How We Test Digital Pianos and Piano Apps
Our published test protocol: what we measure on every digital piano and piano app, the tools we use, how scores are weighted, and how we choose our picks.
Every digital piano and piano app we review goes through the same written test protocol. We measure what we can with real tools, score the rest against fixed rubrics, and publish both the results and the method so you can check our work. This page explains each test, how the scores are weighted, how we pick the winners on our lists, and what we do when we get something wrong.
Current version: Pianoers Test Protocol v1.0, published November 2026. Every review states which version it was scored with.
Who does the testing
- Katarina: lead teacher. Piano teacher since 2001, B.A. in Music from the University of Missouri, member of the Music Teachers National Association (MTNA). She runs the playing and teaching-quality assessments and signs off every score.
- Richard J. Abraham: founder. Pianist since age 8, gigging player and teacher. He runs the measurements, the recordings and the app trials.
Where our products come from, and how we make money
- Products: every review states whether we bought the product ourselves or it was loaned to us. Loaned products are tested exactly the same way.
- How we make money: some links on Pianoers are affiliate links (Amazon and course providers). If you buy through one, we may earn a commission at no extra cost to you.
- What brands can't do: no brand can pay for a review, a higher score or a place on a list, and no brand sees a review before it's published.
Digital pianos: what we test
Every piano is tested in the same room, on the same stand and bench, with the same headphones and the same recording setup. Settings are factory defaults unless the review says otherwise.
1. Key action and weighting (30% of the score)
What we measure: how much force it takes to press a key (downweight, in grams) at C2, C4 and C6, and how much heavier the bass end is than the treble end.
How: we place calibrated weights on the front edge of the key, adding 2 g at a time, until the key goes all the way down. We repeat each key three times and record the average. Katarina also plays the same set of pieces on every model, including scales, repeated-note passages and a slow lyrical piece, and rates control, repetition and evenness with our action rubric.
Why it matters: key feel is the biggest reason beginners keep practising or give up. Learning on keys that are too light can make the move to an acoustic piano harder later.
Reported as: downweight at C2, C4 and C6 (g), graded weighting difference (g), and the action rubric score.
2. Piano sound (20%)
What we assess: how the main grand piano voice sounds through the built-in speakers and through headphones.
How: we record the same passages at soft, medium and loud dynamics through the speakers and through the headphone output. Katarina scores tone, dynamic range, sustain and how natural soft playing sounds, using our sound rubric. We publish the recordings so you can listen for yourself.
Reported as: sound rubric score and audio samples.
3. Speaker output (10%)
What we measure: maximum volume in dB(A) at 1 metre, and whether the sound distorts at full volume.
How: volume at maximum, the same loud chord sequence every time, measured with a sound level meter at 1 m, centred, at ear height when seated.
Reported as: maximum dB(A), distortion yes/no, and a short note on the bass response.
4. Key noise (5%)
What we measure: how loud the keys themselves are with the sound turned off. This matters for apartments and late-night practice.
How: volume at zero, the same passage played loudly, measured in dB(A) 50 cm above the keyboard in the same quiet room. We also record the room's background noise and report it.
Reported as: key noise dB(A) and room background dB(A).
5. Polyphony stress test (5%)
What we measure: whether notes cut out when many notes are sounding at once.
How: we hold the sustain pedal, play a low bass note, then play a fast run up the whole keyboard. On the recording we check whether, and after how many notes, the bass note gets cut off.
Reported as: pass/fail, plus the note count where the first note dropped, if any.
6. Latency (10%)
What we measure: the delay, in milliseconds, between pressing a key and the sound arriving over USB-MIDI, Bluetooth MIDI and Bluetooth audio, where the piano has them.
How: we record the key's mechanical sound and the piano's output on the same audio track, then measure the gap between them in audio editing software. Each measurement is taken 10 times and averaged.
Why it matters: above roughly 20 ms, many players start to feel lag, which makes app-based lessons and recording frustrating.
Reported as: average ms for each connection type.
7. Setup and learning-app compatibility (10%)
What we measure: minutes from opening the box to playing, and whether the piano works with Simply Piano, Flowkey and Skoove, over a cable and over Bluetooth.
Reported as: setup minutes and a compatibility table (works / works with adaptor / doesn't work).
8. Value (10%)
What we assess: the total score compared with the price, and with the other pianos in the same price band.
Reported as: the price and the date we checked it, plus the value score.
Specs we check but don't score
Polyphony, number of voices, weight, dimensions and connections. These come from the maker's spec sheet or owner's manual and are checked against the unit we tested. Where our measurement differs from the official figure, we report both.
Piano apps and online courses: what we test
Every app is tested on a phone, a tablet and a laptop, with a digital piano connected over USB-MIDI, and again using the device's microphone.
1. Note-detection accuracy (20%)
What we measure: whether the app notices when you play the wrong note.
How: we play the same 100-note exercise containing 10 deliberate mistakes (wrong notes, early notes and missed notes), three times over MIDI and three times through the microphone.
Reported as: % of mistakes caught, and false alarms (correct notes marked wrong), for MIDI and for the microphone.
2. Teaching quality (25%)
What we assess: whether the course would actually turn a beginner into a pianist: lesson order, technique and posture guidance, reading, rhythm and practice advice.
How: Katarina works through the first 10 lessons and scores them with our teaching rubric.
3. Curriculum coverage (15%)
What we assess: what the course covers. Each area gets 0 (absent), 1 (partial) or 2 (strong):
| Area | What we look for |
|---|---|
| Reading music | Treble and bass clef, ledger lines, sight-reading practice |
| Rhythm | Counting, metronome work, note values |
| Technique | Hand position, scales, finger independence |
| Theory | Intervals, keys, chord construction |
| Chords and playing by ear | Lead sheets, accompaniment patterns, improvising |
| Repertoire | Real pieces with a steady difficulty curve |
4. Feedback (10%)
What we measure: the share of lessons that give real-time or teacher feedback on your playing, rather than just video.
Reported as: % of the first 10 lessons with feedback, and the type (automatic, teacher video reply or live).
5. Time to first song (5%)
What we measure: how long it takes to play a first complete song with both hands, following only the app's own path, exactly as a beginner would.
Reported as: minutes of lessons before the first two-handed song.
6. Value (15%)
What we measure: the monthly price on the annual plan, the price per hour of lesson content, and what the free tier really includes.
Reported as: the price and the date we checked it, $/hour of content, and free-tier limits.
7. Cancellation and refunds (10%)
What we measure: how many steps it takes to cancel, whether the refund terms match what the sales page says, and how quickly support replies.
How: we cancel a real subscription and record every step, then ask support one standard question and time the reply.
Reported as: steps to cancel, refund window as tested, and support reply time.
How scores are calculated
- Each test produces a score from 0 to 10. Measured results are turned into scores using fixed bands that we set before testing; the bands for each test are shown in every review.
- The overall score is the weighted average of the test scores, using the percentages above, rounded to one decimal place.
- Pianos and apps are scored on separate scales, so a piano's score can't be compared with an app's.
- Once five or more products have been through a test, our reviews show each result next to the average of everything we've tested, so you can see at a glance whether a result is good or bad.
How we choose the picks on our best-of lists
- The product must have been through the full protocol.
- Picks are ranked by overall score within each category, such as "best under $500" or "best for small spaces".
- A lower-scoring product can win a category only if it leads that category's key test, for example the quietest keys for "best for apartments". When that happens, we say so on the list.
- Commission rates, and whether a brand has an affiliate programme at all, are never part of the decision.
Our rubrics
These rubrics make the parts of our testing that need a trained ear repeatable. Each criterion is scored from 0 to 10; the descriptions below show what earns a 2, a 6 and a 10.
Key action rubric
| Criterion | 2 | 6 | 10 |
|---|---|---|---|
| Control | Hard to play softly or evenly | Even at medium dynamics, less control at the extremes | Fine control from pianissimo to fortissimo |
| Repetition | Fast repeated notes miss or stick | Repeats reliably at moderate speed | Fast repeated notes and trills feel effortless |
| Evenness | Noticeable differences between neighbouring keys | Minor unevenness across the keyboard | Every key feels consistent |
| Realism | Feels like a light synth keyboard | Close to an acoustic upright | Feels like a well-regulated acoustic piano |
Sound rubric
| Criterion | 2 | 6 | 10 |
|---|---|---|---|
| Tone | Thin, artificial or harsh | Pleasant, slightly flat or generic | Rich, clear and characterful |
| Dynamic range | Soft and loud sound alike | Clear difference, but soft playing is weak | Wide, natural range from whisper to full |
| Sustain and decay | Notes cut off or loop audibly | Natural decay with small artefacts | Long, smooth, natural decay |
| Speakers vs headphones | Big drop in quality on one of them | Good on one, acceptable on the other | Excellent on both |
Teaching rubric
| Criterion | 2 | 6 | 10 |
|---|---|---|---|
| Lesson order | Random or jumps ahead | Logical, with occasional gaps | Each lesson builds clearly on the last |
| Technique and posture | Not covered | Mentioned but rarely reinforced | Taught early and reinforced throughout |
| Reading and rhythm | Mostly skipped | Covered at a basic level | Built in steadily from the start |
| Practice guidance | None | General tips | Clear practice plans with goals |
| Motivation | Easy to stall or quit | Engaging at first, fades later | Keeps a beginner progressing for months |
Updates and retesting
- Prices are checked whenever a review is updated, and the check date is shown on the page.
- Retests happen when a piano gets a hardware revision or major firmware update, or when an app changes its curriculum, pricing or note-detection engine.
- Older versions: when we change this protocol, we publish the new version, keep the old one online, and note in each review which version it was scored with.
Corrections
If you think we got something wrong, please tell us through our contact page. When we correct a factual error, we fix the review, add a dated correction note at the top of it, and list it below.
| Date | Page | What changed |
|---|---|---|
| — | — | No corrections yet. |
Changelog
| Version | Date | Changes |
|---|---|---|
| v1.0 | November 2026 | First published protocol. |
Coming in v1.1: a 30-day app trial with adult learners, and long-term notes on how pianos hold up after a year of daily practice.
Frequently asked questions
Do you test every piano yourselves? Yes. A piano only gets a score after it has been through this protocol. Products we haven't tested are marked "not tested" wherever they appear.
Why measure key weight in grams? Because "realistic weighted keys" means something different on every product page. Grams at three points on the keyboard let you compare actions directly.
How accurate are your volume measurements? We use the same meter, distance and room for every test, so the numbers are reliable for comparing pianos with each other, which is what they're for.
Do brands ever see reviews early? No. Brands see a review at the same time you do.