Mixing
How to mix vocals over a beat in Logic Pro
You recorded over a two-track beat and the vocal sounds like it was taped on top. Here is the order engineers work in, with the Logic plug-ins and settings that get you most of the way.
Mixing a vocal over a finished beat is different from mixing a full band. The beat is already mixed and often already loud, so you are not building a mix from scratch. Your job is to make room for one voice inside something that is already finished. That changes the priorities: level and space matter more than clever processing.
This guide uses Logic’s built-in plug-ins, so you don’t need to buy anything. The same order works in any DAW.
Before you touch a plug-in
1. Turn the beat down
Most beats you download are mastered, peaking near 0 dBFS. If your vocal has to compete with that, you will end up pushing everything into the red. Drop the beat’s fader by 6 to 8 dB before you start. You’re giving yourself headroom, not changing the song: the master at the end brings the whole thing back up.
2. Comp and clean the vocal
Pick your best takes and build one lead vocal from them. Cut out breaths that are too loud, mouth clicks, and the noise between phrases. A clean, well-performed take will always beat a heavily processed weak one. If you have doubles and adlibs, put each on its own track; you’ll treat them differently from the lead.
3. Set a rough balance
With no plug-ins at all, move the vocal fader until you can hear every word without the vocal feeling separate from the beat. Listen quietly. If it works at low volume, it will work loud. This rough balance is the most important decision in the mix, and everything after it is refinement.
The vocal chain, in order
4. EQ: clean up first
Insert Channel EQ on the vocal.
- High-pass around 80 to 100 Hz for most voices (lower for deep voices). This removes rumble and the low end that clashes with the 808 and kick.
- Cut mud between 200 and 400 Hz if the vocal sounds boxy or thick. A gentle 2 to 4 dB cut with a medium width is usually enough.
- Add presence around 3 to 5 kHz if the words aren’t cutting through. Go easy: 1 to 3 dB.
- Add air with a high shelf around 10 to 12 kHz for a more polished, modern sound.
Our vocal EQ cheat sheet goes through each range in more detail.
5. Compression: control the level
Insert Compressor after the EQ. A vocal that jumps between whispering and shouting is impossible to place in a beat; compression evens it out so it stays at the level you set.
- Start with a ratio of 3:1 to 4:1, a fast-ish attack (5 to 15 ms) and a release around 50 to 100 ms.
- Lower the threshold until the gain reduction meter moves 3 to 6 dB on the loud parts.
- Use the make-up gain so the vocal is the same loudness bypassed and active. Otherwise you’ll just hear “louder” and think it’s better.
For rap and trap vocals, many engineers use two compressors in a row, each doing a little, rather than one compressor working hard. It sounds more natural.
6. De-essing: tame the S sounds
Compression and presence boosts make S, T and SH sounds sharper. Insert DeEsser 2 after the compressor and set it so it only reacts on the harsh consonants, usually somewhere between 5 and 9 kHz. If you can hear the vocal getting lispy, back it off.
7. Space: reverb and delay on sends
Don’t put reverb directly on the vocal track. Create a bus (an aux) with the reverb on it and send the vocal to it. That way you control how much space the vocal gets without dulling the dry signal.
- Reverb: ChromaVerb with a short room or plate, 0.8 to 1.5 seconds. High-pass the reverb return around 300 to 500 Hz so it doesn’t muddy the low end.
- Delay: Stereo Delay synced to the song tempo, a quarter or eighth note, mixed low. Delay adds depth without washing the vocal out, which is why it suits busy beats better than lots of reverb.
The rule of thumb: you should miss the effect when you mute it, but not notice it when it’s on.
Making it sit
8. Carve a little room in the beat
If the vocal still sounds covered, the beat is probably busy in the same range as your voice, usually 1 to 4 kHz. Put a Channel EQ on the beat and dip that range by 1 to 3 dB. Small moves on the beat often do more than big moves on the vocal.
9. Ride the level
Even after compression, some words get lost when the beat gets busier, like in the hook or when a new element comes in. Use volume automation to push those words up a dB or two. Engineers spend a lot of their time on this; it’s the difference between a vocal that’s loud and a vocal that’s always clear.
10. Treat the backing vocals differently
Doubles and adlibs should support the lead, not fight it. Keep them lower, pan doubles left and right, cut more low end from them, and give them more reverb and delay than the lead so they sit further back.
Check it like an engineer would
- Listen on more than one system: earbuds, a phone speaker and a car if you can. A vocal that disappears on a phone usually needs more presence; one that is harsh on earbuds needs more de-essing.
- Compare with a record you like: drop a reference song into the session, turn it down to match your mix’s loudness, and switch between them.
- Take breaks: your ears adjust after 20 or 30 minutes. Come back and the problems will be obvious.
When the mix is done, the last step is mastering, getting the whole song to the right loudness for streaming. See how loud your master should be for Spotify.