Exploring the benefits of de-essers for audio mixing

De-essers are digital audio effect plugins (or pro audio signal processing hardware effects) used in many different things, like music production, audio treatment for podcasts, and even creating when audiobooks. They are meant to reduce the harsh "ess" or excess "sibilant" sounds that can sometimes be prominent in a vocal recording. These are the "s," "sh," and "ch" sounds, which can be a bit annoying if they're too loud.

The difference between split-band and wide band de-esser​

There are two main types of de-essers: split-band and wideband de-essing plugins. A wide band de-esser works on the full range of frequencies in the audio signal. When it detects harsh "ess" sounds—above your specified threshold level in dB e.g., -18 dB—from words like "ssssnake" or "hissss", it turns down the entire digital or analog signal to reduce those ear-piercing high frequencies. You may as well think this as dynamic range compression that is tuned to be sensitive to the frequencies produced by fricative consonant sounds.

On the other hand, a split-band de-essing audio processor divides the source material into two or more frequency bands. It only turns down the band with the harsh "ess" sounds, leaving the lower frequencies untouched. This makes it useful in sound engineering or audio mastering applications e.g., taming transients or acting as a high frequency limiter.

What frequency range is sibilance?​

Take the "s" sound, for instance. You'll usually hear it most clearly between 5,000 and 8,000 Hz. The "sh" sound is similar, but it tends to be more prominent between 4,000 and 6,000 Hz. The "ch" sound is a bit different—it can be heard across a wider range, from 2,000 to 8,000 Hz. And then there's the "z" sound, which is similar to "s" but a bit deeper, typically falling between 4,000 and 7,000 Hz.

Note: The exact spot where fricative consonant sounds become unpleasant can shift a bit depending on who's speaking or singing (e.g., male or female vocals), the type of microphone they're using, and how the recording is set up i.e., the proximity effect—the closer you are to the microphone leads to an increase in low frequency response and vice versa.

Click to expand...

Where does a de-esser go in a vocal processing chain?​

You should insert your favorite de-essing plugin early in the chain. Here's a common sequence:

• Subtractive EQ (to cut infrasonic bass notes)
• De-esser
• Auto-Tune (or automatic pitch correction)
• Compressor/Limiter (if needed)
• Additive EQ (for boosting desired frequency bands)
• Other dynamics processors (gate, expander, etc. if needed)
• Saturation/Distortion (for adding grit or character if needed)
• Modulation effects (chorus, flanger, etc. if desired)
• Delay
• Reverberation
• Gain utility plugin (if needed for automation)