Double tracking vocals: real doubles, ADT and fake doubles

Double tracking — singing the same part twice and playing both takes together — is one of the oldest vocal production techniques and still one of the most effective. It makes a vocal thicker, steadier and more confident without adding a plugin you can hear. Almost every chorus you love uses it. It is also very easy to get wrong: too loose and it sounds like two people, too tight and it sounds like a phasey copy.

Why doubling works

Two takes of the same line are never identical. The singer lands each word a few milliseconds earlier or later, pitches each note a few cents differently, shapes vowels slightly differently. Played together, the ear doesn’t hear two singers — it hears one bigger one:

• Thickness. Tiny pitch and timing differences create a natural ensemble effect, the same thing that makes a choir sound bigger than one voice turned up.
• Steadiness. Small pitch wobbles in one take are averaged with the other, so the result sounds more in tune than either take on its own.
• Width. Two doubles panned apart fill the sides and leave the center to the lead.
• Confidence. A doubled line sounds deliberate. That’s why hooks and choruses get doubled and intimate verses often don’t.

The catch is the balance. The differences have to be small enough to fuse into one voice — in practice, syllable starts within roughly 10 ms — and big enough not to collapse into a copy. Our guide on how to align vocals covers those thresholds in detail.

A short history: ADT

Singers have doubled their own parts since tape machines made overdubbing practical. It’s tiring, though, and in 1966 at Abbey Road the Beatles were tired of it. Engineer Ken Townsend came up with artificial double tracking: the vocal was sent to a second tape machine whose playback was slightly delayed and gently varied in speed, then mixed back in. One take now sounded like two. The technique was used all over the albums that followed, and the name stuck — today “ADT” usually means any effect that fakes a double from a single take.

Later, digital pitch shifters added another trick: a copy of the vocal shifted a few cents up or down and slightly delayed. That “micro-pitch” doubling became a studio staple and is behind many doubler plugins.

Real, fake and in between

The essential difference: a fake double is always the same performance, so it can never go out of time, but it also never adds a second human choice. Real doubles drift — and that drift, tightened just enough, is what sounds like a record.

Recording a real double

The best editing trick is a double that needs little editing. Most of that is decided at the mic.

Set-up

• Same chain as the lead for unison doubles: same mic, same distance, same preamp settings. The doubles should feel like the same voice.
• Vary it for big stacks: a different mic, a step back from it, or a different room on some takes adds natural variety to a wall of backing vocals.
• Record after the lead is comped, so the singer follows the final phrasing — not an early take that later changes.

Headphone mix

• Put the finished lead loud and clear in the headphones, with less of the track than usual.
• A click or a tight drum loop helps the singer lock the rhythm.
• Some singers double better with the lead only in one ear. Try it.

What to ask the singer

• Copy, don’t improve. A double is not a second chance at the lead. Same phrasing, same note lengths, same vowels.
• Softer consonants. Lighter “s”, “t” and “k”. Two hard consonants land as one harsh one.
• Less vibrato. Two vibratos rarely move together; straighter notes on the double blend better.
• No ad-libs. Anything improvised belongs on its own track.
• A few passes, then keep the one that follows the lead best — not the one that sounds best on its own.

Editing the double

1. Align it to the lead. Word starts and hard consonants tight, long vowels left alone. Always to the lead, never to another double.
2. Trim the consonants. Fade the doubles out just before ending “s” and “t” sounds and let the lead finish the word alone.
3. Remove breaths on the doubles. The lead’s breath is enough.
4. Tune lightly. Fix wrong notes, keep the natural movement. Perfectly matched pitch plus perfectly matched timing equals a phasey vocal.
5. Mute what doesn’t work. A double that’s only on the strongest words of a line is often better than one that’s on every word.

Aligning is the slow part. Ordinary Align does it syllable by syllable from two dropped files, keeps the double’s pitch, and marks anything the lead never sang — like an ad-lib that slipped into the double — instead of forcing it into place.

Mixing doubles

Level and panning

EQ

• High-pass higher than the lead. The doubles don’t need the low body; the lead has it.
• Less top end. Slightly darker doubles keep the lead in front and tame doubled sibilance.
• De-ess harder than the lead.

Dynamics and space

• Compress the doubles a bit more than the lead so they sit at a steady level underneath it.
• Share the space. Send lead and doubles to the same reverb so they sound like one performance, usually with a little less send on the doubles.
• Bus them. Route all doubles to one bus and ride that fader through the song.

Fake doubles, done well

No time to re-sing? A fake double can still help:

• Classic ADT: a copy of the vocal delayed somewhere around 10–30 ms with a slow, small modulation of the delay time. Keep it under the lead.
• Micro-pitch: two copies, one a few cents up and one a few cents down, each with a slightly different short delay, panned left and right.
• Filter it: high-pass and darken the fake double just like a real one.
• Don’t overdo width: big delay-based width can sound strange in mono. Check it.

When not to double

Doubling adds weight and takes away intimacy. A verse that should feel like someone talking in your ear often works better single-tracked, so the chorus doubles feel like a lift. If everything is doubled, nothing stands out — the contrast is part of the effect.

FAQ