Where the model stands
The diagonal-ridge roof is confirmed: from the camera solved for photo #17, the model's eaves, apex, ridge-beam end and purlin ends land on the photo. The model reached 85.9% after three critique rounds, short of your 90% target. Below 90% the protocol requires your explicit sign-off, so treat this as the best current version rather than a finished one.
Why I stopped at round 3
The two lowest dimensions are limited by missing evidence, not effort. No photo shows the north or east sides, the bathroom, or the upstairs landing, and the interior photos can't fix depth precisely. More rounds would only polish rendering, worth one or two points at most. New photos are what would move the score; the list is at the end.
Photo against model
Each comparison renders the model through a camera solved from points I marked on your photo: corners, brick edges, header line and mullions. The right panel draws the model's edges in red on the photo.
Floor plans
Elevations
Orthographic renders straight from the model, levels above deck. In true elevation the level ridge reads as a flat top running to the northeast corner. No photo can show that, because the cameras sit nearly in the plane of the roof; it depends on the unverified ridge height there.
Presentation renders
Key dimensions
| Element | Value | Basis |
|---|---|---|
| Footprint | 6.50 × 6.50 m (21′-4″ square) | measured south face, plane fit 3 px; west mullions confirmed in #17 within 2–5 px |
| Door / lower glazing head | +2.03 m above deck | measured 20 courses of 4″ economy brick; standard 6′-8″ door |
| Roof underside at walls | +4.95 at glass corner, +2.17 at SE and NW corners | measured both glass walls, 6 cm RMS |
| Roof pitch | 23.2° along each wall; 31° (≈7.2:12) true fall, perpendicular to the ridge | derived |
| Ridge | Corner to corner, SW to NE; underside +4.95, level | level assumed NE end never photographed |
| Roof depth / overhang | 0.36 m / 0.60 m all edges, 0.15 m at entry | inferred, high apex fit in #17 |
| Glazing module | ≈1.06 m o.c.; brick panel 1.69 m; narrow lite 0.61 m | measured |
| Purlins | Parallel to ridge, 2.3 m off it, bearing on the wide 0.31 m posts | observed beam ends in #11 and #17 |
| Upper floor | +2.40; 13 risers × 177 mm, 12 treads × 235 mm | inferred, medium |
| Headroom | 2.55 m along ridge in bedroom; 2.32 m over top tread | depends on level ridge |
| Area | ≈42 m² main + ≈9.4 m² upper ≈ 52 m² (≈555 sq ft) gross | inferred |
Iteration log
| Dimension | Weight | R1 | R2 | R3 |
|---|---|---|---|---|
| Form & silhouette | 20% | 82 | 90 | 90 |
| Proportions & scale | 20% | 85 | 91 | 91 |
| Roof geometry | 15% | 78 | 86 | 86 |
| Fenestration | 15% | 85 | 90 | 90 |
| Materials & color | 10% | 70 | 74 | 82 |
| Floor-plan plausibility | 10% | 72 | 75 | 75 |
| Render quality | 10% | 55 | 70 | 76 |
| Total | 77.6 | 84.5 | 85.9 |
Round 1 → 2
Mullion tops in #11 sat 0.07–0.17 m above the model's roof line. Refitting all six upper mullions gave z = 4.97 − 0.427·s, which #17 confirmed independently on the west wall. Ridge raised 4.80 → 4.95 m, eaves lowered to 2.17 m, roof depth 0.30 → 0.36 m to match the apex, and the header band clipped under the eave. The raise also resolved the bedroom-height conflict with #3 (2.4 → 2.55 m). Added surface textures, lighting modes and lamp-lit interiors. Furniture moved per the overhead photo #18.
Round 2 → 3
The glass read white and the soffits glowed. Causes: a studio environment map, and lamps lighting through walls because they cast no shadows. Fixes: a forest-gradient environment in world space, faster lamp falloff, darker glass, weathered deck tone. Geometry unchanged.
Checks that didn't work
A ray-cast of #12's furniture placed the first stair step outside the house, so that camera can't measure depth and was used for the west wall only. A floor homography from #18 failed because the wall bases are hidden behind furniture. I didn't substitute guessed points.
Assumptions and unknowns
| Item | Status | Note |
|---|---|---|
| South and west facades, bay rhythm, slider and gate | observed | Camera-fit verified |
| Diagonal ridge, purlins, corner height | observed | Beam ends and silhouette match in #17 |
| Bedroom windows, two per wall near the NE corner | observed | From inside (#3, #20); exterior trim inferred |
| Square plan, symmetric slopes, 0.6 m overhang | inferred, high | West wall fits the same line as the south |
| Level ridge to NE corner | inferred, medium | Fits bedroom photos; drives upstairs headroom and true elevations |
| Stair position, kitchen on north wall, upper floor +2.40 | inferred, medium | Rails align in #12; headroom checks pass |
| Furniture placement | inferred, medium | ±0.4 m |
| Bathroom location and size, eave store, entry canopy shape | inferred, low | Bathroom never photographed; bedroom is a tight 2.1 m deep |
| North and east exteriors, other windows, foundation, structure sizes | unknown | Excluded from scoring; modeled as plain siding on a crawlspace skirt |
What would close the gap
- One photo of the northeast corner from outside, plus the north and east sides. This settles the ridge height and the unseen walls.
- The bathroom, and the upstairs landing looking back at the bedroom door.
- Total square footage from the listing, as a cross-check on the 6.5 m footprint.
Concept and visualization grade only. Structure, energy code, foundation and permit drawings need a licensed architect or engineer before construction.