midicnTheory 中文

Start with this entry's core, a counter-intuitive fact:

While playing, you hear your intention, not the sound you actually make.

The brain fills in the actual sound with "how I mean it to go" — so you feel the playing matched your imagination, while the recording says otherwise. This is not self-deception but an inevitable coupling of hearing and action.

Which makes a recording the most objective teacher: it takes the actual sound out of your hands and turns it into something you can hear repeatedly, without the pressure of real-time action.

A four-step procedure

StepDo this
1 recordrecord a complete run (not fragments — only the whole reveals structure and stamina)
2 let it restlisten later (hours, or the next day)
3 listen objectivelyuse the four layers below; do not think "I know that bit" as you go
4 write down specificsnotes that are actionable

Step 2 cannot be skipped: listening straight after playing, you hear through the memory of the actions just made, and attention lands on "where I struggled" rather than "how it sounds". After a gap, your ears become an audience's ears again.

Four layers to listen for

LayerWhat to hearTypical discovery
1 rhythmis it steady throughout (especially in hard passages and joins)?"I thought it was steady, but I was secretly speeding up in the hard places"
2 dynamicsdo the levels match the intention? (see [[concept:dynamics\dynamic markings]])
3 phrasingis the punctuation clear? (see [[concept:breath-phrasing\breathing and phrasing]])
4 timbreis it even, without superfluous impact? (see [[concept:articulation\articulation]])

Layer 1 delivers the biggest surprises: because the ear barely notices gradual acceleration (see metronome and BPM) — a recording is the only objective check.

Notes must be actionable

This is the entry's most practical rule:

Useless noteUseful note
"it was bad""the crescendo in bars 17–24 starts too late"
"rhythm was unsteady""I rushed half a beat at the join in bar 33"
"the sound was not ideal""bar 51: pressing too hard in the upper register, the sound turns hard"

The difference is actionability: a useful note points directly at the next practice session (the same requirement as "say exactly which beats" under how to practise).

A common mistake

Listening only for the wrong notes.

Only hunting errorsConsequence
you find where the technique failedbut miss how the whole thing sounds
—and a recording's greatest value is precisely the whole: phrasing, shape, direction

So listen once as an audience (what is the overall impression? do I want to keep listening?), and only on the second pass as an inspector. Reverse the order and you will hear nothing but the errors.

Diagram: the gap between intention and reality

While playing you hear intention; the recording holds the actual sound — hence listen later while playing listening back you hear the intention the actual sound is filled in you hear the actual sound the most objective teacher, free of real-time pressure Four layers: rhythm steadiness, dynamic levels, phrasing, evenness of timbre Notes must be actionable: "bad" is useless; "the crescendo starts too late in bars 17-24" is not Common mistake: listening only for errors — that misses the recording's greatest value, the whole

Listen: intention versus actual

This site cannot play your recording — but rhythm can run a small "intention versus actual" exercise: count four beats inwardly first, then press play. Most people find their inner four beats and the actual four are not the same length — the smallest example in the time dimension of the gap between intention and reality.

Common misconceptions

Listen in the library

Suite bergamasque: III. Clair de luneC14:36Debussy's Clair de lune is delicately layered — only a recording reveals whether the dynamics and phrasing matched the intentionDetailsDanse Macabre, S.555C110:20Liszt's transcription of Danse macabre is full of fast passages, where a recording most easily exposes unsteady rhythmDetailsOde to JoyC10:37The Ode of Joy is simple, a good starting point for learning to hear yourself before moving to large worksDetails

All tracks from midicn-lib — inline badge is the license tier (C1 commercial / C2 non-commercial / C3 study only).

Sources

  • 演奏时的听觉反馈以"预期"为主,与录音回放存在系统性偏差,为通行演奏教学与音乐心理学表述
  • 录音自查的常见维度(节奏 / 力度 / 分句 / 音色)与"冷处理后回听"的方法,为通行表述

Used for fact-checking only; all prose is original.