For most vocals in Logic Pro, open the stock DeEsser 2 and start the Frequency around 6 kHz for a male voice or around 8 kHz for a female voice. Lower the Threshold until the gain reduction moves only on the "s" and "t" sounds. Set Max Reduction to about 3 to 6 dB, then adjust by ear.
I'm Warren Young, and I mix vocals in Logic every week. This page covers de-esser settings only. For the order of every plugin on the channel, see my full vocal chain for Logic Pro. For the dynamics side, read my vocal compression settings for Logic Pro.
Every number below is a starting point. Voices, mics and songs differ, so use the values to get close and then trust your ears.
What a de-esser does
Sibilance is the sharp hiss in "s", "sh", "ch" and "t" sounds. It lives in the upper midrange and highs, and a compressor or a bright EQ can make it worse. A de-esser listens to one narrow band. When that band gets too loud, it turns the signal down for a moment, and the rest of the vocal stays untouched.
Logic's stock plugin for this job is DeEsser 2. You find it in the Dynamics group of the plugin menu on any audio channel strip.
Frequency: find where the harshness sits
The Frequency control sets the band the de-esser listens to. Sibilance usually sits in a different place on each voice, so find it instead of guessing.
- Male voices: start around 5 to 8 kHz.
- Female voices: start around 6 to 10 kHz.
To find the exact spot, play a section with strong "s" sounds. Sweep the Frequency knob slowly and watch the gain reduction meter. The setting where the meter reacts most to the hiss is your starting point. Another trick is to boost a narrow band on a spare EQ, sweep it until the "s" sounds are at their worst, and note the frequency. Then remove the boost and use that number in DeEsser 2.
The Filter menu changes the shape of the detection filter. Try each filter on the same phrase and keep the one that catches the hiss without reacting to the vowels.
Threshold: set it from the meter
Do not set the Threshold by a number. The right value depends on how loud the recording is and how bright the mic is. Play the harshest phrase and lower the Threshold until the gain reduction meter moves on the sibilants.
The meter should barely move on normal words and vowels. If it moves on every syllable, the Threshold is too low. If you cannot hear any change on the harsh "s" sounds, it is too high.
Max Reduction: cap how much it can cut
Max Reduction limits the amount of gain reduction the de-esser can apply. A low cap keeps the vocal natural. A high cap removes more hiss but can leave a soft, lisping sound.
- 3 to 4 dB: light control for a mic that is already smooth.
- 4 to 6 dB: a good start for most pop, R&B and rap vocals.
- 6 to 8 dB: for bright mics or heavily compressed vocals. Check that the vocal still sounds clear.
Relative or Absolute mode
The Mode menu offers two ways of working. Pick one by ear, because each one reacts differently to loud and quiet takes.
- Relative: responds to both loud and quiet signals. It is a good first choice for a vocal with a wide range from soft verses to loud choruses.
- Absolute: the classic de-esser behaviour. It works best on high level signals, so it suits a vocal that is already compressed and even.
I start in Relative on most vocals and switch to Absolute only if the result feels uneven from phrase to phrase.
Split or Wide
The Split and Wide setting decides what gets turned down. In Split, DeEsser 2 processes only the band you set. In Wide, it processes the full range.
Split is the safer choice for a clean vocal because only the harsh band moves. Try Wide if Split sounds thin or if the hiss spreads over a broad range. Wide changes the whole sound, so use your ears and keep Max Reduction low.
Starting settings by voice
Use this table as a first pass. Set the Threshold by the meter, not by a number.
| Voice | Frequency (start around) | Max Reduction | Mode | Split or Wide |
|---|---|---|---|---|
| Male lead | 5 to 7 kHz | 3 to 5 dB | Relative | Split |
| Male rap vocal | 6 to 8 kHz | 4 to 6 dB | Relative | Split |
| Female lead | 7 to 9 kHz | 3 to 5 dB | Relative | Split |
| Bright female or breathy vocal | 8 to 10 kHz | 4 to 6 dB | Relative | Split |
| Heavily compressed vocal | by sweep | 4 to 6 dB | Absolute | Split |
| Background vocals | 6 to 9 kHz | 4 to 6 dB | Relative | Split |
These values are my estimates from experience, not fixed rules. A dark mic or a soft singer may need less. A bright condenser mic may need more.
Where to put the de-esser in the chain
Most producers put the de-esser after the compressor. Compression brings quiet sibilance forward, so the de-esser catches it afterwards. A common order is corrective EQ, compression, de-esser, then tone EQ and effects. If your tone EQ adds a lot of high end, check the sibilance again after that boost.
Some engineers also place a de-esser before the compressor so the compressor does not react to loud "s" sounds. Try both and keep the one that sounds better on your vocal. For the EQ side, see the vocal EQ cheat sheet. For the dynamics before the de-esser, read what vocal compression is and why it works.
Step by step: de-ess one vocal
- Insert DeEsser 2 after your compressor.
- Loop a phrase with strong "s" sounds.
- Set Mode to Relative and choose Split.
- Sweep Frequency until the meter reacts most to the hiss.
- Lower the Threshold until the meter moves only on sibilants.
- Set Max Reduction to 3 to 6 dB.
- Bypass the plugin and compare at the same volume. The "s" sounds should be softer, and the vocal should not sound duller.
- Play the full song and check again in the loudest chorus.
Common mistakes
- Threshold too low. The de-esser works on every word and the vocal sounds dull or lisped. Use the meter.
- Max Reduction too high. It removes hiss but also removes clarity. Keep it as low as the job allows.
- Wrong frequency. If the hiss stays, you are listening in the wrong place. Sweep again.
- De-essing too early and boosting later. A bright EQ after the de-esser can bring the sibilance back.
- Using it to fix a bad recording. If the mic is far too bright, a different mic position or a gentle EQ cut helps more than heavy de-essing.
- Judging by solo. Check the vocal against the beat, because sibilance often hides in a busy mix.
Skip the setup with a ready-made chain
If you would rather start from a finished vocal chain, my Logic presets give you a full channel strip to build on. Open one on your vocal, then adjust the settings to your recording. Browse the Logic Pro X vocal presets collection to see the styles.
Want to try one first? Download the free free vocal presets for Logic Pro.
FAQ
What frequency should I set on the de-esser in Logic Pro?
Start around 5 to 8 kHz for male voices and around 6 to 10 kHz for female voices. Then sweep the Frequency knob while a harsh phrase plays. Keep the setting where the gain reduction meter reacts most to the "s" sounds.
What threshold should I use on DeEsser 2 for vocals?
There is no fixed number. Lower the Threshold until the gain reduction meter moves on sibilants and stays almost still on other words. Then set Max Reduction to about 3 to 6 dB and compare with the plugin bypassed.
Should I de-ess a vocal before or after compression in Logic Pro?
Most producers de-ess after compression, because the compressor brings quiet sibilance forward. Try both positions on your vocal. Keep the one that sounds smoother, and check again after any high-frequency EQ boost.
For the complete signal path, return to the vocal chain for Logic Pro.
