diff --git a/README.md b/README.md index e47430e..bdc1775 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # CubeAssembler -Scan a physical 2×2–7×7 puzzle cube from six photos and assemble its state in +Scan a physical 2×2–15×15 puzzle cube from six photos and assemble its state in your browser. [Open the scanner](https://wstein.github.io/cube-assembler/) · @@ -19,7 +19,7 @@ with no model download or production server. ## Scan a cube in five steps -1. **Choose a cube and colors.** Select a size from 2×2 to 7×7. Automatic +1. **Choose a cube and colors.** Select a size from 2×2 to 15×15. Automatic colors compares the live face with built-in and saved palettes, or you can choose a palette yourself. 2. **Capture six faces.** Show four sides while turning the cube, then its top @@ -40,7 +40,7 @@ capture, since different faces can look identical on even cubes. ## What works - Camera capture or six uploaded photos, cropped or uncropped. -- Cube sizes 2×2–7×7, with manual color correction when needed. +- Cube sizes 2×2–15×15, with manual color correction when needed. - Physical-validity checks and an unfolded net plus 3D view. - WRG/URF facelets and Orbit64 compact state tokens. - Local, browser-only scanning; an optional development server saves fixtures. diff --git a/docs/color-profiles.md b/docs/color-profiles.md index 37d3836..3d88c9e 100644 --- a/docs/color-profiles.md +++ b/docs/color-profiles.md @@ -2,7 +2,7 @@ **Cube** and **Colors** are separate settings. A cube records its name, size, and sticker gap. A color profile records six sticker colors and can be reused -across sizes. The Generic cube offers each size from 2×2 to 7×7 with a 60% +across sizes. The Generic cube offers each size from 2×2 to 15×15 with a 60% sticker core and a 40% total gap. **New cube** copies the selected size and gap so you can name and adjust a variant. Built-in cubes are read-only. diff --git a/docs/parity.md b/docs/parity.md index 5778fe0..21f7040 100644 --- a/docs/parity.md +++ b/docs/parity.md @@ -15,12 +15,12 @@ For a 3×3, four conditions are necessary and sufficient: 4. Corner and edge permutation parity agree. All cube sizes have eight corner cubies, so the corner color-triplet and -orientation checks apply from 2×2 through 7×7. A 2×2 has no edges or centers; +orientation checks apply on every size from 2×2 up. A 2×2 has no edges or centers; its complete model comes from corner checks alone. ## Larger cubes -On 4×4–7×7, each edge has N−2 wing pieces. Every wing's two colors must be +On cubes of 4×4 and larger, each edge has N−2 wing pieces. Every wing's two colors must be one of the 12 canonical neighboring color pairs, and each pair must occur exactly N−2 times. Wings are counted across depths; the validator does not yet check the distinct depth orbits or generalize full edge permutation parity diff --git a/docs/scanner.md b/docs/scanner.md index 9c79d1f..4d18e74 100644 --- a/docs/scanner.md +++ b/docs/scanner.md @@ -13,8 +13,8 @@ it. [`gridAlignment.ts`](../src/client/vision/gridAlignment.ts) estimates an out before its fine seam search, or searches around the centered guide when an outline is unclear. The outline scan covers faces 0.7–1.3 times guide size with centers up to 22% of a guide width away. The seam search stays within -half a cell of that estimate to avoid slipping a row on 5×5–7×7 faces. -For 6×6 and 7×7, an additional outer-edge check rejects seam fits whose +half a cell of that estimate to avoid slipping a row on faces of 5×5 and larger. +For 6×6 and larger, an additional outer-edge check rejects seam fits whose grid lines miss the face. The live view then asks you to move the face toward the center instead of accepting the shifted reading. diff --git a/docs/user-guide.md b/docs/user-guide.md index ad658b2..172f06d 100644 --- a/docs/user-guide.md +++ b/docs/user-guide.md @@ -2,7 +2,11 @@ [Open CubeAssembler](https://wstein.github.io/cube-assembler/). Scanning, review, assembly, and ZIP import/export run in the browser. Choose a cube size from 2×2 -through 7×7 before taking photos. +through 15×15 before taking photos. Face detection has been tested on real camera +photos of 2×2, 3×3, 4×4, 5×5 and 7×7 cubes. Sizes 6 and 8 to 15 have only +synthetic images, and only sizes 8, 9, 10, 12 and 15 were measured. Upright +faces are mostly found and read correctly; from 9×9 on, a tilted face often +loses the grid or shifts it, which gives wrong colors. Hold the face upright. ## Scan a cube @@ -106,7 +110,7 @@ Turns snap into place like a magnetic cube: while you drag a layer or the whole softly when capture **Sound** is on. Layer turns update the 2D net, facelet notation, and Orbit64 token. The move list records completed turns; **Undo** reverses the last one, and **Reset** returns to the captured state. -**Scramble** uses more moves for larger cubes and turns inner layers on 4×4–7×7, +**Scramble** uses more moves for larger cubes and turns inner layers on cubes of 4×4 and larger, while leaving the fixed middle layer alone on odd cubes. **Random State** mixes the cube at once into a random valid state, as speed cubers do for practice: the cube spins once and its stickers change at the @@ -133,7 +137,7 @@ face images are read directly. In Guide grid mode, select **Cropped face** for an image that is already cropped but does not use a `face-*.jpg` filename. **One picture of a cube net** works too: a screenshot or export of the cross -(U above F, L F R B in a row, D below) of a 2×2 to 7×7 cube, on a dark or +(U above F, L F R B in a row, D below) of a 2×2 to 15×15 cube, on a dark or light background, with or without gaps, lines or face titles. The app finds the stickers and the size, cuts the picture into six face images and goes on as with six photos, so you see the order preview and the cube check as usual. Odd cubes diff --git a/index.html b/index.html index b7508d1..22ee522 100644 --- a/index.html +++ b/index.html @@ -3,7 +3,7 @@
- +