Mixing Vocals: How to Set a De-Esser | Soundways

Why Use a De-esser?

De-essers are most often used to control high frequency peaks found in a vocal performance.  These peaks, known as sibilance, occur when singers use consonant sounds like "s" and "t."  Siblance sounds unpleasant, harsh, and needs to be controlled.  For this, we typically use a frequency dependant compressor, also known as a de-esser.  

De-essers have some other applications, usually in mixing or mastering.  If a mix is overly bright, a de-esser can be used to control the high frequencies in a more transparent manner than traditional EQ.    

Step 1:  Identify Sibiliants

First, listen to the quality of the singer's voice and the vocal recording.  If you find that there are sometimes harsh high frequencies, use a de-esser.

Identify sibilant peaks.  Listen for the loudest peak and use this section to adjust the settings on your de-esser.  To adjust these settings, insert a de-esser on your vocal track and place Reveal on your master bus.  When listening for transients, use Device mode.  This will allow you to precisely identify any harsh resonances in the 2kHz - 5kHz range.  

Step 2:  Set the center frequency and threshold

Now that you have identified the loudest peak, set the center frequency and threshold of your de-esser.  The center frequency is the range that will be compressed the most once the threshold is reached.  This setting should be adjusted to the frequency area in which the sibilance occurs.  This is typically between 4kHz - 7kHz.  

The threshold setting determines how much level can pass through the de-esser before compression is triggered.  Lower the threshold until gain reduction starts to occur.  Now listen as you adjust the threshold.  Find the point at which sibliance is controlled before the vocal becomes dull.  Note that the de-esser should only react during instances of siblance, not the entire performance.

Step 3: Quality Check with Reveal

Use Reveal to quality check your de-esser settings.  Turn on Device mode and listen to your vocal.  This mode allows engineers to easily identify any harsh peaks that can occur around 3kHz.  If you find that in this context your vocal still sounds harsh, adjust the threshold of the de-esser until you achieve a smooth sound.  Using Reveal engineers can be confident that their mixes will translate accurately to all consumer devices.