Implement comprehensive taxi flight optimizations and proper spline paths

Major improvements:
- Load TaxiPathNode.dbc for actual curved flight paths (no more flying through terrain)
- Add 3-second mounting delay with terrain precaching for entire route
- Implement LOD system for M2 models with distance-based quality reduction
- Add circular terrain loading pattern (13 tiles vs 25, 48% reduction)
- Increase terrain cache from 2GB to 8GB for modern systems

Performance optimizations during taxi:
- Cull small M2 models (boundRadius < 3.0) - not visible from altitude
- Disable particle systems (weather, smoke, M2 emitters) - saves ~7000 particles
- Disable specular lighting on M2 models - saves Blinn-Phong calculations
- Disable shadow mapping on M2 models - saves shadow map sampling and PCF

Technical details:
- Parse TaxiPathNode.dbc spline waypoints for curved paths around terrain
- Build full path from node pairs using TaxiPathEdge lookup
- Precache callback triggers during mounting delay for smooth takeoff
- Circular tile loading uses Euclidean distance check (dx²+dy² <= r²)
- LOD fallback to base mesh when higher LODs unavailable

Result: Buttery smooth taxi flights with no terrain clipping or performance hitches
This commit is contained in:
Kelsi 2026-02-08 21:32:38 -08:00
parent 941dac446d
commit 536b3cea48
9 changed files with 249 additions and 32 deletions

View file

@ -76,6 +76,7 @@ public:
bool isSitting() const { return sitting; }
bool isSwimming() const { return swimming; }
bool isInsideWMO() const { return cachedInsideWMO; }
bool isOnTaxi() const { return externalFollow_; }
const glm::vec3* getFollowTarget() const { return followTarget; }
glm::vec3* getFollowTargetMutable() { return followTarget; }

View file

@ -35,6 +35,7 @@ struct M2ModelGPU {
uint16_t textureAnimIndex = 0xFFFF; // 0xFFFF = no texture animation
uint16_t blendMode = 0; // 0=Opaque, 1=AlphaKey, 2=Alpha, 3=Add, etc.
uint16_t materialFlags = 0; // M2 material flags (0x01=Unlit, 0x04=TwoSided, 0x10=NoDepthWrite)
uint16_t submeshLevel = 0; // LOD level: 0=base, 1=LOD1, 2=LOD2, 3=LOD3
glm::vec3 center = glm::vec3(0.0f); // Center of batch geometry (model space)
float glowSize = 1.0f; // Approx radius of batch geometry
};
@ -58,6 +59,7 @@ struct M2ModelGPU {
bool collisionTreeTrunk = false;
bool collisionNoBlock = false;
bool collisionStatue = false;
bool isSmallFoliage = false; // Small foliage (bushes, grass, plants) - skip during taxi
// Collision mesh with spatial grid (from M2 bounding geometry)
struct CollisionMesh {
@ -310,9 +312,11 @@ public:
void clearShadowMap() { shadowEnabled = false; }
void setInsideInterior(bool inside) { insideInterior = inside; }
void setOnTaxi(bool onTaxi) { onTaxi_ = onTaxi; }
private:
bool insideInterior = false;
bool onTaxi_ = false;
pipeline::AssetManager* assetManager = nullptr;
std::unique_ptr<Shader> shader;

View file

@ -166,6 +166,12 @@ public:
*/
void unloadAll();
/**
* Precache a set of tiles (for taxi routes, etc.)
* @param tiles Vector of (x, y) tile coordinates to preload
*/
void precacheTiles(const std::vector<std::pair<int, int>>& tiles);
/**
* Set streaming parameters
*/
@ -294,7 +300,7 @@ private:
std::unordered_map<TileCoord, CachedTile, TileCoord::Hash> tileCache_;
std::list<TileCoord> tileCacheLru_;
size_t tileCacheBytes_ = 0;
size_t tileCacheBudgetBytes_ = 2ull * 1024 * 1024 * 1024; // 2GB default
size_t tileCacheBudgetBytes_ = 8ull * 1024 * 1024 * 1024; // 8GB for modern systems
std::mutex tileCacheMutex_;
std::shared_ptr<PendingTile> getCachedTile(const TileCoord& coord);