De-esser
Definition
Audio processor that reduces sibilance (harsh “s” and “sh” sounds) in vocal recordings by applying frequency-specific dynamic compression, typically in the 4-10kHz range.
Context
De-essers are a standard part of vocal processing chains in DAW mixing workflows. They can operate in wideband mode (compressing the entire signal when sibilance is detected) or split-band mode (only compressing the targeted frequency range). Common de-esser plugins include FabFilter Pro-DS, Waves DeEsser, and stock de-essers included in most DAWs.
Related Terms
Mixing Techniques (from mixing-talk)
In mixing-talk’s vocal mixing discussions (2,741 messages), de-essing is one of the most discussed vocal processing steps:
- Placement in the chain: The community consensus is to place the de-esser after compression, since compression can exaggerate sibilance. Some engineers use a gentle de-esser both before and after compression
- Multiple light instances vs one heavy instance: Using 2-3 de-essers doing light work produces more transparent results than one de-esser working hard
- Split-band vs wideband: Split-band mode (only compressing the sibilant frequency range) is generally preferred for transparency
- Manual de-essing: The most natural approach — using clip gain to manually reduce individual sibilant peaks — but the most time-consuming
- BatMeckley’s FabFilter approach (pinned): Using FabFilter Pro-MB’s dynamic bands as a precision de-esser for particularly problematic vocal resonances that standard de-essers can’t handle