Choir Intonation: How Singers Tune to Each Other Without a Piano
A piano fixes its pitches before you ever touch a key. A choir has to agree on all of them, live, every time.
Short answer: when a choir sings without an instrument, there is no fixed pitch grid underneath them the way there is on a piano, so every chord is tuned in real time, by ear, through the singers listening to each other. Measurements of real choir singers found that this is not a vague impression: singers measurably adjust their pitch toward acoustically pure, beatless intervals when singing a cappella, a tendency documented directly by tracking actual sung frequencies rather than by asking listeners what they thought they heard (Ternström and Sundberg, 1988). The mechanism and its practical consequences are laid out in a dedicated survey of the topic (Howard, 2015). None of this replaces the tuning-system physics covered on our equal temperament page or the beat-counting mechanics on our tuning-by-ear page; this page is about what changes when the tuning has to be negotiated between several people listening to each other instead of decided in advance by one instrument.
What "in tune" means when nothing is fixed
A piano commits to twelve pitches per octave before a single note is sung, in equal temperament, so every interval on it is a known, fixed compromise. An unaccompanied choir has no such commitment. Each singer produces a pitch, listens to what the other voices are doing, and adjusts, and the chord that results is whatever the group converges on in that moment. Grove Music Online's entry on intonation treats this real-time, ear-led tuning as a distinct category from tuning to a fixed instrument, precisely because the target is not a predetermined frequency, it is whatever makes the sounding chord lock together (Grove, "Intonation," 2001).
The direction that convergence tends to take is not random. Left to their own ears, singers tend to narrow the choir's major thirds and adjust other intervals toward the small-integer ratios described on our just intonation page, because those ratios are exactly the ones where the chord's overtones line up and the beating disappears. A choir singing with a piano cannot do this; the piano's pitches are fixed, so the choir has to compromise toward the instrument instead. That is the acoustic reason a good a cappella choir and the same choir singing with piano accompaniment can sound subtly different on the same chord, even when both performances are considered "in tune."
What Ternström and Sundberg actually measured
Rather than relying on listener impressions, Ternström and Sundberg recorded real choir singers and measured the actual frequencies they produced across sustained passages (Ternström and Sundberg, 1988). The study quantified how precisely singers held a pitch, how much a chord's intonation shifted over the course of a phrase, and how those shifts related to the chord's harmonic content. It is one of the foundational pieces of evidence that choir intonation is not simply "good ears versus bad ears": trained singers systematically adjust pitch in ways that track the acoustic structure of the chord they are inside, which is a different and more specific claim than saying some singers are more accurate than others.
Vertical listening versus horizontal listening
Every singer in a choir is doing two listening tasks at once, and they can pull in different directions. Horizontal listening is tracking your own part: is this note the correct distance from the note I just sang. Vertical listening is checking your note against everyone else sounding at that instant: does this chord lock, or is it beating. A melodic line that is perfectly in tune with itself, judged note to note, can still sit slightly wrong inside the chord if the singer is not also listening vertically, which is why conductors often ask a choir to "listen down" into the harmony rather than only forward along their own line.
| Listening mode | What you are checking | Typical adjustment |
|---|---|---|
| Horizontal | your own melodic interval, relative to the note you just sang | matching a remembered scale-degree distance in your own part |
| Vertical | your note relative to the notes sounding under and over you right now | narrowing or widening slightly to remove beating against the other voices, the same mechanism explained on our beats page |
Howard's survey of the topic frames staying in tune in a cappella singing as a continuous balancing act between these two demands, not a single skill that either is or is not present (Howard, 2015). A singer who is excellent at one and weak at the other will sound fine alone and slightly wrong in the ensemble, or the reverse.
Practical mechanics for rehearsal
Because the target is a beatless chord rather than a fixed frequency, the practical technique is the same beat-listening skill covered on our tuning-by-ear page, applied voice to voice instead of instrument to tuner. Build chords from the bass upward rather than from the top down, since the bass note anchors which overtones the rest of the chord needs to line up with. Sustain problem intervals, especially thirds, well past the point where you would normally move on, and listen specifically for whether the roughness is increasing or decreasing as you adjust; that is the same beat-rate cue used to tune an instrument, just applied to two singers instead of two strings. And separate the two listening tasks in rehearsal deliberately: sing a phrase attending only to your own line, then repeat it attending only to how your note sits inside the chord under a held drone from the rest of the section.
What our games do here, honestly
None of Hearkle's eleven games involves more than one voice at a time, so nothing here trains vertical, ensemble listening directly; that is a real gap, and a choir rehearsal is the only place that specific skill actually gets built. What our games can sharpen is the underlying judgement choir intonation depends on: noticing how far off a pitch is from a reference, which is exactly what a singer is doing, faster and less consciously, every time they nudge a note to kill a beat in a chord. Hearfork trains that judgement directly, on single tones against a reference.
How small a pitch gap can you actually hear? Choir intonation is built from dozens of tiny adjustments like this, judging exactly how far off a note is and nudging it the right way. Hearfork measures that threshold directly, one tone at a time.
Keep reading
- Equal temperament vs just intonation: the ratio math behind why some chords lock and others do not.
- Beats and tuning by ear: the beat-counting mechanism singers and instrumentalists both rely on.
- Pitch drift and holding a note: the separate, memory-based reason ensembles can slide off pitch over time.
- Why you cannot sing in tune: the solo perception-to-production chain that choir intonation builds on top of.
Frequently asked questions
Why does an a cappella choir sometimes go flat over the course of a piece?
That is usually a different mechanism from the chord-tuning adjustments described on this page. Losing pitch gradually across a long passage is closer to the memory-and-matching drift covered on our pitch drift page: without an external reference, each singer's held note leans on a fading internal memory of the target, and a whole section can migrate together without anyone noticing in the moment. Chord-by-chord tuning adjustments, by contrast, are the choir actively converging on a beatless chord in real time, which is a feature, not a drift error.
Is choir intonation the same thing as being able to sing in tune solo?
No. Singing a melody accurately alone is mostly a perception-to-production chain, covered on our page about why people sing out of tune. Choir intonation adds a second, simultaneous task on top of that: continuously checking your note against everyone else sounding at that instant and adjusting, which is a listening skill a solo singer never has to exercise.
What did Ternström and Sundberg actually measure?
They recorded real choir singers and tracked the exact frequencies they produced during sustained singing, rather than relying on what listeners thought they heard. That let them quantify how precisely singers held pitch and how a chord's tuning shifted over a phrase, providing direct acoustic evidence for adaptive, ear-led tuning in trained choirs (Ternström and Sundberg, 1988).
Why does a chord "lock"?
When a chord's pitches sit close to small-integer frequency ratios, the overtones of the different notes coincide instead of clashing, so the beating and roughness that come from near-but-not-quite-matching partials disappear. The physics of why that happens is covered in full on our equal temperament and just intonation page; this page is about the fact that unaccompanied singers audibly chase that locked state by ear.
Should choirs tune to equal temperament or just intonation?
Left unaccompanied, choirs tend to drift toward the purer, beatless intervals of just intonation on their own, because that is what their ears are rewarding as they adjust. Singing with a fixed-pitch instrument like a piano removes that option, since the instrument's pitches are already fixed in equal temperament, so the choir compromises toward the instrument instead. Neither is a mistake; they are different constraints.
What is the difference between horizontal and vertical listening?
Horizontal listening tracks your own melodic line, checking that each note is the right distance from the one before it. Vertical listening checks your note against the other voices sounding with you right now, listening for whether the chord is beating. Choir intonation depends on doing both at once, and the two demands can genuinely pull a singer in different directions.
Sources: Howard (2015), Intonation and Staying in Tune in A Cappella Choral Singing, in The Oxford Handbook of Singing, Oxford University Press; Ternström and Sundberg (1988), Intonation precision of choir singers, The Journal of the Acoustical Society of America 84(1); Intonation, Grove Music Online, Oxford University Press.