Intelligent Adaptive De-Essing with Automatic Sibilance Tracking

Biliousis S, Erkut C. Intelligent Adaptive De-Essing with Automatic Sibilance Tracking. In: Convention Paper. Audio Engineering Society; 2026. Paper 10268. Available from: https://aes.org/publications/elibrary-page/?id=23182

@inproceedings{Biliousis2026_23182,
  author = {Biliousis, Stefanos and Erkut, Cumhur},
  title = {{Intelligent Adaptive De-Essing with Automatic Sibilance Tracking}},
  note = {Paper 10268},
  year = {2026},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=23182}
}

TY  - CPAPER
TI  - Intelligent Adaptive De-Essing with Automatic Sibilance Tracking
AU  - Biliousis, Stefanos
AU  - Erkut, Cumhur
M1  - Paper 10268
PY  - 2026
DA  - 2026/05/28
UR  - https://aes.org/publications/elibrary-page/?id=23182
PB  - Audio Engineering Society
LA  - en
AB  - High-fidelity vocal processing is frequently compromised by sibilance, a phenomenon characterized by stochastic high-frequency energy that presents unique dynamic range challenges. While traditional de-essing techniques often rely on static frequency bands, they fail to account for inter-speaker variability and changing dynamics. This paper presents an adaptive real-time de-essing application, developed using the JUCE framework, which automatically detects and suppresses sibilant frequencies. The proposed methodology integrates a derivative-based frequency tracking algorithm to estimate a spectral-centroid-like statistic without the computational overhead of the Fast Fourier Transform (FFT). This is coupled with a dual-path envelope detection system and a relative threshold logic to distinguish sibilance from the wideband signal. Additionally, a dynamic harmonic exciter is implemented to restore high-frequency presence during non-sibilant periods. Objective spectral analysis confirms the systems ability to selectively attenuate energy in the 211 kHz range while maintaining spectral transparency and minimizing artifacts. Across three vocal files, the adaptive method improves attenuation-frequency alignment relative to static processing by an average of 99.58 Hz, supporting the core contribution of automatic center-frequency tracking under speaker-dependent sibilance variation.
ER  -