Mixing Screaming Vocals

Starting points we use on screamed vocals in metal, metalcore and hardcore. Every voice needs its own adjustments.

Screamed vocals turn up mostly in metal and its subgenres, but also in punk, hardcore and some rock and rap. In principle they’re mixed like any vocal. In practice they’re louder, harsher, noisier and less tonal than sung parts, and they have to cut through walls of distorted guitar. Here is how we approach them, step by step.

Recording quality decides most of the result

A scream covers a wide frequency range, jumps between pitches and noise, and hits much harder than a sung line. The mix can shape that, but it can’t repair a take that clipped at the preamp or one full of room sound. Two things matter most.

• The microphone. A dynamic mic close to the mouth is the safe choice. It handles the level, picks up less room and is less likely to exaggerate harsh upper mids. Our pick is the Shure SM7B: it has a smooth response and switches for bass roll-off and presence boost, but it needs a lot of clean preamp gain. A handheld dynamic like the SM58 works too, and some vocalists scream better holding a mic. Condensers can work in a well-treated room with a careful gain setting, and plenty of records use them. In an untreated home room, a dynamic will usually give you the cleaner take. Our guide to choosing a microphone for recording vocals at home goes further.
• Gain staging. Set the input gain on the loudest scream of the song, not on a quiet verse. Record at 24-bit and leave plenty of headroom. A clipped scream is the one problem that mixing can’t fix.

How to mix screaming vocals

Screams usually need more compression and more saturation than sung vocals, and more control in the upper mids. The goal is a vocal that sounds as aggressive as the performance but stays readable and doesn’t hurt at high volume.

1. Prepare the takes

Remove noise between phrases first. Screams are compressed hard, and compression brings every breath, gasp and bit of headphone bleed up with them. Move the breaths you want to keep to their own track so you can set their level separately, or use a de-breath plugin if cutting by hand takes too long. Put screams and sung parts on separate tracks from the start, because they will need different chains.

2. Equalization

High-pass first, and firmly. A scream has almost nothing useful in the lowest octaves, and what’s there fights the bass and kick. Then listen for the problem areas: often a boxy area in the low mids and harshness somewhere in the 2–4 kHz range. A gentle top-end lift is common on a dull dynamic mic, but go carefully. A big high-frequency boost on a scream adds fizz, not power.

3. Compression in stages

Screams take a lot of compression in total, but spread it out. A typical setup:

• Fast compressor for peaks: an 1176-style FET or a Distressor, fast attack, fast release, catching the loudest hits. All-buttons mode on an 1176 adds grit when you want it as part of the sound.
• Second compressor for level: a slower or optical compressor working gently, so the vocal holds a steady place against the guitars.
• Automation: ride the phrases that still poke out or drop away. It sounds more natural than adding more gain reduction.

Each compressor colours the sound in its own way, so try a few and keep the one that suits the voice. If you can clearly hear the vocal pumping or the noise floor breathing between words, one stage is working too hard.

4. Dynamic EQ for harshness

The harsh range of a scream is also where its consonants live, so a static cut makes it duller and harder to understand. A dynamic EQ band or a resonance suppressor around 2–4 kHz only reduces that range when it spikes. A de-esser helps too, depending on the singer and the language.

5. Saturation for aggression

Saturation is where screams get their character. Tape, tube or distortion adds harmonics that make the vocal sound bigger and help it cut through guitars, without the brittleness of an EQ boost. A common approach is parallel distortion: send the vocal to a bus, distort it heavily, filter the top and bottom, and blend it under the clean scream. Read more in our guide to saturators. The aim is a vocal that sounds overdriven but still pleasant, not uncontrolled distortion.

6. Reverb and delay

Dense, loud rock and metal mixes leave little room for long reverb. A long tail on a scream disappears into the guitars and only blurs the attack. Use a short room or plate, low in level, and a tempo-synced delay that’s filtered and quiet. Save longer reverb and delay throws for breakdowns, gaps and the ends of phrases, and automate them in.

7. Doubles and the place in the mix

Double-tracked screams panned left and right are common in metal and metalcore, with the lead scream in the centre. Keep the doubles tight in timing and a little lower. Then check the vocal against the whole band. In heavy music the vocal usually sits further into the guitars than it would in pop or rap, where it rides on top. Every word doesn’t have to be clear, but the vocal should never vanish.

Mixing a rock or metal song with screams?

Screams are one of the hardest vocals to get right, and they are judged next to commercially released heavy records. For the rest of the mix, see how to mix rock music and mastering rock music. You can hear a rock before/after in our homepage samples. If you’d rather hand the session over, we mix rock and metal regularly and keep screams aggressive without the harshness.

If the vocal needs more than a plugin, our manual vocal editing (Melodyne, by hand) starts at $57 per song, mixing by an engineer from $113, and a full rock mix with mastering from $133.