|
MayaFlux 0.5.0
Digital-First Multimedia Processing Framework
|
|
inline |
Set the scalar multiplier applied to the network's output buffer after processing.
| scale | Linear scale factor (1.0 = unity relative to the network's own output normalization, 0.0 = silence, >1.0 = amplify) |
The meaning of unity depends on the network type. ResonatorNetwork sums all resonator outputs and divides by sqrt(resonator_count) (RMS normalization) before this scale is applied, keeping perceived loudness roughly stable as resonators are added or removed. This is a statistical approximation: it assumes resonators are not strongly phase-correlated. When resonators share a single exciter and are tuned closely (the common formant-synthesis case), their peaks can align more than RMS assumes, so true worst-case output can exceed what RMS normalization implies. FinalLimiterProcessor at the root audio buffer remains the actual safety backstop regardless of this scale.
If a specific relationship to raw resonator count is needed (e.g. compensating toward the sum a single resonator alone would produce), scale relative to get_node_count(): scale ~= sqrt(get_node_count()) approaches raw uncompensated sum, scale ~= 1.0 / sqrt(get_node_count()) approaches per-resonator average.
Definition at line 309 of file NodeNetwork.hpp.
References scale.