapr3D mask STEP](../stls/freddy_fazbear_mask.step): AP242 export in millimetres, with 17 solids. The two main head halves are separate solids, as are the muzzle, jaw, ears, and smaller parts. Its overall bounds are 518.761 × 227.383 × 386.090 mm **across the assembly**, not eye-opening dimensions. The older [STL export](<../stls/freddy_original_solids_cut_parts.step (3).stl>) has matching overall bounds but loses the STEP's units and solid structure. The earlier `freddy-fazbear-mask-BBL.3mf` and `EYES_For_MESH.stl` mentioned in records are not currently present in this workspace.
- The Waveshare ESP32-S3-Touch-LCD-2.1 is nominally 75 mm across per [Waveshare's dimensions](https://www.waveshare.com/wiki/ESP32-S3-Touch-LCD-2.1). Waveshare also publishes an [official 3D assembly](https://files.waveshare.com/wiki/ESP32-S3-Touch-LCD-2.1/ESP32-S3-Touch-LCD-2.1_structure.zip); its overall CAD envelope measures 75 × 75 × 9.5 mm. Confirm this reference assembly matches Joshua's Rev2.0 boards before treating small features as final. It does **not** include the plugged-in USB and UART cables. The board cannot pass through either front eye opening and must be placed from behind.

## STEP-based eye geometry

I imported the mask STEP with the Shapr3D CAD connection and measured front-to-back clear sightlines on a 0.5 mm X/Z grid using [this repeatable analysis](analyze_mask_step.py). Coordinates below are in the STEP's millimetres; front is approximately negative Y. These are **projected clear apertures**, not caliper measurements of the printed rim or a guaranteed mount clearance.

| STEP head half | Clear X bounds | Clear Z bounds | Widest horizontal span | Tallest vertical span | Clear-area center | Largest fully clear circle |
| --- | ---: | ---: | ---: | ---: | ---: | ---: |
| Negative X | −66 to −3 mm | 168.5 to 223.5 mm | ~62.5 mm | ~54.5 mm | (−33, 194.9) mm | ~52.2 mm diameter at (−30, 196) mm |
| Positive X | 29 to 92 mm | 168.5 to 223.5 mm | ~62.5 mm | ~54.5 mm | (58.9, 194.9) mm | ~52.2 mm diameter at (56, 196) mm |

The clear-area centers are about 92 mm apart; the centers that maximize an unoccluded circular view are about 86 mm apart. Those are different aesthetic placement choices, not prescribed mounting-hole positions. The mask aperture is wider than it is high. The nominal 2.1-inch display circle is about 53 mm, so a perfectly full, unobstructed display circle is unlikely at every angle; a narrow, deliberate edge overlap may look more integrated. In a coarse 2 mm scan of an untilted 75 mm disc centered at either largest-circle point, the head shell extends rearward to approximately Y = −20.5 mm near the upper rim. The shell depth varies strongly around the opening, making the adjustable rear cradle and tapered cosmetic surround useful.

For a **candidate straight-on CAD pose**, I rotated the official board model so its display faces toward negative Y, placed its center at (X, Z) = (−30, 196) or (56, 196) mm, and checked the 75 mm circular body plus the main display/PCB/standoff solids against the corresponding STEP head half. At board-origin Y = −18 mm the negative-X 75 mm body still intersects the shell by about 7 mm³; at Y = −17 mm that body no longer intersects. At Y = −15 mm, none of the selected major board solids intersects either head half. This is a starting position only: it does not quantify minimum surface gap, include all 513 vendor sub-solids or plugged-in cables, prove the 3D model matches the Rev2.0 units, or prove a good off-axis view. The chosen powered appearance may require changing center, tilt, or setback and rechecking interference.

## Provisional mount approach

Use two independent, removable rear cradles or retainers, one per eye, with the lit display centered behind the visible opening. Support the module by its housing/PCB mounting points with nonconductive spacers; do not clamp the glass, touch surface, display ribbon, or components. Keep a small adjustable setback so the display does not bear on the uneven printed rim. Provide access to the eye boards' `USB` ports for service/power and to the 4-pin UART headers for camera RX/common ground; leave the separate UART USB-C ports unplugged because they disable the headers. Route wires with strain relief and clearance from the brow/bridge/muzzle. The final print should be split or removable so either eye can be replaced without removing the whole mask.

Joshua's chosen appearance is **integrated with the mask**, rather than an exposed deep black socket. Pair each rear cradle with a separate, removable cosmetic eye surround that follows the irregular printed opening and visually carries the mask surface inward toward the lit circle. An opaque, matte inner wall should conceal the open cavity, gray display bezel, wiring, and any visible indicator light; finish its visible face to match the mask's eventual paint/texture. Leave a narrow, even dark reveal at the display edge instead of pressing trim against the glass or covering active pixels. The surround may need a tapered/stepped profile because the printed rim and display are at different depths. Keep the two eyes independently serviceable, and allow small center/tilt/setback adjustments before locking their positions.

This design direction now has [dimensioned first-fit carrier CAD](eye_mount/README.md). It is still a prototype, not print-fit approval: its separate trim annulus is only an appearance gauge, not the final tapered surround. Avoid bonding the PCB directly to the glued center seam. Do not glue or cut the printed mask based on the 2.48 in front annotation alone. A first print should be a low-cost test fit of one eye, not the finished pair.

## Remaining checks before mount CAD

1. Confirm the STEP export is the same scale and assembly state as the printed mask. The CAD removes the need to estimate the two projected eye silhouettes by hand, but glue, print tolerance, and assembly alignment can change the real fit.
2. Register the official board 3D assembly against each STEP head half at a chosen center, tilt, rotation, and setback; test interference with the shell, brow, bridge, and muzzle. The 75 mm reference-disc check above is only a conservative starting envelope.
3. Verify the **actual** board's rear mounting-point/thread geometry and its powered visible circle against the vendor CAD, then reserve space for the inserted USB and UART plugs, cable bend, and strain relief. Those cables are absent from both CAD models.
4. With both boards temporarily held at the selected positions, compare straight-on powered views at normal visitor distance with close-ups. Confirm the finish/color of the mask and whether the thin intentional display-edge overlap and cosmetic surround achieve the desired integrated appearance.

The first-fit rear carrier/pad/sleeve set is now modeled and clears both mask head halves in CAD. The official vendor assembly also clears the carrier and trim at this candidate pose; the printed Rev2.0 boards and plugged cables still need checking. Print one test set, verify both physical sides and the powered front appearance, then finalize the retained cosmetic surrounds and service removal. One eye has photographed powered placement and plugs inserted; the final center/rotation/setback, second eye, repeatable attachment, fixed cable routing, and strain relief remain unverified. Do not size the final surround from these perspective photos alone.