Vocal mics tend to have a bright frequency response and, when matched to a singer with high sibilant content, things may be problematic from the start. Methods to try to tame wild ess sounds include using foam windsocks, turning the mic slightly off-axis to the singer or in extreme cases, using a different mic. While equalization could be applied at this point, it is advisable to leave that until later in the process.
Minor problems can often be addressed across the board with notch filtering. Set a parametric EQ band for narrow width (a high Q setting). Boost this band as high as it can go and sweep the frequencies between 2 kilohertz (kHz) and 10 kHz. This will help identify the most objectionable sibilant frequency. Reverse the boost to cut to de-ess the entire performance as long as the cut does not affect clarity or vocal sounds.
Multiband compressors split the frequency spectrum into two or more sections. For more aggressive de-essing, set a compression band for the affected frequencies and set its threshold to compress when ess sounds occur, but not through the remainder of the vocal. Adjusting attack and release times of the compressor can often help to blend the de-ess compression naturally, so the effect is not noticeable. Run de-ess multiband compression before any overall compression on the vocal performance.
Make a copy of the vocal track and EQ it to accent the sibilant frequencies. Do not send this track to your mix, but rather use it to feed the sidechain of a compressor applied to the original vocal track. The sidechain controls the actions of the compressor based on the ess-emphasized track, but the effect of the compressor is on the original track, bringing down the level of sibilant frequencies as they occur. Again, careful adjustment of compressor controls will keep the sound of processing natural.
Digital audio workstations allow engineers to see tracks on a waveform basis, in great detail. To de-ess through automation, identify the spots in the wave form where sibilants occur. Create a volume envelope to attenuate these frequencies. This has the effect of an engineer pulling back a fader as each ess sound occurs, but with a speed and accuracy beyond what could be accomplished in real time.
De-essing plug-ins and presets are common. However, de-essing is not a one-size-fits-all solution. Be prepared to adjust parameters, like frequency, gain reduction and threshold, to match your situation.