| Age | Commit message (Collapse) | Author |
|---|
|
Converted ShaderToy shader (Saturday cubism experiment) to Scene1Effect
following EFFECT_WORKFLOW.md automation guidelines.
**Changes:**
- Created Scene1Effect (.h, .cc) as scene effect (not post-process)
- Converted GLSL to WGSL with manual fixes:
- Replaced RESOLUTION/iTime with uniforms.resolution/time
- Fixed const expressions (normalize not allowed in const)
- Converted mainImage() to fs_main() return value
- Manual matrix rotation for scene transformation
- Added shader asset to workspaces/main/assets.txt
- Registered in CMakeLists.txt (both GPU_SOURCES sections)
- Added to demo_effects.h and shaders declarations
- Added to timeline.seq at 22.5s for 10s duration
- Added to test_demo_effects.cc scene_effects list
**Shader features:**
- Raymarching cube and sphere with ground plane
- Reflections and soft shadows
- Sky rendering with sun and horizon glow
- ACES tonemapping and sRGB output
- Time-based rotation animation
**Tests:** All effects tests passing (5/9 scene, 9/9 post-process)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
|
Implements multi-layer convolutional neural network shader for stylized
post-processing of 3D rendered scenes:
**Core Components:**
- CNNEffect: C++ effect class with single-layer rendering (expandable to multi-pass)
- Modular WGSL snippets: cnn_activation, cnn_conv3x3/5x5/7x7, cnn_weights_generated
- Placeholder identity-like weights for initial testing (to be replaced by trained weights)
**Architecture:**
- Flexible kernel sizes (3×3, 5×5, 7×7) via separate snippet files
- ShaderComposer integration (#include resolution)
- Residual connections (input + processed output)
- Supports parallel convolutions (design ready, single conv implemented)
**Size Impact:**
- ~3-4 KB shader code (snippets + main shader)
- ~2-4 KB weights (depends on network architecture when trained)
- Total: ~5-8 KB (acceptable for 64k demo)
**Testing:**
- CNNEffect added to test_demo_effects.cc
- 36/36 tests passing (100%)
**Next Steps:**
- Training script (scripts/train_cnn.py) to generate real weights
- Multi-layer rendering with ping-pong textures
- Weight quantization for size optimization
handoff(Claude): CNN effect foundation complete, ready for training integration
|
|
Multi-input composite shaders with sampler support.
- Dynamic bind group layouts (N input textures + 1 sampler)
- dispatch_composite() for multi-input compute dispatch
- create_gpu_composite_texture() API
- gen_blend.wgsl and gen_mask.wgsl shaders
Guarded with #if !defined(STRIP_GPU_COMPOSITE) for easy removal.
Tests:
- Blend two noise textures
- Mask noise with grid
- Multi-stage composite (composite of composites)
Size: ~830 bytes (2 shaders + dispatch logic)
handoff(Claude): GPU procedural Phase 4 complete
|
|
Complete Phase 2 implementation:
- gen_perlin.wgsl: FBM with configurable octaves, amplitude decay
- gen_grid.wgsl: Grid pattern with configurable spacing/thickness
- TextureManager extensions: create_gpu_perlin_texture(), create_gpu_grid_texture()
- Asset packer now validates gen_noise, gen_perlin, gen_grid for PROC_GPU()
- 3 compute pipelines (lazy-init on first use)
Shader parameters:
- gen_perlin: seed, frequency, amplitude, amplitude_decay, octaves (32 bytes)
- gen_grid: width, height, grid_size, thickness (16 bytes)
test_3d_render migration:
- Replaced CPU sky texture (gen_perlin) with GPU version
- Replaced CPU noise texture (gen_noise) with GPU version
- Added new GPU grid texture (256x256, 32px grid, 2px lines)
Size impact:
- gen_perlin.wgsl: ~200 bytes (compressed)
- gen_grid.wgsl: ~100 bytes (compressed)
- Total Phase 2 code: ~300 bytes
- Cumulative (Phase 1+2): ~600 bytes
Testing:
- All 34 tests passing (100%)
- test_gpu_procedural validates all generators
- test_3d_render uses 3 GPU textures (noise, perlin, grid)
Next: Phase 3 - Variable dimensions, async generation, pipeline caching
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
|
Phase 1 implementation complete:
- GPU compute shader for noise generation (gen_noise.wgsl)
- TextureManager extensions: create_gpu_noise_texture(), dispatch_noise_compute()
- Asset packer PROC_GPU() syntax support with validation
- ShaderComposer integration for #include resolution
- Zero CPU memory overhead (GPU-only textures)
- Init-time and on-demand generation modes
Technical details:
- 8×8 workgroup size for 256×256 textures
- UniformBuffer for params (width, height, seed, frequency)
- Storage texture binding (rgba8unorm, write-only)
- Lazy pipeline compilation on first use
- ~300 bytes code (Phase 1)
Testing:
- New test: test_gpu_procedural.cc (passes)
- All 34 tests passing (100%)
Future phases:
- Phase 2: Add gen_perlin, gen_grid compute shaders
- Phase 3: Variable dimensions, async generation
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
|
- Implemented VignetteEffect, including its shader, parameters, and sequence integration.
- Added VignetteEffect to demo_effects.h, shaders.cc/h, and asset definitions.
- Updated seq_compiler to handle VignetteEffect parameters.
- Added VignetteEffect to test suite and updated expected counts.
- Ensured all changes build and tests pass.
- Added vignette_effect.cc implementation file.
- Updated CMakeLists.txt to include the new effect file.
- Updated assets/demo.seq to include the VignetteEffect.
- Updated assets/final/demo_assets.txt with the new shader asset.
|
|
Fixes crashes in demo64k and test_3d_render caused by uninitialized ShaderComposer. Moves InitShaderComposer() call before effect initialization in gpu.cc and adds explicit call in test_3d_render.cc. Also fixes include paths for generated assets.h in multiple files.
|
|
Split src/gpu/demo_effects.cc into individual .cc files for each effect
and created separate files for post-processing helpers and WGSL shaders.
Updated src/gpu/demo_effects.h to be a central header for all effect-related
declarations and adjusted CMakeLists.txt accordingly.
|
