Comping and Alignment
Pick the best fragments from several takes and join them into one seamless track.
Related
The problem
A single pass rarely gets everything right: one phrase has tuning, another has feel. Comping chooses the best fragments from several takes and joins them into one track whose seams no one hears.
Signal chain
Record the section three to five times, each on its own track and aligned to the same start point, only one playing at a time. Join the chosen fragments with short crossfades so room tone, breath and tails stay continuous.
Placement and settings
Put edit points before a consonant or inside a breath, never inside a sustained note. Use crossfades of 5 to 20 ms, scaled to the material. Check each phrase in mono, then play the whole take to hear whether level and tone jump at the joins.
What to listen for
Listen first for a click or sudden silence at the join, then for tone changes as two versions of the same phrase alternate. Play the whole section and notice whether the breathing is broken and the accents still fall in the same place.
Common misconceptions
- Comping only removes wrong notes → it chooses interpretation, not just accuracy.
- Shorter fades are always safer → too short a fade leaves a click.
- Alignment is a visual job → after matching waveforms, confirm attacks by ear.
Listen in the library
All tracks from midicn-lib — inline badge is the license tier (C1 commercial / C2 non-commercial / C3 study only).
Sources
- AES Convention paper on edit-point perception and crossfade length
- Eargle, Handbook of Recording Engineering, editing and assembly chapters
Used for fact-checking only; all prose is original.