One of the biggest problems facing producers who process vocal takes is that of sibilance. Sibilance refers to the harsh consonants, notably ess and eff, that are exaggerated due to either the singer’s voice or delivery (sometimes getting too close to the microphone can exaggerate sibilance). The microphone is merciless in picking up sibilance and taming these nasties can be a nightmare if you don’t know how.
The process of taming or controlling sibilance is called De-Essing which makes perfect sense as what we are trying to do is to de-ess, ie, remove the sss or attenuate it. Quite often the producer will manually search for sibilance and use volume automation to attenuate the sibilant frequencies. However, this can be time consuming and quite laborious and in these instances we prefer to use a dedicated de-esser. The de-esser is designed specifically to treat sibilance in vocal recordings.
The problem with treating sibilance is trying to locate exactly where it resides in the vocal waveform. Sibilance invariably can cover quite a wide band of frequencies and I have often treated them from 4 kHz all the way to 12 kHz. Additionally, you need to be careful not to attenuate too heavily as the process can suck the life out of the band of frequencies being processed: remember that we are attenuating a range of frequencies and not individual frequencies.
Even the best tracked vocals can exhibit sibilance.
In the video I explain what a de-esser is, how it works and how to use it on female vocals. I use the wonderful Toneboosters Sibilance plugin and show you how to customise the settings for achieving the optimum results. I show you how to locate and isolate sibilant frequencies in a vocal recording and explain why certain frequencies are more problematic than others.
Plugin used in this video:
Topics covered in this video are:
- What is Sibilance
- Best practices
- Sibilance location
- Sibilance and bandwidth
- Filtering for Texture
- Toneboosters Sibilance and how to use it
- Threshold manipulation
- Selecting the correct Algorithm
- Range and Knee in Sibilance