Wojtek Wiesławski
Tutorials

Struggling to Sit Vocals in the Mix? Do This Before the Compressor

Manual syllable levelling and a separate breath track — a clear, steady vocal without choking it.

Chapters

The vocal is heavily compressed and choking, yet in places it still disappears and becomes unintelligible. This problem keeps coming up in one-to-one consultations and on the Mix SOS livestreams. In this episode Wojtek Wieslawski shows a simple trick done before the compressor. The examples use a vocal from Dilluvium’s album “Rising from the Dark”.

Why a vocal chokes and disappears at the same time

Because singers don’t perform with even dynamics: a whisper one moment, a scream the next. Ideally this is handled while tracking by riding the input level, but for a mix you usually get what you get. On a loud syllable the compressor reduces 10⁠–⁠12 dB and chokes it; on a quiet one it barely works. If the release is too long, a quiet syllable right after a loud one hits a compressor that hasn’t let go yet, and vanishes. A second compressor after the first only brings out artifacts.

Step 1: manual levelling before any plugins

Turn loud syllables down and quiet ones up by cutting the track into pieces and changing the event volume (clip gain). This has to happen before every plugin. In Cubase the fader works post-insert by default, so the compressor would still receive an uneven signal.

  • Enable cutting on zero crossings (in Cubase: snap to zero crossing) or you’ll get clicks. Without that option, use short crossfades.
  • Only fix what sticks out, not every syllable.
  • Judge by ear, not by eye: the same peak on the waveform isn’t the same loudness, and high notes and bright vowels sound louder.
  • Catch harsh sibilants and strong plosives (p, b) along the way ⁠—⁠ compression with make-up gain only exaggerates them.

The result: you learn the track’s weaknesses, and the compressor gets a predictable signal and works in its sweet spot.

Step 2: breaths on a separate track

The compressor turns down loud syllables, make-up gain raises everything, so the gap between syllables and breaths shrinks. With a fast release a breath can end up as loud as the first syllable of a line. Wojtek cuts out every breath and moves it to a separate track ⁠—⁠ vertically only, so the timing doesn’t shift. The breath track gets the same sends, reverb, delay and EQ as the vocal, or it would sound like it came from another room. Then a single fader sets breath level for the whole song.

Why manually, not with Vocal Rider?

An automatic tool watches the level, but listening syllable by syllable you hear the artifacts: a breath that jumps out, a word ending sung too quietly, an “s” that’s too sharp. Good production is in the details.

FAQ

Why level a vocal with clip gain instead of the fader?
Because the fader usually works after the plugins. Clip gain changes the signal before it reaches the compressor.

Should I remove breaths from a vocal?
Better to control them: move them to a separate track with the same effects and set their level with one fader.

Tools in this episode

  • Cubase (clip gain, snap to zero crossing)

Work on your music

Explore mixing services

View the course (in Polish)

Music courtesy of: Dilluvium ⁠–⁠ Rising from the Dark (2025).