# Terrace Tower · source research and next-package brief

PI-BLD-0003-V01 · legacy 03-A · 16 October 2026 · **CONCEPT / NOT TO COMMON SCALE**

Research baseline: main `aa3b95526ce18bd7c74cc0a232c62e1b0f82535b`. This brief prepares the following building while Split Complex and the isolated sample advance. It does not change the asset's dimension contract, issue a pack, select a controlling view, accept a shell, or authorize district construction. Current decisions §7 and the foundation hold still control the work.

## Sources inspected

All repository sources below were read at the pinned baseline. Paths and links identify evidence; this document is a research companion, not a new source of dimensional authority.

| Source | What it establishes |
| --- | --- |
| [Asset README](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/README.md) | Concept-only P2 asset; no issued pack, dimensions, shell or placement; proposed Tier C lobby/dwelling/shared-terrace scope; no named story assignment. |
| [Visual notes](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/VISUAL_NOTES.md) and [reference index](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/REFERENCE_INDEX.csv) | Card 02, row 1 column 3 is the documented asset view. No dedicated location render is listed. Card 01 is context only. |
| [Asset brief](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/ASSET_BRIEF.json) and [dimension contract](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/BUILDING_DIMENSIONS.json) | Envelope, floors, core, entrance, player/camera values, reference authority, shell and approval remain unset. PI-DEV-VICE is a production proposal; enterability is null. |
| [Preparation R01](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/assets/buildings/PI-BLD-0003-V01/PREP/R01/README.md) | Resolve stepped slabs, continuous core, reusable dwelling conditions, planting/wet interfaces and terrace access before claiming fit. |
| [Shared plan](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/docs/crown-key-production/R01/PLAN.json) and local `SPLIT_REFERENCE_REVIEW.md` | Isolated sample and Quik feedback precede package scaling; Split Complex comes before Terrace Tower. Research can proceed without claiming completed tooling or tests. |
| [Building R05](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/protocols/building/R05/GUIDE.md), [Spatial Imagery R01](https://github.com/215GrandAve/lost-souls/blob/aa3b95526ce18bd7c74cc0a232c62e1b0f82535b/production/workflows/spatial-imagery/R01/GUIDE.md), `docs/current-decisions.md` §7 | Keep dimension status, measured model/camera provenance, named source hierarchy and independent shell/engine acceptance. |

The uploaded `Crown_Key_Final_Card_Set.zip` and local `reference_review/02_Towers_and_Variants_Global_IDs.png` were checked. The sheet bytes match the index SHA-256: `a84ef3a27eea546ea48c3c154dc4d23c00e0230395dd87b48d221c11e1eb269b`. The image was visually inspected. Its footer says bare architecture, vegetation separate, not to common scale. Card 01 was not visually inspected for this brief; no parcel location is inferred from it.

## Visual invariants

Preserve these observed relationships in concept studies. Directions below describe the **card view**, not a selected building front or world orientation.

- A taller, rounded blue-glass volume rises at the right/rear; the lower mass descends toward the left/front. Keep that asymmetric composition legible.
- Broad warm-ivory floor bands sweep into soft curved/S-shaped terrace edges. The cascading mass is not a stack of rectangular garden boxes.
- Recessed glazing and warm amber interior light sit behind the terrace bands. Keep the contrast between blue glazing, ivory structure and warm occupied spaces.
- The ground treatment is broad, curved and arched, with a low wrapping deck/base. The small glass volume at the right remains a distinct unresolved element.
- The upper tower is comparatively restrained and has a flat top/small rooftop expression in the card. A Crown Landmark crown, Split Complex bridge or Twist Tall helix would change this building's identity.
- Planting can accompany the architecture, but remains a separate scope. Do not count vegetation as slab geometry or silently expand it into new shared amenities.

The card shows several terrace bands; the notes describe about six visible stepped levels. Neither observation establishes floor count, floor-to-floor height, terrace depth, structural thickness or a complete level schedule. The rear and hidden sides remain unobserved.

## Candidate massing, core and setback approach

These are design questions and proposed study directions, not selected geometry. Author no dimensions from the perspective image.

**Massing study:** begin with two explicitly identified zones: rounded taller tower and cascading terrace mass. Add the lower arched base and right-side glass volume as separately traceable components. A neutral conceptual diagram should reveal which connections are proposed, rather than making an unseen rear facade appear source-confirmed. Compare a single connected building against a connected podium with distinct upper masses only if that difference affects circulation or shell reuse.

**Core study:** test a continuous core within the taller volume as the first candidate. Show how each selected terrace floor reaches it. A core that terminates at a setback, a route that crosses a private dwelling, or lifts/stairs that shift without an explained transfer cannot be hidden by finish imagery. A separate secondary core or circulation link is an alternative to evaluate only if the chosen massing/access program needs it. Core location, count, lift/stair arrangement and dimensions remain open.

**Setback study:** record each distinct slab outline and elevation, rather than deriving an arbitrary repeated offset from the card. Distinguish terrace usable area, planting/reserved zones, structural intrusions, doors and circulation. Keep the soft curved edge signature while checking the narrowest usable depth and access to each selected terrace. No universal unit is assumed to fit all steps.

**Envelope/interface study:** annotate candidate planting build-up, drainage/wet zones, door thresholds, guards and maintenance access in section as unresolved interfaces. The game geometry needs clear collision and playable boundaries; a concept drawing does not certify real-world waterproofing or structural design. Avoid assigning a pool or extensive roof program without creator selection.

**Interior study:** retain proposed Tier C only as a candidate: lobby, representative terrace dwelling and selected shared terrace. First compare a typical dwelling against a setback-edge condition. Reuse is demonstrated by fit and access, not by scaling a room to every floor. Keep unselected floors and exterior-only areas explicit.

## Conflicts and decisions still open

No direct image-to-image conflict can be established from the single documented asset view. The gaps below are incomplete evidence, not permission to improvise accepted geometry.

| Question | Current evidence / risk | Owner and next evidence |
| --- | --- | --- |
| Controlling exterior and front | Reference authority and orientation are null; card view is three-quarter. | Legend selects visual authority; layout author supplies annotated orientation/massing study. |
| Connected building or separate masses | Tower/terrace relationship reads visually connected, but circulation and structural joins are unseen. | Design/layout author compares topology; Legend selects. |
| Tower plan and hidden faces | Rounded facade does not distinguish circular, oval or faceted plan. | Layout author proposes plan and rear/side studies, marked newly designed. |
| Floor and setback sequence | Visible terrace bands are not a schedule. | Codex/layout owner proposes explicit levels/slab outlines later; Legend reviews the contract. |
| Core continuity and access | No core, entrance clearance or route dimensions exist. | Layout author provides plans/sections; Quik later returns measured traversal evidence. |
| Small right-side glass volume and roof | Connection, use and rear extent unknown. | Legend selects scope from annotated options. |
| Private/shared/playable terraces | Tier C is proposed; enterability is null. | Legend selects accessible spaces and limits before detailed room/route work. |
| Planting and wet interfaces | Required coordination topic, no technical details issued. | Design author separates reserved zones, playable areas and component interfaces for review. |
| Similar terraced family | PI-BLD-0005-V02 uses straight decks and diagonal frame members. | Preserve separate global identities; do not substitute that design for Terrace Tower. |
| Parcel and scale | No placement, metric envelope or frame revision selected. | Reconcile the foundation and measured sources before final placement/acceptance. |

## Required authored views and measurements

After the selected reference and proposed contract exist, derive the following from **one model revision**, with units, front/north convention, level IDs, section cuts and saved cameras. A generated image that resembles an orthographic drawing is still concept art without that provenance.

| View / drawing | What must be testable |
| --- | --- |
| Ground/site and roof plans | Envelope, entrance, tower/terrace/right-volume relationship, roof scope and proposed parcel interface. |
| Each distinct terrace floor plus typical-floor instance list | Slab outline, setback from the level below, clear dwelling area, core/corridor access, real openings, terrace depth and planted/reserved zones. |
| Section through taller tower and cascading mass | Core continuity, floor elevations, clear heights, slab/planting build-up, setback transitions and any circulation transfer. |
| Section through representative terrace doorway and edge | Clear opening, threshold/collision, furnishing/camera clearance, planting separation, guard and wet-zone interface. |
| Front/rear/left/right elevations | Complete proposed envelope and explicit distinction between observed card features and authored hidden faces. |
| Full-building three-quarter plus reverse three-quarter | Card silhouette comparison and visibility of the newly resolved rear/side relationships. |
| Furnished dwelling and shared-terrace access views | Selected kit fit, private/public route separation and player/camera usability. |

Keep measured evidence for envelope bounds; distinct slab footprints/elevations; core/shaft continuity; narrowest clear routes and terrace zones; entrance/door clear openings; actor/component scale and import orientation. Actual values remain in the asset contract, with status and basis. Do not duplicate proposed numbers in this brief. Until district placement is established, window context stays neutral or explicitly provisional.

## Work after sample feedback

1. Carry Quik's actual import, orientation, traversal and package-usefulness feedback into the shared standard. A successful sample does not accept Terrace Tower or Crown Key.
2. Complete the Split Complex pilot and capture the revisions that make its contract/drawings/handoff useful. Keep Terrace research bounded so the unbuilt backlog stays small.
3. Present Terrace's controlling-reference, massing connection, orientation and accessible-scope decisions to Legend. Resolve the minimum needed to author a coherent proposed contract.
4. Author proposed envelope, level/setback schedule, continuous core, selected room/terrace scope and interface values in the existing per-asset contracts. Keep unknowns null; trace every proposed value and approval separately.
5. Produce deterministic plans/sections and a neutral blockout. Test the representative dwelling and the most constrained selected terrace/access condition before decorating every level.
6. Create concept/detail imagery against the selected reference and, when available, the saved model/cameras. Record overlay differences before promoting any view to ALIGNED REVIEW or ACCEPTED BUILD REFERENCE.
7. Hand Quik a small primary build list, editable sources, schedules, open issues and manifest. Issue status, shell review, actual engine verification and district placement remain separate decisions.

**Current result:** source-backed research complete; no new model, dimension values, imagery, acceptance or engine test is claimed by this document.
