A good vocal mix is mostly about order and restraint. Beginner mixes rarely sound wrong because something is missing from the chain. They sound wrong because every step does too much. Here is a simple chain, in the order you would use it, with the reason for each step.
Before the plugins: a clean take and a rough balance
No chain rescues a bad recording, so start by picking the best takes and joining them into one clean performance. Fix timing only where it is obvious. Lower loud breaths instead of deleting all of them, because a vocal without breaths sounds artificial.
Then set a rough balance. Pull the vocal fader to a level where you can clearly understand the words over the beat, and leave it there while you work. Make your decisions with the whole track playing. A vocal that sounds great alone is often wrong in the mix.
Step 1: clean-up
Start with a high-pass filter to remove rumble and low noise below the voice. For most voices somewhere around 80 to 100 Hz is a safe place to begin, a little higher for a thin voice. Raise it until the voice starts to lose body, then back off a notch.
Next, deal with the loud and quiet spots by hand. Use clip gain, the volume of each clip, to lift a quiet word or lower a harsh syllable or a popping "p". This does more for a natural sound than any compressor, and it lets the compressor work gently later.
Step 2: EQ, first to remove, then to add
Remove before you add. Sweep a narrow boost across the range to find what is bothering you: mud somewhere in the low mids, a boxy tone, a harsh note. Then cut it a little, two or three decibels is often enough.
After that, add gently. A wide, soft lift in the presence area, roughly the upper mids, makes the words clearer. A little extra air at the very top adds sparkle if the recording is clean. Avoid big boosts. If the voice does not fit, try cutting the same area in the beat, so the voice has its own space.
Steps 3 and 4: compression and de-essing
Compression
The goal is a steady voice, not a squashed one. A moderate ratio, such as 2:1 to 4:1, is a good start. Set the threshold so the loudest words are reduced by around three to six decibels. A medium attack lets the start of each word through, which keeps it alive, and the release should let the compressor recover between words, following the rhythm of the lyrics.
Two gentle compressors in a row often sound more natural than one hard one. Use a faster setting for rap and a slower one for a ballad. And remember that the fader and clip gain control loudness. Compression controls consistency.
De-essing
Sharp "s", "t" and "ş" sounds sit mostly in the high frequencies, somewhere around five to eight kilohertz. Compression tends to bring them forward, which is why the de-esser usually comes after it. Reduce only as much as you need. Too much de-essing gives the voice a lisp, and it takes away the clarity you just built.
Check it on small speakers or a phone. Sibilance is the thing that gets obvious on a weak speaker.
Steps 5 and 6: space and fit
Reverb and delay on sends
Put reverb and delay on separate send tracks, not directly on the vocal, so the dry voice stays clear and you can shape the effect on its own.
- Reverb: a short room or plate gives natural space. Choose a longer one for a slow song. Cut the low end and the very top of the reverb return, so it doesn't make the mix muddy. A little pre-delay, a few tens of milliseconds, keeps the voice in front of the reverb.
- Delay: set it to the tempo of the song, quarter or eighth notes, keep it quiet and filter its top. It often works best only on the ends of phrases, which you can do with automation.
A good test: if you can clearly hear the effect all the time, turn it down. You should notice it when it stops.
Fit the vocal to the beat
Check the level again with everything in place. The voice should sit on top without looking separate from the music. Lower the volume and see if you can still follow the words.
Listen in mono, on a phone speaker, on headphones and, if you can, in a car. Compare with a finished song in a similar style, and notice how loud its vocal sits and how bright it sounds. Backing vocals and doubles go lower, a bit darker and with more reverb than the lead.
Take breaks. A mix you finish tired often sounds different the next morning.
Common mistakes
- Listening to the vocal solo for too long.
- Boosting hard around the same frequency that the de-esser then removes.
- Adding a lot of reverb to hide a weak take. Re-record it instead.
- Loading a preset chain without listening to what each step does.
- Pushing the vocal bus loud to compete with the beat, instead of fixing the beat.
The best chain is the shortest one that makes the voice sound good. If a step does not clearly improve the sound, switch it off and move on.