In this lesson
Ad-libs are arrangement moves
Ad-libs that sing over the lead compete with it for attention. Put them in the gaps, away from the centre, and they push the song forward instead.
Updated · 5 min read · Lesson 11 of 12 in Vocal Production
In short
- A listener can follow one line of words at a time, and an ad-lib in the same voice as the lead is the hardest kind to separate.
- Time is the strongest separator: an ad-lib that lands in the gaps of the lead needs little else.
- Give each ad-lib a slot, a position and a register, and cut the ones that sing words over the lead.
In this lesson (8)
The hook got busier and smaller
The lead vocal is great. The delivery is right and the tuning is solid. Then you add three tracks of ad-libs and harmonies, recorded in one pass over the whole song, and the lead disappears. The hook sounds like a crowd. Turning the ad-libs down loses the energy, and panning them just makes the stereo picture messy.
The ad-libs are being treated as decoration, something to fill every space. A listener can only follow one line of words at a time. Ad-libs work when they are arranged: placed in time, in space and in register so they answer the lead instead of talking over it.
Why it matters
Two sets of words at once
When an ad-lib sings words over the lead's words, the listener has to choose which to follow. Speech research shows how hard that gets when the voices are alike. When Brungart (2001) asked listeners to pick out a target phrase against one competing phrase, they did worst when both came from the same talker and best when the two talkers were of different sexes. Most of the difficulty came from confusing the two voices, not from one covering the other's frequencies.
An ad-lib is usually the same singer on the same mic, which is the hardest case of all. So ad-libs that carry words should stay out of the lead's words. Ad-libs that must overlap should stop being new words: a held vowel, a shout, or a repeat of the lead's last word.
The science
What separates two vocal parts
describes the cues the brain uses to split a mixture of sound into separate streams (Bregman, 1990). For two vocal parts, these are the ones you control:
- Time. Parts that start at different moments separate most easily. An ad-lib in a gap of the lead needs little other help.
- Register. An octave up or down, or a clearly different range, sets a part apart.
- Position. Sound from a different direction is easier to follow on its own. In Cherry's 1953 experiments, two messages played to separate ears were far easier to tell apart than two messages mixed together. Panning in a mix is a milder version of the same cue.
- and depth. A filtered, distorted or more reverberant ad-lib sounds like a different source from a dry, full lead.
No formula adds these up, and you do not need all of them at once. Use as many as it takes for the ad-lib to sound like a reply, starting with time.
Try it
Clear the gaps
0 of 6 steps done
The ad-libs should now sound like answers that drive the groove, and the lead should read clearly through the whole chorus.
Common mistake
Ad-libbing through the whole song
Artists often record ad-libs in one continuous pass, singing along with the lead and adding shouts and harmonies from start to finish. Keep all of it and you have a second lead vocal that never stops, competing with every line.
Edit the pass like any other part. Keep the phrases that land in the gaps and a few doubles on key words, and delete the rest. With fewer ad-libs, each one gets noticed.
Takeaway
Give every ad-lib a slot
Treat each ad-lib as an arrangement part with a time slot, a position and a register. Keep the centre and the lead's words for the lead. Ad-libs that answer in the gaps, from the sides and a little further back, push the song forward instead of crowding it.
Auditory scene analysis
How the brain sorts a mixture of sound into separate sources, such as a voice, a kick and a pad.
Timbre
What lets you tell two sounds apart at the same pitch and level: the balance of harmonics and noise, and how they change over time.
3 questions
Check yourself
Sources
Show all 3Hide sources
- Bregman, A. S. (1990). Auditory Scene Analysis: The Perceptual Organization of Sound. MIT Press.
- Brungart, D. S. (2001). Informational and energetic masking effects in the perception of two simultaneous talkers. Journal of the Acoustical Society of America, 109(3), 1101-1109.
- Cherry, E. C. (1953). Some experiments on the recognition of speech, with one and with two ears. Journal of the Acoustical Society of America, 25(5), 975-979.
Next lesson · 12 of 12
Keep the quiet detail a gate would cutThe fading ends of words and the soft starts of lines sit just above the noise, right where a gate acts. Clean the gaps by hand, then compress.Previous: Why a vocal double works best unnoticed

Written by Virzy Guns
Songwriter and producer, founder of Virzy Guns Production. Now building HealingWave.