How to Mix Vocals on a Trap Beat: Chain, EQ, Autotune

Most home vocals don't sound bad because of the mic. They sound bad because the mix is done in the wrong order. The voice gets EQ'd before it's cleaned up, compressed while it's still peaking, and tuned after the compressor has already squashed the pitch information. Then the beat gets turned down until the vocal "sits", and the whole record ends up quiet and flat.

This guide gives you a vocal chain in the order that works, with ranges you can start from and one full verse walked through at 146 BPM. Open it mid-session and go down the list. Every number here is a starting point. Your voice, your mic and your room will move them.

Set your levels before any plugin

Gain staging is the boring step that fixes half your problems. Do it first.

• Recording level: peaks between -12 and -6 dBFS. Never touch 0. In 24-bit, quiet is safe and clipped is ruined.
• Beat level: drop the beat so its loudest hits peak around -8 to -6 dBFS on the master. You need that headroom for the vocal and for whatever limiter you use later.
• Vocal level: pull the fader until the lead sits around -6 to -3 dBFS peak before processing. Then judge it against the beat, not in solo.

Never mix the vocal in solo for more than a minute at a time. Solo makes you carve out things the beat will cover anyway, and you end up with a thin voice.

If you're working from a purchased beat, a WAV file gives you cleaner material to work with than a compressed MP3. See licensing for what each tier includes.

The chain order that works

Run this order on the lead vocal track:

1. High-pass filter at 80-100 Hz. Everything under that is room rumble and breath thumps.
2. Subtractive EQ. Cut mud and boxiness.
3. De-esser (before compression, so the compressor doesn't push the S sounds up).
4. Pitch correction.
5. Compressor one: the leveler.
6. Compressor two: the fast catcher (optional).
7. Saturation: a little warmth and edge.
8. Tone EQ: the additive boost.
9. Sends: reverb and delay, not inserts.

Why cut before boosting? Because boosting a muddy vocal makes the mud louder. Clean first, then brighten.

Subtractive EQ starting points

• Cut 200-400 Hz by 2-4 dB with a medium Q if the voice sounds boxy or cardboard-like. Sweep to find the worst spot.
• Cut 500-800 Hz by 1-3 dB if it sounds nasal or "honky".
• Check 2-4 kHz. If it's harsh on your ears, a 1-2 dB dip there often helps more than a de-esser.

Autotune and pitch correction

On trap and rage records the tuning is part of the sound, so decide how audible you want it.

• Key and scale: set it to the beat's key. If you don't know the key, find it first. A wrong scale gives you notes that wobble between pitches in a bad way.
• Retune speed: 0-10 ms is the hard, robotic snap. 15-30 ms is tuned but still human. 40 ms and up is subtle correction.
• Humanize: turn it up a little on long held notes so they don't sound frozen.

Tune before the compressors. A compressor changes the volume contour of a note, and the tuner reads that contour. Feed it a steady, uncompressed signal and it behaves better.

Record a few lines with no tuning and a few with the tuner running in your headphones. Many artists find their pocket faster when they hear the effect live.

Compression: two light passes beat one hard one

One compressor working 10 dB gives you pumping and breath noise. Two compressors working 3-4 dB each sound smoother.

Compressor one (the leveler):

• Ratio 3:1 to 4:1
• Attack 10-30 ms, so the start of each word still punches through
• Release 50-100 ms, or auto
• Threshold set for 3-5 dB of gain reduction on the loud phrases

Compressor two (the catcher):

• Ratio 4:1 to 8:1
• Fast attack, 1-5 ms
• Fast release, 20-50 ms
• Only 1-3 dB of gain reduction, just to catch the peaks that slipped through

Watch the gain reduction meter, not the knobs. If the needle is sitting at 8 dB all the time, back off the threshold.

Ad-libs, doubles and stacks

Ad-libs are texture, and they should sound different from the lead.

• Level: 4-8 dB below the lead.
• Pan: hard or wide, around 30-70% left and right. Keep the lead dead center.
• EQ: high-pass them higher than the lead, 200-300 Hz, and let them live in the upper mids.
• Effects: more delay and reverb than the lead, or a filter that thins them out.

For doubles on hook lines, record two takes of the same line and pan them 20-40% left and right. Don't copy-paste the lead and nudge it. You'll get phasing, not width.

Worked example: one 16-bar verse at 146 BPM

You have a 146 BPM beat and a 16-bar verse. At 146 BPM a bar lasts roughly 1.64 seconds, so the full verse runs about 26 seconds.

Step 1: delay times. One beat is 60,000 ÷ 146 ≈ 411 ms. So:

• 1/8 note ≈ 205 ms
• Dotted 1/8 ≈ 308 ms
• 1/4 note ≈ 411 ms

Step 2: set up the lead. High-pass at 90 Hz. Dip 2.5 dB at 300 Hz, and 1.5 dB at 2.8 kHz because the voice is a bit harsh. De-ess around 6-8 kHz for 2-3 dB of reduction on the S sounds.

Step 3: tune. Retune speed 15 ms, the beat's key set, humanize low.

Step 4: compress. Compressor one: 4:1, 20 ms attack, 80 ms release, about 4 dB of reduction. Compressor two: 6:1, 2 ms attack, 30 ms release, about 2 dB.

Step 5: tone. Add 2 dB with a wide shelf around 10 kHz for air. Add 1.5 dB around 4-5 kHz if the lead still gets buried by the hi-hats.

Step 6: delay send. Use a 1/4 note (411 ms) delay with the feedback around 20-30%, high-passed at 300 Hz and low-passed around 5 kHz so the repeats stay out of the way. Automate the send up only on the last word of every second bar.

Step 7: reverb send. Short plate, 0.8-1.2 s decay, 20-40 ms pre-delay. Keep it low. If you can clearly hear the reverb, it's too loud.

Step 8: check. Loop bars 9-12 with the beat at full volume. Can you hear every word? If not, raise the vocal by 1 dB before touching any EQ.

Common mistakes and fixes

• Too much reverb. The voice gets pushed back and the words blur. Fix: cut the send until you only miss it when you mute it.
• Boosting 10 kHz to sound "expensive". It only makes S sounds and noise louder. Fix: de-ess first, then add the shelf gently, 1-3 dB.
• Compressing with a slow attack on fast flows. The start of every word gets through unchecked and the line sounds uneven. Fix: set the first compressor to 10-20 ms and the second to 1-5 ms.
• Mixing only in headphones. The low end and the width will mislead you. Fix: check on speakers, a car and a phone. If the vocal vanishes on the phone, add a touch around 1-3 kHz.
• A vocal that fights the 808. Fix: keep the high-pass on the vocal, and don't boost anything below 150 Hz on the voice.
• Loud ad-libs. They take attention from the lead. Fix: pull them down until they only show up when you mute them.

Quick checklist

• Record peaks between -12 and -6 dBFS, no clipping.
• Beat peaks around -8 to -6 dBFS on the master before the vocal goes in.
• High-pass the lead at 80-100 Hz, then cut mud at 200-400 Hz.
• De-ess first, then tune, then compress.
• Use two light compressors, 3-5 dB of gain reduction each at most.
• Keep the lead centered, spread ad-libs and doubles wide, and sit ad-libs 4-8 dB lower.
• Use reverb and delay on sends, with delay times based on the BPM.
• Check the mix on speakers, a phone and a car before you export.

Take a beat you already have, strip your session back to a clean chain in this order, and mix one verse before touching the hook. If you need something to practice on, browse the beats and pick one that suits your voice.

---

Fresh from PROD. SIGIL: COLD SOLVENT · AMBER COIL — all beats