De-essing (also desibilizing) is any technique intended to reduce or eliminate the excessive prominence of sibilant consonants, such as the sounds normally represented in English by "s", "z", "ch", "j", "t" and "sh", in recordings of the human voice. Sibilance lies in frequencies anywhere between 2 and 10 kHz, depending on the individual voice.
De-essing (also desibilizing) is any technique intended to reduce or eliminate the excessive prominence of sibilant consonants, such as the sounds normally represented in English by "s", "z", "ch", "j", "t" and "sh", in recordings of the human voice. Sibilance lies in frequencies anywhere between 2 and 10 kHz, depending on the individual voice.
==History== De-essing originated in 1939 at Warner Bros. for film soundtracks. In the 1960s, Ortofon used treble limiting in vinyl cutting. The 1970s saw the first dedicated units like the Orban 516EC. The dbx 902, released in 1980, became a studio standard and shaped modern designs.
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).