How It Works
Sibilance — the sharp, hissing sound produced by consonants like S, T, SH, and CH — is a natural part of speech, but microphones and signal chains often exaggerate it. Condenser microphones with presence boosts, bright preamps, and high-frequency EQ all compound sibilance, making it piercing and fatiguing in a mix. A de-esser targets this specific problem.
A de-esser is essentially a compressor with a sidechain filter tuned to the sibilant frequency range, typically between 4 kHz and 10 kHz. When it detects energy in that range exceeding the threshold, it reduces the gain — either on the full signal (broadband mode) or only on the sibilant frequency range (split-band mode). The result is that harsh S sounds are tamed while the rest of the vocal remains untouched.
Setting up a de-esser involves identifying the specific sibilant frequency for the particular vocalist (it varies from person to person and microphone to microphone), setting the threshold so that only the harsh sibilants trigger reduction while normal vocal brightness is preserved, and choosing between broadband and split-band modes. Split-band mode is generally more transparent because it only affects the offending frequency range rather than ducking the entire signal.
Why It Matters for Your Mix
Sibilance is one of the most immediately noticeable problems in a vocal mix. Harsh S sounds are fatiguing to listen to, cause listeners to reach for the volume knob, and become even more pronounced after compression (since compression brings up the average level, making sibilant peaks relatively louder). On earbuds and headphones — where most music is consumed today — sibilance is especially painful. A well de-essed vocal sounds smooth and present without drawing attention to the processing. The listener should not notice the de-esser working; they should only notice the absence of harshness. Getting de-essing right is essential for professional vocal production in any genre.
Common Mistakes
Over-de-essing until vocals sound lispy
Removing too much sibilance makes the singer sound like they have a lisp — "s" sounds become "th" sounds, and the vocal loses its natural presence and air. Set the threshold so only the harshest peaks are caught, and aim for 3-6 dB of reduction on the worst sibilants.
Setting the frequency too wide
A de-esser frequency range that is too broad will affect more than just sibilance — it will duck any vocal energy in that range, including non-sibilant brightness and presence. Narrow the frequency target to the specific sibilant frequency of your vocalist, typically a focused range around 5-8 kHz.
Placing the de-esser after heavy compression
Compression raises the average level and can make sibilance worse. Placing the de-esser before compression ensures that the sibilant peaks are tamed before the compressor amplifies them. Some engineers use de-essers both before and after compression with gentle settings on each.
What RoastYourMix Measures
RoastYourMix does not detect sibilance or isolate the vocal, so it cannot tell if a de-esser is needed. It measures the high mids (2-8 kHz) and highs (8-20 kHz) of your whole mix in dB against the typical range of thousands of real uploads, which shows if the top end is hot overall.
See De-essing in Action
Upload your mix and see its loudness, true peak, tonal balance and stereo width measured.
The quick answers
Where should I place a de-esser in my signal chain?
The most common placement is after EQ but before compression. This way, any EQ boosts in the presence range have already been applied (and potentially worsened sibilance), and the de-esser cleans things up before the compressor amplifies the remaining sibilance further. Some engineers also place a second gentle de-esser after compression.
Can I de-ess instruments other than vocals?
Yes. Cymbals, acoustic guitars, and even some synth sounds can have harsh high-frequency content that benefits from de-essing. An overdriven guitar amp can produce harshness in a similar frequency range. Use a de-esser whenever you need to dynamically tame a specific frequency range.
What is the difference between broadband and split-band de-essing?
Broadband de-essing reduces the volume of the entire signal when sibilance is detected, which can cause momentary dips in overall level. Split-band de-essing only reduces the volume within the targeted frequency range, leaving the rest of the signal untouched. Split-band mode is usually more transparent.