Merge pull request #1626 from drsnuggles8/feature/groom-wet-coats-scene-lighting-1615-1618

feat(groom): wet coats and scene lighting for coats (#1615, #1618)

Author: drsnuggles8

.gitattributes

@@ -44,6 +44,8 @@ OloEditor/SandboxProject/Assets/Grooms/Bernese/Bernese.ologroombinding filter=lf
 tools/dog-authoring/sources/Samoyed/Samoyed.bin filter=lfs diff=lfs merge=lfs -text
 tools/dog-authoring/sources/SamoyedLidSource/Samoyed.bin filter=lfs diff=lfs merge=lfs -text
 tools/dog-authoring/sources/Bernese/Bernese.bin filter=lfs diff=lfs merge=lfs -text
+# The golden's coat source (build_dog_groom.py's archive, which its cooked coat records).
+OloEditor/SandboxProject/Assets/Models/Dog/Dog.abc filter=lfs diff=lfs merge=lfs -text
 
 # Mark vendor as vendored (hides from GitHub linguist and other tooling)
 vendor/** linguist-vendored

OloEditor/SandboxProject/Assets/Grooms/Bernese/Bernese.ologroom

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:270c3496fe051acab66abe4b4bfe3fbf31cf764ec848ec0c76ec95e9826afa7e
-size 98542247
+oid sha256:019ab73f2a07844fb1d0fe36f333285040fbb11c0ff967feee3e010abf5157b3
+size 100476730

OloEditor/SandboxProject/Assets/Grooms/Dog/Dog.ologroom

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c0d9fc7945036d8104a309b3710de591b38695c92f0d6316d4aef93afe9cecbe
-size 116251543
+oid sha256:620197de3a2580a6a326cfb7a1bbd1221ce23f3026adbab651f3c3c07de7c8ac
+size 118364531

OloEditor/SandboxProject/Assets/Grooms/Dog/Dog.ologroombinding

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74fbcc6bde87001d168ac8fd79a8481734c237d985db9ba3985a1339cd85fe93
-size 162850982
+oid sha256:24db73294ad4b1bfc953bd5d0a79316ea78b6ee58ac051064777ce1d42ead432
+size 162851014

OloEditor/SandboxProject/Assets/Grooms/Samoyed/Samoyed.ologroom

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b92c9b8fc3eff71ba115b5aaa6f216b5bf1fa67bcecab7d76c8646be3e7cbfa2
+size 57000219

OloEditor/SandboxProject/Assets/Models/Bernese/Bernese.abc

@@ -1,5 +1,25 @@
 {
   "units": "metres, glTF Y-up, the head faces +Z",
+  "assembly": {
+    "pelt": [
+      0.58,
+      0.48,
+      0.38
+    ],
+    "lid": [
+      0.2,
+      0.14,
+      0.1
+    ],
+    "fibre": [
+      1.0,
+      0.93,
+      0.8
+    ],
+    "pigment": "baseColor"
+  },
+  "coatSource": "Assets/Models/Dog/Dog.abc",
+  "walkSpeed": 0.312,
   "eyeRadius": 0.0235,
   "lidInnerRadius": 0.0245,
   "socketRadius": 0.02625,

OloEditor/SandboxProject/Assets/Models/Dog/Dog.abc

@@ -14,15 +14,16 @@ writes it as Alembic hair curves, which the engine cooks into `Grooms/Dog/Dog.ol
 | file | what it is |
 |---|---|
 | `build_dog.py` | the Blender 5.x script that writes everything below, from nothing, and runs the next one |
-| `build_dog_groom.py` | the coat: grown on `Dog.gltf`, written to `.dog-groom/Dog.abc` at the repository root |
+| `build_dog_groom.py` | the coat: grown on `Dog.gltf`, written to `Dog.abc` |
+| `Dog.abc` | the coat's source (Git LFS): one hair object per coat group, which the engine cooks into `Grooms/Dog/Dog.ologroom` |
 | `Dog.gltf` + `Dog.bin` | one skinned mesh (~256k triangles) on a 50-bone rig, eight materials, six clips |
 | `DogCoatColor.png` | 2048² coat colours: DogSkin's albedo and the groom's colour map |
 | `DogNoseNormal.png`, `DogNoseOcclusion.png` | 2048² each, the nose leather's own UV square: its cobblestones, and the occlusion that keeps its nostrils dark |
 | `DogTongueColor.png`, `DogTongueNormal.png`, `DogTongueOcclusion.png` | 1024² each, the tongue's own UV square: its colour (redder root and edges, paler tip, the median groove, the papillae's pale tips, a veined underside), its filiform papillae, and its occlusion |
 | `DogGumColor.png`, `DogGumOcclusion.png` | 1024² each: pink gums with black pigment patches, and the oral cavity's occlusion |
 | `DogLipOcclusion.png`, `DogTeethOcclusion.png` | 1024² each: the lips' and teeth's occlusion, all of the mouth's baked in the Pant pose |
 | `DogIrisColor.png` | 2048², the eyeball's albedo on a planar UV across the gaze: the painted iris (fetched at the refracted iris point), its crisp pupil, and the sclera round it |
-| `Dog.rig.json` | eye radius, eye centres and gaze, lid and socket radii, the clip list |
+| `Dog.rig.json` | eye radius, eye centres and gaze, lid and socket radii, the clip list, the dressing (`assembly`: pelt, lid and fibre colours and the pigment mode), the coat source and the walk speed |
 | `DogEyeball.gltf` + `.bin` | the unit sphere each eye entity draws (128 x 64: the engine's primitive sphere reloads at 16 segments) |
 
 Units are metres, Y-up, and the head faces +Z, the convention `Horse.gltf` and `Fox.gltf` use. The
@@ -106,9 +107,18 @@ undercoat, guard hair, long hair and strays per region, Voronoi locks, darkened
 sun-bleached tips. A last pass, `keep_off_other_parts`, cuts any strand short of the other side
 of an ear's edge: the face and cheek fur under a flap otherwise grew straight through it (52% of
 the face's guard hair) and showed as a slit of hair at the flap's front edge. Fur dipping into the
-skin it grows on is left alone. It writes one hair object per coat group into `.dog-groom/Dog.abc`: ~410k
-strands, 57 MB. The archive is a build output, git-ignored, and it stays out of `SandboxProject/Assets`,
-because the editor registers everything it finds there.
+skin it grows on is left alone. It writes one hair object per coat group into `Dog.abc`, beside
+the model: ~410k strands, 57 MB, the coat's source, committed through Git LFS as the other breeds'
+are. `Dog.rig.json` names it (`coatSource`), the shipped coat records it as its provenance, and the
+Content Browser's "Import as Groom" cooks it into the same bytes
+(`TheShippedCoatIsWhatTheEditorImportWrites`). The archive also carries the card tier
+(`groom_card_cell_size`, `groom_card_points`: `CARDS`), so every cook of it builds the same cards.
+Root UVs are written in the engine's convention, Blender's v up (#1560).
+
+Blender's Alembic exporter writes the hair objects in a different order on every run, and the
+importer numbers the groups in archive order, so a regenerated archive is the same coat in another
+curve order: every curve-indexed jitter and stride-based LOD subset moves. Cook the shipped coat from
+the committed archive, and regenerate the archive only to change the coat.
 
 Blender's Alembic exporter writes a hair object's points and widths and its custom properties, and
 nothing else a groom needs. So each group's coat description (`groom_role`, `groom_tint`,

OloEditor/SandboxProject/Assets/Models/Dog/Dog.rig.json

@@ -11,7 +11,8 @@
                          Sit, Walk and Pant (30 fps, in place)
     DogCoatColor.png     the coat colours, both DogSkin's albedo and the groom's colour map
     Dog.rig.json         the eye radius, centres and gaze (the engine draws each eye as its own
-                         sphere entity attached to eye_L / eye_R), the clip list
+                         sphere entity attached to eye_L / eye_R), the clip list, the dressing
+                         (ASSEMBLY), the coat's source (COAT_SOURCE) and the walk's speed
     DogEyeball.gltf      the unit sphere each eye entity draws, at close-up tessellation
 
 Nothing is hand-edited after generation (issue #1533, criterion A6). The surface is a signed
@@ -3365,6 +3366,16 @@ def to_gltf(v):
     return [float(v[0]), float(v[2]), float(-v[1])]
 
 
+# How the golden is dressed (the engine's DogShowcaseEvidenceTest and the exported Dog.olo read it
+# from Dog.rig.json's "assembly", as they do the other breeds'), linear RGB: DogSkin's and DogLid's
+# base colours, which multiply the coat map, and the strands' base colour under it. "baseColor"
+# pigment tints the light the fibre returns, as #1533 graded the golden.
+ASSEMBLY = {"pelt": [0.58, 0.48, 0.38], "lid": [0.20, 0.14, 0.10], "fibre": [1.0, 0.93, 0.80], "pigment": "baseColor"}
+# Where the coat's Alembic source is (build_dog_groom.py's default_abc_path), as the cooked coat
+# records it: project-relative, beside the model, where the Content Browser imports it.
+COAT_SOURCE = "Assets/Models/Dog/Dog.abc"
+
+
 def write_rig_json(path):
     import json
 
@@ -3374,6 +3385,9 @@ def write_rig_json(path):
         eyes[side] = {"bone": f"eye_{side}", "centre": to_gltf(c), "gaze": to_gltf(f), "up": to_gltf(up)}
     data = {
         "units": "metres, glTF Y-up, the head faces +Z",
+        "assembly": ASSEMBLY,
+        "coatSource": COAT_SOURCE,
+        "walkSpeed": float(WALK_SPEED),
         "eyeRadius": float(EYE_R),
         "lidInnerRadius": float(EYE_R + LID_GAP),
         "socketRadius": float(EYE_R + SOCKET_CLEAR),
@@ -3647,8 +3661,8 @@ def main():
     """blender -b --factory-startup --python build_dog.py -- [out_dir] [preview_dir]
 
     Writes Dog.gltf + Dog.bin + DogCoatColor.png + DogNoseNormal.png + DogNoseOcclusion.png +
-    Dog.rig.json into out_dir, then the coat (build_dog_groom.py) into .dog-groom/Dog.abc (default: this
-    script's directory) and, when preview_dir is given, Workbench renders of the rest pose,
+    Dog.rig.json into out_dir, then the coat (build_dog_groom.py) into Dog.abc beside this script
+    (Assets/Models/Dog/, the rig's coatSource) and, when preview_dir is given, Workbench renders of the rest pose,
     every clip, a blink and the pant."""
     import bpy