An intelligent dynamic range compression (DRC) algorithm, using the CA-DAFX processing architecture, produces the optimal amount of dynamic range for multitrack recordings. The algorithm exploits the interdependence of input audio features, incorporates best practices, and uses subjective evaluation. The classical parameters of a typical compressor (ratio, threshold, knee, attack, and release) are dynamically adjusted depending on extracted features and control rules. Two new audio weighting features, percussiveness and low-frequency strength, were proposed to incorporate the transient nature and spectral content of the signal. The authors applied multiple linear regression models to the subjective results to formulate the ratio and threshold automations that follow the choices of the human operators. The results showed that the algorithm can compete with or outperform semiprofessional mixes in terms of four different perceptual criteria: the appropriateness of the amount of DRC applied, the degree of imperfection, the ability to stabilize the erratic level fluctuations, and overall Preferences.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.