Ear training / testing
Perfect pitch test: how to take one that means something
Most online perfect pitch tests are too short, too predictable, and too easy to game with a remembered reference. Here is what a valid test controls for, and how to read your result honestly.
A perfect pitch test asks you to name pitches with no reference given. That sounds simple, but a test that does not control for a handful of specific confounds will hand out false positives freely, which is why so many people come away convinced they have absolute pitch when they have something else.
What a valid test controls for
- No reference note, ever. Not at the start, not between items. One reference at the beginning turns the whole thing into a relative pitch test.
- A silent washout period. No music for at least ten minutes beforehand. Hearing a song in a known key primes you.
- Enough items. Twelve items gives you a real chance of scoring well by luck. Thirty or more is the minimum for a usable signal.
- Randomised octaves and a wide range. Many partial possessors are accurate only in the middle two octaves.
- Multiple timbres. Piano only will flatter anyone who has played piano for years. Include sine tones, strings, and voice.
- Black notes as often as white. Tests weighted toward white notes overstate accuracy considerably.
- No humming, no singing, no instrument. Any of these converts the task into a relative one against your own produced reference.
Reading your score
| Result | Most likely interpretation |
|---|---|
| Near-perfect, instant, across timbres and octaves | Genuine absolute pitch. |
| Strong on piano, weak elsewhere | Timbre-specific pitch memory from long familiarity. |
| Consistently one or two semitones off in the same direction | A stable internal reference that is slightly mistuned. Common and useful. |
| Good only after hearing the first note | Relative pitch. This is the valuable one. |
| Around 1 in 12 correct | Chance. No pitch labelling. |
That third row deserves a note. A consistent offset is not a failure. It means you are carrying a genuine absolute reference and merely have the label attached one step off. That is a real and trainable form of pitch memory.
Test construction matters more than test length alone: Athos and colleagues found that performance distributions in large-sample absolute pitch testing are strongly bimodal only when octave, timbre, and note-class are properly randomised. Poorly randomised instruments produce a smeared unimodal distribution that makes partial and genuine absolute pitch indistinguishable.
Athos et al., PNAS, 104(37), 2007.
What to do with the result
If you scored at chance, nothing is wrong with your ears, and this has no bearing on whether you can become a good musician. See how rare absolute pitch is for why that is not the constraint people assume.
Whatever you scored, the productive next step is the same: train intervals and scale degrees. Those are the skills that let you transcribe, improvise, and harmonise, and unlike absolute pitch they respond to practice at any age.
Take the test
Our ear trainer includes a pitch identification mode built to the constraints above: randomised octave, mixed timbre, black and white notes weighted equally, and no reference tone. It runs in the browser with no signup.
Some links in this article are affiliate links to Musical U, Pianote, ArtistWorks, Yousician, or Amazon Associates. If you subscribe or buy after clicking, we may earn a small commission at no additional cost to you. Affiliate revenue helps us keep the interactive tools and the essays free. It does not change what we recommend. Full policy at /affiliate-disclosure/.