Direct Gen~ code translation implementing complete Huovilainen model.
This implementation represents the most faithful translation of the original Max/MSP Gen~ moogLadderFilter code to C++, preserving every mathematical operation, coefficient, and processing step from the visual programming environment. It serves as the reference implementation for validating other optimized versions while maintaining bit-exact compatibility with the original Gen~ code.
@translation_methodology The translation process involved systematic conversion of Gen~ operators to C++ equivalents:
Gen~ Operator Mapping:
- history operators → class member variables for state storage
- param objects → function parameters or class members
- expr calculations → inline mathematical expressions
- clamp operators → custom clamp functions with exact behavior
- fixdenorm → denormal number protection functions
Preservation Priorities:
- Bit-exact Output: Identical results to Gen~ version
- Coefficient Fidelity: All empirical constants preserved exactly
- Processing Order: Maintained original calculation sequence
- State Management: Exact replication of delay line behavior
- Parameter Mapping: Identical control value interpretation
@research_validation_importance This scalar implementation serves critical research functions:
- Reference Standard: Validates optimization correctness
- Algorithm Documentation: Preserves original research implementation
- Performance Baseline: Establishes unoptimized performance metrics
- Educational Tool: Demonstrates direct translation methodology
- Debugging Reference: Provides known-good implementation for comparison
@implementation_characteristics
- Processing Model: Sample-by-sample scalar operations
- State Variables: Direct mapping from Gen~ history operators
- Coefficient Precision: Double-precision where original used it
- Parameter Control: Exact replication of Gen~ parameter behavior
- Memory Layout: Optimized for cache locality despite scalar processing
@performance_profile
- Computational complexity: ~120-150 floating-point operations per sample
- Memory footprint: ~80 bytes (20 float state variables)
- Branch prediction: Moderate branching from conditional operations
- Cache behavior: Sequential access pattern, cache-friendly
- Optimization potential: Excellent baseline for vectorization analysis
- Author
- [Implementation team based on original Gen~ code]
- Date
- 2025
- Version
- Final (matching Gen~ structure with corrected rc/cutoff logic)
1.14.0