In short
- Interval timing — judging and reproducing durations from a few hundred milliseconds to several seconds — is a trainable perceptual skill, separate from reading a clock.
- Timing error grows roughly in proportion to the interval: a ten per cent miss on one second is a ten per cent miss on three. This is Weber's law applied to time.
- Counting is a workaround that replaces the sense with arithmetic. It works, and it stops you training the thing itself.
- Comparing two durations and reproducing one are different skills; most real tasks need the second.
- Practice with immediate, measured feedback improves accuracy and shrinks bias. Nothing here claims to improve attention, productivity or anything beyond the timing itself.
What a sense of time is
Researchers call it interval timing: the ability to judge how long something lasted, or to produce an interval of a given length, in the range from a few hundred milliseconds to a few tens of seconds. It is distinct from circadian timing (knowing roughly what time of day it is) and from millisecond motor timing (the coordination inside a single movement). It is the range of musical phrases, spoken sentences, film cuts and the gap before a punchline.
You have no organ for it. There is no retina for duration. The sense is built from what the brain does with everything else that was happening while the interval ran, which is why it is so easily pulled around — by attention, by emotion, by how much was going on — and why two people can disagree about whether a pause was long.
What is known about it
Error grows with the interval
The most robust finding is that timing follows Weber's law, as most senses do: the variability of your judgement grows in proportion to the duration being judged. If your reproductions of a one-second interval scatter by about a tenth of a second, your reproductions of a three-second interval will scatter by about three tenths. The proportion — the coefficient of variation — stays roughly constant across the range, though studies disagree about exactly how constant. A practical consequence: judge timing errors as a percentage of the target, not in seconds. A quarter-second late on a one-second interval is a serious miss; on a four-second interval it is a good result.
The pacemaker-accumulator picture
The dominant model for decades, scalar expectancy theory, pictures an internal pacemaker emitting pulses that an accumulator counts while the interval runs; the count is compared with counts stored in memory from past intervals to produce a judgement. It explains the proportional error neatly. It has also been challenged — some researchers prefer models based on the decay of memory traces, and others on the dynamics of neural populations — but every model agrees on the behaviour: proportional error, sensitivity to attention, and improvement with feedback.
Perceiving a duration and producing one are different
Comparing two intervals — was the second longer? — is perceptual timing. Holding a key for the same length as an interval you just heard is motor timing: the duration has to be translated into an action, and the action has its own delays. Studies that test both find they are not simply the same timer read twice. Most of what people want from a better sense of time — landing the cut, placing the beat, holding the pause — is the motor kind, so that is the one to practise.
Why counting is a workaround
Counting replaces the sense with a procedure. Instead of feeling that an interval was about two seconds, you count "one-and-two-and" at a learned rate and read off the answer. It works, which is why people do it, and it is entirely respectable in a rehearsal. But it trains the rate of your counting, not the sense of duration, and it falls apart whenever counting is unavailable: while you are speaking, while you are playing something else, while you are watching for the moment to act.
A test of the sense itself has to make counting useless. A short interval with no regular pulse to count, a different length every time, and a scoring that cares about the exact length rather than the nearest whole second will do it. Percore's Duration session is built on exactly this: the interval is somewhere between about one and three and a half seconds, it is not announced, and there is, as the app's own description puts it, no way to count it out loud that will not lose you the round.
How to practise
Each of these gives you a target, an answer, a measured result and a record, which is what turns repetition into training. Do one, briefly, every day, under the same conditions.
- Reproduce intervals without counting. Hear or see an interval, then hold a press for the same length. Get the error as a percentage, and watch which way it goes: most people are consistently short or consistently long, and a bias is easier to correct than scatter.
- Hold a tempo without a click. Start with a metronome, let it drop out, and keep tapping. Measure how far you drift over eight or sixteen bars. A metronome app that mutes bars does this; so does a Tempo exercise that gives you a pace and asks for it back.
- Predict the arrival. Watch something move at a steady rate behind an occluder and tap when it would reappear. This trains the forward sense — when will it happen — which is the one a performer or an editor uses most.
- Change one condition at a time. Morning versus evening, quiet versus noisy, before and after coffee. Your sense of time moves with your state, and seeing that in your own numbers is half the lesson.
Where Percore fits
Percore's free daily practice includes Duration: a pulse breathes with a low tone for an unannounced length between about one and three and a half seconds, then stops, and you press and hold anywhere until it feels like the same length. Releasing submits. The score is the error relative to the target, so it is a Weber-honest measurement — a miss is judged against the interval it happened in. Everyone gets the same targets on the same day, and the series is compared only with your own recent weeks.
Pro adds the other three timing modules for practice outside the daily four: Tempo (match a pace, not just a duration), Prediction (anticipate timing before it happens) and Reaction (respond the instant a signal appears — the one timing skill that is about speed rather than judgement). None of them teach rhythm notation, and the rhythm training apps list says which apps do.
Find out how your sense of time runs, free, in half a minute.
Download on the App StoreFrequently asked
Can you train your sense of time?
Yes. Interval timing improves with practice when each attempt gets immediate, measured feedback. Accuracy improves and, often more usefully, a consistent bias — always short, always long — becomes visible and shrinks.
Why does time feel faster or slower on different days?
The sense of duration is built from what the brain was doing while the interval ran. Attention, arousal, fatigue and how much was happening all pull it around, which is why your reproductions vary with your state and why a daily measurement is worth more than a single one.
Is counting cheating?
Not in real life — count whenever it helps. But in a test of your sense of duration it measures your counting rate instead of the sense, and a well-designed test makes it impossible or costly.
What is a normal error on a duration test?
It depends on the interval. Variability is roughly proportional to the length being judged, so judge yourself in percentages. Being within about ten per cent of a few-second interval without counting is a good result for most people.
Does a better sense of time make you a better musician?
It is one component. A musician also needs rhythm — the patterns — and the motor skill to execute them, which rhythm trainers and practice build. A steadier internal sense of duration and tempo is the foundation those sit on, not a substitute for them.