Comping

Comping, short for compositing, is the process of recording several takes of the same part and assembling the best sections of each into one final performance, called the comp.

It is standard practice for lead vocals, but engineers comp guitar solos, bass, drums and MIDI performances in exactly the same way. Done well, nobody can hear it: the listener just hears a great take.

What is comping?

Comping is a selection and editing process. You record a part several times over the same section of a song, audition the takes, pick the strongest version of each line, phrase or word, and join those pieces into one continuous performance. The joins are smoothed with short crossfades so the result sounds like a single uninterrupted take.

In plain language: a singer records the song five times, and the producer builds one “perfect” vocal out of the best moments of all five.

How the industry defines it

Sound On Sound describes comping as recording several takes of a lead performance so you can create a final, composite track. Avid’s Pro Tools documentation describes it as constructing a performance by selecting and combining the best parts of multiple takes, and Apple’s Logic Pro guide calls its version Quick Swipe Comping: swiping across takes to select the best parts and combine them into one master comp.

Where the word comes from

“Comp” is short for composite: one performance made from several. Recording engineers were compositing performances long before software gave the process its own buttons, first by cutting tape and later by bouncing the best parts of several tracks onto one.

Comping in jazz means something else

Musicians also use “comping” for something completely unrelated. In jazz, comping (from accompanying or complementing) is the rhythmic chordal accompaniment a pianist or guitarist plays behind a soloist, reacting to the soloist’s phrasing rather than repeating a fixed pattern. Same word, different craft. In a recording or production context, comping always means compositing takes.

“Comp” can also mean compressor

In studio shorthand, “comp” is also used for a compressor (“put some comp on the snare”). Context usually makes it clear, but beginners often mix the two up. Compositing takes has nothing to do with dynamics processing.

The history of comping

The tape era

Multitrack tape made comping possible. Once engineers could record several performances on separate tracks, they could combine the best parts of each. On tape this was done in two ways. The first was splicing: physically cutting the tape with a razor blade and joining sections from different takes, usually with an angled cut so the transition acted like a tiny crossfade. The second, and more common for vocals, was bouncing: recording several vocal passes on different tracks, then playing them back through the console and switching between them in real time while recording the result onto a free track.

Both methods were slow and risky. A bad splice could ruin a master, and every bounce to tape added a generation of noise. With 16- and 24-track machines in the 1970s and 1980s, engineers could dedicate several tracks to vocal takes, and building the comp by muting and unmuting console channels became a studio skill in itself.

The DAW era

Digital audio workstations turned comping from a risky, destructive job into a fast, non-destructive one. At first, engineers simply stacked takes on separate tracks and cut and muted regions by hand. Dedicated comping interfaces followed:

Year	DAW	What changed
2007	Logic Pro 8	Take folders and Quick Swipe Comping: loop-recorded takes are packed into a folder and you swipe across them to build the comp.
2008	Pro Tools 8	Playlist view shows every take under the track, with shortcuts to copy selections to the main playlist.
2011	Cubase 6	Lane-based multi-take comping.
2021	Ableton Live 11	Take lanes, bringing linear comping to a DAW built around loops (Live 11 manual).
2021	Bitwig Studio 4	Comping in the arranger and inside clips in the Clip Launcher (MusicTech).
2023	Reaper 7	Fixed item lanes and swipe comping (The REAPER Blog).

How comping works

Non-destructive editing

When you comp in a DAW, nothing is deleted. Each take stays on disk as its own audio file. The comp is just a list of instructions telling the DAW which part of which file to play, and where. That is why you can change your mind at any point, and why most DAWs let you keep several alternative comps of the same part.

Edit points and zero crossings

A digital waveform swings above and below a zero line. If you cut from one take to another at a point where the two waveforms are at very different values, the signal jumps instantly and you hear a click or pop. Cutting at a zero crossing, where the waveform passes through zero, reduces that risk, and most DAWs can snap edits to the nearest zero crossing. In practice, though, two different takes rarely line up perfectly, which is why crossfades matter more than zero crossings.

Crossfades: equal power versus equal gain

A crossfade fades the outgoing take down while the incoming take fades up. The shape of the two curves decides whether the level stays steady across the edit:

• Equal power crossfades use curved fades, so each side is at about 70.7% (−3 dB) at the midpoint. For unrelated material, such as a breath, a consonant, a room tone or a different word, this keeps the perceived loudness constant. Most vocal comp edits fall into this category.
• Equal gain crossfades use straight lines, so each side is at 50% at the midpoint. For near-identical, in-phase material, such as the same sustained vowel from two takes lined up cycle by cycle, the two halves add up to exactly full level. Using equal power here would produce a small bump of about 3 dB in the middle of the fade.

Crossfade length matters too. A few milliseconds is enough to kill a click on a cut in silence or on a consonant. Longer crossfades are useful across sustained sounds or noisy room tone, but if they are too long you may hear both takes at once, which sounds like a brief doubling or phasing.

Hiding edits with masking

The ear is less likely to notice a change when something louder or noisier is happening at the same moment. Engineers use this masking effect by placing edits on hard consonants such as t, k, p or ch, or right on a drum hit, where the burst of noise covers small differences in tone between the two takes.

Comping in your DAW

Every major DAW does the same thing under the hood, but the names differ, which is why beginners often cannot find tutorials for their own software.

DAW	What it is called	How it works
Pro Tools	Playlists	Takes are stored as alternate playlists and shown in playlist view under the main track. Select the best part of a take and copy it to the main playlist (Control+Option+V on Mac); takes can be rated to filter the best ones.
Logic Pro	Take folders, Quick Swipe Comping	Loop recording packs takes into a take folder. Swipe across a take to add that part to the comp; crossfades are added automatically.
Ableton Live	Take lanes	Loop recording creates take lanes under the track. Select part of a lane and send it to the main lane.
Cubase	Lanes, Comp tool	Show lanes to see the takes and click sections with the Comp tool to bring them to the front.
Fender Studio Pro (formerly Studio One)	Layers	Takes are recorded to layers; selected ranges are promoted to the main layer.
Reaper	Fixed item lanes, swipe comping	Takes are recorded to lanes; swipe across them to build the comp lane.
Bitwig Studio	Comp takes	Comp in the arranger or inside clips in the Clip Launcher, with adjustable boundaries between takes.
FL Studio	Playlist tracks	No dedicated swipe-comping tool; record passes to separate playlist tracks, slice and mute.

Once the comp is finished, most engineers consolidate (flatten, bounce or render) it into a single clip before tuning and processing. This bakes in the crossfades and prevents plugins such as Melodyne from losing track of the audio if the comp is changed later.

How to comp vocals: a practical workflow

1. Set up the session so the takes match

A comp only sounds seamless if the takes sound alike. Keep the mic, mic position, preamp gain and the singer’s distance from the mic the same for every take, and record takes you plan to combine in the same session where possible. A line recorded on a different day, or with the singer half a step further back, will jump out in tone even when the performance is right.

2. Record enough takes, but not too many

Get a full pass of the song first, so the singer finds the story of the lyric. Then record the sections that need it (verse, pre-chorus, chorus), which keeps energy matched to each part of the arrangement. A handful of strong passes, often four to eight, is usually enough. Beyond that, voices tire, takes start to sound the same, and choosing between twenty near-identical versions becomes paralysing.

3. Take notes while you record

A lyric sheet with notes (“take 3, great verse 1”) or markers in the DAW saves hours of blind auditioning later. Some producers do it in their heads: vocal producer Kuk Harrell told Sound On Sound that he comps as the session goes along, recording in Pro Tools playlists and “making mental notes” rather than physical ones (Sound On Sound, October 2019).

4. Choose by feel first, pitch second

Listen for emotion, tone, phrasing and timing before pitch. A slightly flat note with the right feeling can be tuned; a perfectly pitched line with no life cannot be fixed. Pick the best take as your base and replace only what needs replacing, rather than building every word from a different take.

5. Put the edits in the right places

• In silence between phrases, the safest place of all.
• On or just before hard consonants (p, t, k, ch), where airflow stops for a moment and the noise masks the join.
• Around whole breaths. Never cut through the middle of a breath: keep each breath with the phrase that follows it and cut in the gap before the breath, or you risk a “double breath” that sounds like the singer is gasping.
• Inside a steady vowel only when you must, using a very short equal gain crossfade on waveforms that line up.

6. Take a break, then comp again

Perspective fades during long editing sessions. Producer Ian Kirkpatrick explained to Sound On Sound that on Dua Lipa’s “Don’t Start Now” he comps the vocal once, hides it, comps it again from scratch, then does it a third time. He tunes and times all three comps (labelled 1C, 2C and 3C) and builds the final vocal from the best of them (Inside Track, May 2020). You do not need three full comps on every song, but taking time away and listening again with fresh ears works.

7. Listen without looking

Finally, turn off the screen and play the comp from start to finish, ideally on a different pair of speakers or from across the room. Waveforms make you hear edits you would never notice otherwise; the only question that matters is whether the performance sounds like one person singing one take.

Comping across genres

• Pop and EDM: comping is often detailed, phrase by phrase or even word by word, and is followed by tight tuning and timing edits.
• Hip-hop and trap: timing and delivery come first. Comps focus on the rhythm of each line, with doubles and ad-libs comped on their own tracks.
• Rock, jazz and classical: engineers usually comp in larger blocks (whole verses or sections) to keep the performance and the interplay between musicians intact. Over-comping here quickly removes the human feel that makes the recording worth listening to.

After the comp: tuning, timing and cleanup

Comping, timing and pitch correction are separate jobs, and the usual order is comp first, then timing, then tuning, then processing. Tuning a take you later throw away is wasted time, and a strong comp needs far less correction.

• Pitch correction: tools such as Celemony Melodyne or Auto-Tune are applied to the consolidated comp.
• Alignment: tools such as Synchro Arts VocAlign match doubles and backing vocals to the timing of the comped lead.
• Cleanup: spectral repair tools such as iZotope RX remove clicks, mouth noise or hum that editing can expose, and can reduce breath levels.

Common mistakes and myths

• The “Franken-take”: comping syllable by syllable for technical perfection until the vocal loses its flow and emotional arc.
• The double breath: cutting between two takes in the middle of a breath, so the singer seems to inhale twice.
• No crossfades: butting clips together without fades almost always leaves clicks.
• Mismatched takes: combining takes recorded with different mic positions, levels or days, so the tone changes from line to line.
• Tuning before comping: wasted work, and it can create problems when the comp changes.
• Myth: comping is cheating. Comping is standard on virtually every commercial record, whatever the singer’s ability. The studio is an unforgiving, microscopic environment, and comping is how producers keep the best moments of a performance.
• Myth: comping and pitch correction are the same thing. Comping chooses and joins real performances; pitch correction changes the notes afterwards.

Comping vs related techniques

Comping vs	The difference
Punching in	Punching in re-records a short section over an existing take. Comping records full takes first and assembles them afterwards.
Double tracking	Double tracking records a second performance to play alongside the first for thickness. Comping creates one lead performance; the doubles are usually recorded and edited to match it.
Jazz comping	Accompanying a soloist with chords. Unrelated to editing takes.
Take folders vs track alternatives (Logic Pro)	Take folders hold several takes of the same part for comping. Track alternatives store completely different versions or ideas for the same track.

Related terms

• Crossfade: what makes every comp edit inaudible.
• Auto-Tune: pitch correction that comes after the comp.
• Transient: a useful place to hide an edit.
• Recording: capturing the takes you will comp.
• Stems: consolidated tracks, including comped vocals, delivered for mixing.
• Compressor: the other “comp” in studio slang.

FAQ