What we mean by a good ear
Say that someone has a good ear and most people picture perfect pitch: hearing a note in isolation and naming it on the spot. In practice, musicians rarely need that. Tuning to the rest of an ensemble, working out a melody by listening to it, checking whether your singing has drifted away from the accompaniment — none of these ask you to name a note. What they ask is how far apart two pitches are. The musicians praised for having a good ear are usually the ones who judge that distance accurately. Treating pitch perception as a single ability blurs together skills that behave quite differently, which is why the phrase causes so much confusion.
Counting the distance between notes
Pitch can be stated as a frequency, but subtracting one frequency from another gives a number that does not match what listeners hear. Ten hertz low in the range is a conspicuous jump, while the same ten hertz higher up is barely noticeable, because the ear responds to ratios rather than differences. The unit that solves this is the cent, defined as one hundredth of a semitone. Adjacent keys on a piano sit 100 cents apart, and an octave spans 1200 cents. On this scale the same number of cents corresponds to the same perceived distance anywhere in the range, which is why cents are the standard way to discuss fine pitch differences.
How fine a difference people notice
The smallest gap at which two notes stop sounding identical is called the discrimination threshold, and it moves considerably with conditions. Longer notes are easier to compare than short ones. Timbres rich in harmonics can supply extra cues, and background noise pushes performance down. Even for one listener the value shifts across the range. Trained musicians are reported to detect smaller differences than untrained listeners, but that gap is not a fixed endowment; it reflects familiarity with the task of comparing notes as much as anything about the ear. This is why no single number can stand as a person's threshold: comparisons only mean something when the conditions are held steady.
Hearing is a different axis
Pitch perception and hearing are easy to conflate, yet they sit on different axes. Hearing asks how quiet a sound you can still detect. Clinicians answer it in a sound-treated room with calibrated equipment, finding the faintest audible level at each frequency and recording the results. Pitch perception asks something else: given two notes you can hear perfectly well, can you tell which is higher? The loudness stays constant throughout. Someone with unremarkable hearing may or may not be good at discriminating pitch, and the reverse holds too. If anything about the way you hear concerns you, a pitch test offers no useful evidence — that is a question for a doctor.
Keeping the equipment out of the result
Any measurement built on sound risks producing a score that reflects the playback hardware rather than the listener. High frequencies are the clearest trap: built-in speakers often fail to reproduce them, so a note that goes unheard leaves no way to tell whether the ear or the device fell short. The way around this is to stop measuring absolute sensitivity and rely only on relative judgement. When two notes share the same loudness and timbre and differ only in pitch, deciding which one is higher barely depends on the frequency response of the speakers or the volume setting. Keep the notes within the range of ordinary speech and any device will reproduce them. Designing a measurement means building conditions in which nothing except the thing you want to measure can affect the result.
Reading it as your own record
Pitch discrimination responds to practice, and that improvement is hard to separate from simply getting used to the task. Once you know how to break the sound apart and where to direct your attention, the same ear produces better scores. That fact argues for watching your own trajectory rather than comparing yourself with other people. Measure repeatedly under the same conditions and you start to see how sleep, alertness, time of day and the noise around you feed into the result. The pitch perception test on Bench works this way: it narrows the gap between two notes step by step and records the smallest difference you could still hear. The useful signal is not the size of the number but how it moves.