Skip to content
Singler

The science behind Singler

Every exercise, measurement and tip in Singler is based on voice research, motor learning or established coaching practice. Here you can see what it rests on and how solid the evidence is. Reading this is optional; the app works without it.

Our principles

  • We cite sources and say honestly how strong the evidence is.
  • Scores and tips are coaching aids, not medical or clinical assessments.
  • Compare yourself with yourself over time, not with norms.

How you learn

Real-time visual feedback

Evidence: Promising, still being studied

Seeing your pitch while you sing gives you information your ears alone can't provide in the moment. In a study with 56 adults, from untrained to trained singers, those who practiced with a live pitch display improved their accuracy; a group without the display did not. Reviews see it as a promising tool, especially combined with guidance.

How Singler uses this

The pitch graph and the meter show your voice live against the target. Tip: try some attempts without looking, so you don't come to depend on the screen.

Sources (2)

The right level of challenge

Evidence: Well researched

Motor learning works best when the task is neither too easy nor too hard for your current skill. Too easy gives little to learn from, too hard overwhelms. This idea is known as the challenge point framework.

How Singler uses this

After each round of "Match the note" the tolerance adapts: with 80 % accuracy or more it gets tighter, below 40 % it gets wider. Target notes come from your comfortable range.

Sources (1)

Practice regularly

Evidence: Well researched

For learning motor skills, several shorter practice sessions spread over time work clearly better than one long session. A meta-analysis of 63 studies found a medium to large effect.

How Singler uses this

Rounds are deliberately short (10 notes). Ten to fifteen minutes a day work better than an hour once a week. The progress page counts your practice days.

Sources (1)

Singing in tune

Singing in tune can be learned

Evidence: Well researched

About 85–90 % of people without vocal training sing reasonably in tune. Most of the remaining 10–15 % hear pitch differences just fine. The difficulty is usually in translating a heard pitch into the voice, which is a coordination skill and can be practiced.

How Singler uses this

"Match the note" trains exactly this link: hear a note, sing it back, see the result. It measures accuracy, how quickly you find the note, and whether you tend to sing sharp or flat.

Sources (3)

Melodies instead of single notes

Evidence: Established coaching practice

Singing in tune means moving between pitches, not just holding one. That's why research on singing accuracy often has people imitate short note sequences. Hearing the melody first and singing it inwardly is a long-standing teaching method; there is little direct research comparing approaches.

How Singler uses this

Melodies are generated in a major key within your comfortable range. They start and end on the key note and move mostly in steps; harder levels are faster and use larger jumps. The first 200 ms of each note aren't scored, since the voice needs a moment to move to the new pitch.

Sources (1)

Cents and tolerance

Evidence: Well researched

Pitch differences are measured in cents: 100 cents make one semitone. Untrained listeners notice differences of roughly 15 cents, trained musicians about 2 cents, and a few hours of ear training narrow the gap. Singing research usually counts a deviation of 50 or 100 cents as a clear error.

How Singler uses this

The green band is your tolerance. It starts at ±50 cents (a quarter tone) and narrows to ±15 cents as you improve. Going tighter would demand differences that are hard to hear at all.

Sources (3)

Starting on the note

Evidence: Established coaching practice

Sliding up into a note from below ("scooping") is a common habit. Voice teachers recommend hearing the note inwardly before singing and starting directly on it with a relaxed onset. There is little direct research on this; it is long-standing teaching practice.

How Singler uses this

If the first 150 ms of a note are clearly flat and then rise, the attempt counts as a scoop. If that happens often, you'll get a tip with an exercise.

Sources (0)

No direct studies; based on common voice-teaching practice.

Your voice

Range and voice type

Evidence: Established coaching practice

Your range runs from the lowest to the highest note you can produce. Voice teachers classify voices mainly by tessitura (where the voice feels at home), timbre and where the registers change, not by range alone. Practicing constantly at the extremes strains the voice.

How Singler uses this

The range test measures your extremes. The comfortable range used for exercises is estimated a few semitones inside them. The voice type is only a rough indication from your range.

Sources (1)

Holding a note (phonation time)

Evidence: Well researched

Maximum phonation time, the longest you can hold a vowel on one breath, is a standard measure in voice clinics. In one study, healthy adults over 70 held a vowel for about 21 seconds (women) and 23 seconds (men); values vary between studies. Single attempts fluctuate; the average of several attempts on different days is reliable.

How Singler uses this

"Hold the note" measures the duration and also how much of it you spend on target. Compare yourself with your own earlier results; there are no grades against norms. Don't squeeze out the last bit of air.

Sources (2)

Steadiness and drift

Evidence: Established coaching practice

A steady pitch on a held note reflects even airflow and fine control of the larynx. When breath support fades, the pitch often sinks slowly toward the end. This comes from coaching experience rather than from controlled studies.

How Singler uses this

Steadiness is the remaining wobble once the overall trend is removed; vibrato doesn't count against it. Drift is the trend itself, in cents per second.

Sources (0)

No direct studies; based on common voice-teaching practice.

Vibrato

Evidence: Well researched

Vibrato is a regular, slight oscillation of pitch. In a study of ten classical singers it averaged about 6 times per second and became a little faster toward the end of notes. The typical classical range is about 4.5 to 6.5 Hz; in many pop styles vibrato is used less or not at all.

How Singler uses this

Singler looks for a regular oscillation between 4 and 8 Hz with an extent of at least 20 cents. Vibrato is a matter of style: it's shown, not scored.

Sources (1)

Warm-up and vocal health

Warming up

Evidence: Promising, still being studied

Most singers feel that their voice works better after warming up. Measurable effects have been found, for example in vocal stability and comfort, but they vary a lot from person to person. The evidence is mixed, the risk is low.

How Singler uses this

The warm-up takes about five minutes. It's optional, and each tip links to the step that fits it.

Sources (2)

Semi-occluded vocal tract exercises

Evidence: Well researched

Humming, lip trills and singing through a straw partly close the mouth. This creates back pressure that helps the vocal folds vibrate efficiently with less effort. A review of studies with singers found the most consistent benefits for vocal comfort, resonance, phonation time and flexibility.

How Singler uses this

Four of the five warm-up steps use this principle: humming, lip trills, straw phonation and sirens on "ng".

Sources (2)

Vocal health

Evidence: Well researched

Drinking enough supports the vocal fold tissue; research on this is growing but not yet conclusive. According to the clinical guideline, hoarseness that doesn't improve within four weeks should be examined by an ENT doctor, and sooner with pain, breathing problems or sudden loss of voice.

How Singler uses this

Tips are coaching suggestions, never diagnoses. If your voice is tired or hoarse, rest is better than practice.

Sources (2)

How the app measures

How the app hears your voice

Evidence: Technical method

Singler estimates the fundamental frequency of your voice with the McLeod Pitch Method, a method developed for musical instruments and voice that measures how similar the sound wave is to itself after a time shift.

How Singler uses this

Analysis runs about every 5 ms in a background thread on your device. A clarity value filters out noise, a short median filter removes outliers. Audio never leaves your device.

Sources (1)

Microphone calibration

Evidence: Technical method

Pitch detection is reliable when your voice is clearly louder than the background noise (signal-to-noise ratio). The threshold must lie between the two: too low lets noise through, too high cuts off quiet notes.

How Singler uses this

Calibration measures 3 seconds of silence and 3 seconds of singing and puts the threshold in between, closer to the noise. A distance of 20 dB or more is very good. Once a note has started, the threshold is relaxed so quiet passages aren't cut off.

Sources (0)

No direct studies; based on common voice-teaching practice.

How the scores are calculated

Evidence: Technical method

The skill scores are simple, transparent rules of thumb, not clinical measurements. They are meant to show your development over time.

How Singler uses this

Pitch accuracy: share of time inside the tolerance band. Note onset: how quickly you hit the note, without scooping. Steadiness: 5 cents of wobble = 100, 30 cents = 0. Breath: 20 seconds = 100, based on phonation time studies. Range: two octaves = 100.

Sources (2)