How to Configure a De-Esser in Audacity

Excessive sibilance—harsh, piercing "s," "sh," and "t" sounds—can ruin an otherwise clean vocal recording. This guide provides a straightforward walkthrough on how to set up and configure a de-esser in Audacity to eliminate harsh frequencies. You will learn how to obtain the appropriate plugin, dial in key settings like frequency and threshold, and ensure your audio sounds natural without introducing a lisp.

1. Install a De-Esser Plugin

Audacity does not include a dedicated native de-esser by default, but it supports Nyquist and VST plugins. The most reliable free option is the Nyquist De-Esser plugin:

1. Download the de-esser.ny plugin file from the official Audacity Nyquist plugins archive.
2. Open Audacity and navigate to Tools > Plugin Manager.
3. Click Add, locate the downloaded .ny file, select it, and click Enable.
4. The plugin will now appear under the Effect menu.

2. Identify the Problem Frequencies

Before adjusting the plugin, find where the harsh frequencies sit in your recording:

1. Select a section of your vocal track containing a sharp "s" sound.
2. Go to Analyze > Plot Spectrum.
3. Look for a pronounced peak in the high-frequency range. Sibilance typically occurs between 4 kHz and 9 kHz (lower for deeper male voices, higher for female voices). Note the frequency range where the spike occurs.

3. Configure the De-Esser Settings

Select your audio track (or highlight the vocal section) and open Effect > De-Esser. Configure the main parameters as follows:

• Frequency Band (Pass Band): Set this to match your spectral analysis. A standard starting point is 5000 Hz to 8000 Hz. This tells the plugin to target only high-frequency spikes while leaving lower vocal fundamentals untouched.
• Threshold (dB): This determines how loud the sibilance must be before the de-esser activates. Set your playback loop on an "s" sound. Lower the threshold gradually from 0 dB down until the gain reduction meter triggers only during the harsh sibilant moments.
• Ratio or Amount (dB Reduction): Control how much the sibilance is attenuated. Aim for a reduction between -3 dB and -6 dB. Applying more than -6 dB can make the speaker sound as though they have a lisp.
• Filter Type: If given the option, use a Band-Pass or High-Pass filter depending on whether the harshness is concentrated in a tight range or extends to the top of the frequency spectrum.

4. Preview and Fine-Tune

1. Click the Preview button inside the plugin window.
2. Listen carefully to words containing "s," "z," and "c."
3. If the voice sounds muffled or muffled "th" sounds replace the "s," raise the threshold or decrease the attenuation amount.
4. If the harshness persists, widen the frequency range slightly (e.g., adjust the lower bound from 5000 Hz down to 4000 Hz).
5. Click Apply once the balance between clarity and smoothness is achieved.