Pitch

What it is

Pitch is a perceptual precision game in the Senso collection that targets one of the most demanding feats in auditory cognition: holding an exact frequency in memory and retrieving it accurately seconds later. Unlike melody or harmony exercises that rely on the relationship between notes, Pitch strips everything back to a single, isolated sine tone and asks you to reproduce it from memory alone.

Each round presents a pure tone — a mathematically simple waveform with no overtones — generated directly by the Web Audio API. Your goal is to drag a slider up or down to sweep a live tone until it matches the one you just heard, then lock in your answer. The game uses headphones for best results because small frequency differences are much harder to detect on speakers. Over five rounds, your score accumulates, and your total reveals just how finely your auditory system can anchor and recall an absolute frequency.

How to play

At the start of each round a pure sine tone plays for approximately one second. Listen carefully — that tone is your target. After playback ends, a draggable control appears. Drag upward to raise the frequency of a live comparison tone and downward to lower it. The live tone plays continuously as you drag, letting you zero in by ear in real time. When you are satisfied that the live tone matches what you heard, release and confirm your answer.

Target frequencies are sampled uniformly in perceptual space between 120 Hz and 1000 Hz. There are five rounds per game and a maximum score of 50 points. Wear headphones, keep the volume at a comfortable but clearly audible level, and give yourself a quiet room — background noise is the enemy of precise pitch memory.

The science

The fleeting persistence of a sound in the mind is handled by echoic memory, the auditory branch of sensory memory. Echoic memory stores a high-fidelity trace of incoming sound for roughly two to four seconds — considerably longer than iconic (visual) memory, which fades in under a second. This extended window was documented through the suffix effect: an irrelevant spoken word appended to a list of items to be recalled disrupts memory for the final items, as if the suffix overwrites the echoic store (Darwin, Turvey, and Crowder, 1972). In Pitch, your task is to read out a specific attribute — frequency — from that decaying echoic trace before it is gone.

The physical structure of the inner ear explains why frequency distances are measured in ERB units rather than plain Hz. The basilar membrane inside the cochlea is tonotopically organized: high frequencies excite hair cells near the base, low frequencies near the apex. This spatial frequency map is not linear in Hz; the membrane allocates roughly equal physical length to each perceptually equal step in pitch. The Equivalent Rectangular Bandwidth (ERB) scale, derived by Glasberg and Moore (1990), models this cochlear frequency resolution by describing the effective bandwidth of each auditory filter. One ERB unit at 500 Hz spans about 80 Hz; one ERB unit at 4000 Hz spans roughly 500 Hz. Measuring your error in ERB units therefore reflects how large your mistake feels to the ear, not how large it looks on a frequency axis.

The game's error metric is the absolute difference between the ERB-rate of your guess and the ERB-rate of the target: |erb(guess) - erb(target)|, where erb(f) = 21.4 * log10(0.00437 * f + 1). An error of zero means a perfect match. At the half-score threshold — the point where you earn 5 out of 10 points for a round — the tolerance is 2.2 ERB units. To put that in concrete terms, 2.2 ERB units near 440 Hz corresponds to roughly a musical third, so the game is substantially more demanding than casual pitch awareness.

It is worth distinguishing absolute pitch from relative pitch. Absolute pitch (sometimes called perfect pitch) is the rare ability — present in roughly one person in ten thousand — to name or produce a musical note without any external reference. Relative pitch, by contrast, is the common skill of judging the interval between two notes. Pitch tests something close to absolute pitch: you must hold a frequency in memory with no reference tone to compare against. Even trained musicians with excellent relative pitch typically find this difficult, because the brain does not normally need to encode raw Hz values. The just-noticeable difference in frequency (the frequency difference limen) can be as small as a fraction of a Hz for trained listeners under ideal conditions, but memory for an absolute frequency over a delay of seconds is far coarser — most people's performance here reflects echoic retention, not fine-grained perceptual acuity.

Scoring explained

Each of the five rounds is worth up to 10 points, for a maximum total of 50. Your score for a round falls from 10 toward 0 as your ERB error grows. The scoring curve is calibrated so that a perfect match earns a full 10 and an error at the half-score threshold — 2.2 ERB units — earns 5. Errors larger than that threshold continue to score above zero but diminish quickly, rewarding even imperfect attempts over a complete miss.

Scores in the 40–50 range indicate exceptional absolute frequency memory, well above the typical untrained listener. A score of 25–35 is a respectable result for someone without musical training. Scores below 20 suggest that the echoic trace is fading or being disrupted before you can lock in — try eliminating distractions, using headphones, and responding more quickly after the tone ends.

Tips to improve

• Use headphones. Small frequency differences that are unmissable on headphones can become inaudible on laptop speakers, especially in the low-frequency range.
• Hum the tone silently to yourself immediately after it plays. Subvocalizing a pitch activates motor memory in addition to auditory memory, giving you a second encoding of the target frequency.
• Respond quickly. Echoic memory fades within a few seconds; the longer you wait before starting to drag, the more the trace degrades.
• Approach the target from below. Start with the slider near its minimum, sweep upward, and stop when the live tone begins to feel uncomfortably high compared to your memory. Then fine-tune downward. Many people find ascending sweeps easier to judge than descending ones.
• Pay special attention to the low bass range. The ERB scale compresses this region, so a small drag produces a relatively large perceptual shift — move slowly when the tone sounds very low.
• Between rounds, avoid listening to music or other pitched sounds. New tonal input can interfere with your residual memory of the previous target frequency.

FAQ