How to Mix Vocals at Home — The Full Process

Step 1 — Clean Up Before You Process

The first mistake most beginners make is jumping straight to EQ and compression without cleaning up the recording first. Apply a high-pass filter at 80-100Hz to remove low-frequency rumble and handling noise. Then run noise reduction to strip out any background hiss, HVAC, or room tone. These steps are invisible when done right — they just remove problems without affecting the voice.

Step 2 — EQ the Vocal

Vocal EQ is split into two stages: corrective and additive.

Corrective cuts: around 200-400Hz for "boxy" buildup (cut 3-6dB with a medium Q), around 1-2kHz for nasal midrange harshness, and 5-8kHz for sibilance harshness that the de-esser will handle more precisely later.

Additive boosts: a gentle shelf or bell boost around 3-5kHz adds "presence" — that forward, cutting quality that makes a vocal audible in a dense mix. A high-shelf boost above 10kHz adds "air" and openness. Both boosts should be subtle (1-3dB) — you're enhancing, not creating.

Step 3 — Compress

Compression controls the dynamic range — the difference between the vocalist's loudest and quietest moments. A typical starting point: 4:1 ratio, fast attack (5-10ms), medium release (60-100ms), threshold set so 6-8dB of gain reduction on the loudest phrases. The goal is consistency — the vocal should sit at roughly the same level throughout the song without you having to ride the fader constantly.

Many engineers use two compressors in series: a fast one to catch peaks (1176-style) and a slower one for overall dynamic shaping (LA-2A-style). If you only have one compressor, go with the 4:1 ratio above.

Step 4 — De-Ess

A de-esser is a frequency-specific compressor that targets harsh sibilance — the "s," "t," and "sh" sounds that get exaggerated by microphone proximity and previous compression. Set the frequency between 5-8kHz (sweep it while the vocal plays to find where the harshness peaks). Keep the threshold high enough that it only triggers on problem moments, not all the time.

Step 5 — Reverb and Delay

Space effects go last, after all the corrective processing. Use a short room or plate reverb (pre-delay of 20-30ms, decay time of 0.8-1.5s) to add dimension. High-pass the reverb return above 200Hz so it doesn't cloud the low end. A slap delay (60-120ms, one repeat, -10dB below dry) adds depth without sounding like a delay effect. Both should support the vocal, not compete with it — if you can clearly hear the reverb, it's probably too much.

The Shortcut — Let AI Handle It

If you want to skip the manual chain or use it as a reference point, Engineer Guy applies the full vocal processing automatically. Upload your vocal, and the AI runs noise reduction, EQ, compression, de-essing, and level correction — then tells you exactly what it did. You can use it as your final result or as a starting point to learn from.