审阅范围 · Review scope
限定范围独立审阅。请保留正文中的出处与限制；不代表全库最终验收、实际演奏或学习效果验证。
Independent review within the stated scope. Keep the evidence and limitations stated in the text in view. This is not final corpus acceptance or validation of performance or learning outcomes.
本页保留作者送审冻结字节，因此正文或数据可能仍有“候选／等待审阅／未发布”标记。当前版本已通过限定独立复核并发布；旧标记只记录冻结时状态。通过范围是双语含义、限定来源文字／静帧及原创固定例题检查，不含原声听校、连续动作、完整源谱、实际演奏、身体适用性或学习效果验证。
This page preserves the author’s frozen review bytes, so prose or data may still contain candidate, pending-review or unpublished labels. This version has passed bounded independent review and repository publication; those labels record author-time status only. Acceptance covers bilingual meaning, bounded source text/stills and original fixed-fixture checks, not source audition, continuous motion, full source scores, actual performance, physical suitability or learning outcomes.
先问谁在移动：内声部、接近音与接近和弦
看到一个半音，先别急着给整段改和弦名称。圈出它所属的声部，再问：其他音保持吗？它要到哪一个目标、何时到达？移动一个音和移动整组和弦，可能用到相同琴键，却是不同的任务。
本页是一份完整纸面任务，不需要钢琴、唱音或播放；没有可选弹奏路线，也不分配手、指法或踏板。先做I与A，再读C的整组比较。三条声部只是同时记录的音线，不是三只手，也不自动对应三个谱表。
G是限定基础乐理，O是本库原创设计，T2是后文有定位的作者说明。所有练习音列与时间表均为原创。源课51讲整和弦接近；本页单音接近A是原创预备对照，不能说成原课只移动一个音。
把音名、声部与目标分开
G：一个半音是标准钢琴相邻两键的距离，黑键也算。♯升一个半音，♭降一个半音；自然音C–D–E–F–G–A–B–C的步长为2、2、1、2、2、2、1。中央C记C4，过B到C才换八度编号。88键从A0＝1起，C4＝40；本页D4＝42、E4＝44、F4＝45、F♯4＝46、A4＝49。
G：声部是一条可追踪的音线。内声部在当时更低和更高的音之间；移线也可能发生在最高音或低音。先追踪同一条线，别把“有一个音变化”误读成“整个和弦一起移”。
G／O：本页先指定目标集合D4–F♯4–A4，以D给和弦命名，这个基准叫根音；本例为D大三和弦，根上方距离0、4、7半音。从0计拍，目标窗口固定为偏移4–6，即第二小节前两拍。集合相同不保证相同和声功能；本页不根据声音、曲式或真实曲目推断功能。
G／O：I／A／C是版本代号；I不是这里的罗马数字一级和弦，C也不是这里的C和弦名称。I在两个保持的外声部之间改中间音。A先给一个邻近目标的单音，再进入目标集合。C先给整个目标集合的半音移位版，再让三音一起回到目标。“接近”描述相对目标的关系，不保证好听、合风格或可加入任意原谱。
G／O：C的前一组写C♯4–E♯4–G♯4，键号41、45、48，是本例D大三和弦整体低一个半音的C♯大三和弦；随后三音各升一键到42、46、49。按字母从C数起：C为1、D为2、E为3，所以三音保留E；本例E♯4与F4同为琴键45。格子和数据里的#与♯都表示升号，把对应自然音升高一个半音。A的单音写F4；同键不等于相同字母角色，也不让单音自动变成完整和弦。
原创对照：先圈出真正改变的部分
遮住“检查结果”，先追踪每版的下、中、上声部。I只圈中间音的改变；A圈目标前单独出现的音；C圈三条线各自的半音。再预测：三版在目标窗口是否都有D–F♯–A？是否都在窗口开始时重新起音？完整时间表在下一节。
| 版本 | 检查结果 |
|---|---|
| I | 两外声部保持0–6；内线改变并在3到达目标，跨4保持。 |
| A | 3–4只有F4；4–6才同时有目标三音。 |
| C | 3–4整个C♯大三和弦；4–6三音各升一键到D大三和弦。 |
G／O：I的中线F♯→F→E→F♯，键号46→45→44→46，步长−1、−1、＋2；不是每一步都是半音。保持D与A时，F自然音会形成D–F–A的不同竖向集合；E时是D–E–A。因此“以D作较长范围参照”不等于每一瞬间都还是D大三和弦。本练习只是受单线移动概念启发，不宣称所有这类线条都叫line cliché。
检查与修正
I固定两外声部，只改内线；A在偏移3–4只有F4，不能给它补上未写出的伴奏；C在3–4同时有C♯、E♯、G♯。A与C的中间琴键相同，但C另外有两音。三版在4–6都有目标集合，I已在3到达并保持，不在4重新起音；A与C在4同时起目标三音。
如果把I的外声部也移了，回到原表只恢复D4与A4的保持；如果把A说成整组接近，数3–4实际写出的音，只有一个；如果把C中的E♯改成F，键号虽对，指定大三和弦的字母结构已变。先改这个明确错误，再复核整条时间线。
目标窗口相同，起音时间可以不同
固定4/4、四分音符＝60 BPM，每小节四拍、每拍1秒；每版两小节，偏移0是第一小节开始，偏移4是第二小节开始。每格1拍，音名是新起音，=是保持同一音且不重起，.是该声部休止。先根据下表预测总秒数、每版起音事件数与有新音开始的不同时点数（同一时刻多个音只算一个时点），再看答案。
同一版本内三条声部同时读，I／A／C三个版本分开比较，不叠在一起。保持跨过小节线也不自动重起；所有声部都在偏移6释放并休止到8。这里的释放是符号终点，不是已测的琴声消失；没有踏板或音频实现。
I
下声部 [D4 = = =] [= = . .]
中声部 [F#4 F4 E4 F#4] [= = . .]
上声部 [A4 = = =] [= = . .]A
下声部 [. . . .] [D4 = . .]
中声部 [. . . F4] [F#4 = . .]
上声部 [. . . .] [A4 = . .]C
下声部 [. . . C#4] [D4 = . .]
中声部 [. . . E#4] [F#4 = . .]
上声部 [. . . G#4] [A4 = . .]核对时间、改写与重测
每版都是8拍＝8秒，不把三条同时声部相加成24秒。I有6个音符事件，集中在0、1、2、3四个时点；A有4个事件，C有6个事件，二者都集中在3、4两个时点。A的下／上声部0–4休止，中声部0–3休止；C三条线都先休止0–3。三个版本末尾各声部6–8均休止。
只把A改成C：将3–4的F4按本例和弦结构改拼E♯4（同键45），在下／上声部各加入C♯4／G♯4的一拍音，并把这两声部的前休止0–4缩为0–3。4–6目标与6–8休止完全不动。恢复A时删去新加的两音、恢复0–4休止并将中音写回F4；不能移动目标时间或把空出的拍删除。
稍后遮答案，重画A与C的3–4及4–6，再说明为什么I在4没有新起音；这是保留重测，不是陌生谱迁移证据。回到真实乐谱时，先核版本、调号、所有声部与原有节奏，标出目标及连接，才考虑是否需要另做这种比较；不要为了“丰富”而直接加音或移动原谱。
来源与边界
原创候选，等待独立复核，尚未发布。文字、选定静帧及原创计算的检查不等于原声、动作、完整课程、实际演奏、应用或学习效果验收。
B51-S01 · T2
第48课展示较长和声范围内移动一个声部，并对同一片段给出不同层次的标记。前例动最高声部，后例才有内线；我们保留“追踪谁在动”的观察，不把作者的整体Em读法当作每个瞬间音集合相同，也不接受普遍情绪或任意小调都适用的概括。
ASR14–17、19–22、28–39、56；75／87秒选定音位与变音，108／118秒两种标记，234／246秒选定内声部。23–27的唱名不确定，未用来定音。156／169秒只看长例版面，未逐音验收。
B51-S02 · T2
第51课的方法是目标和弦整体移高或移低半音后接回目标，不是只把一个音移半音。我们只采用这个操作区别及选定标记；复杂原谱的全部配置、节奏、衔接和实际效果没有验收。本页单音A及D大三和弦比较是原创设计。
ASR16–18、29–39、42–43、45–53、69–70；58秒方法文字，75秒基准标记，102秒上方半音、133／144秒下方半音标记。88秒模糊帧不作记号证据；165／187秒仅版面。
B51-S03 · T2
作者也提醒，反复套用可能显得刻意。这是他的风格建议；本包没有试听或听众试验来验证这种效果。真实作品还需核旋律、低音、声部连接、拍位与风格；本包没有完成这些作品级判断。
ASR55–63的作者意见；165秒三版本对照版面仅供定位，不证明听众效果。
已读两课全部84＋86＝170段既有ASR，检查16个静帧／16个不同图像哈希。57段保留为简短归属的候选支持；113段未用于归属，包含不确定唱名、未用叙述、推广、重复及未经核实的情绪／适用性概括。未用不等于确认错误，保留不等于逐句正确。没有新ASR、听校、连续动作、完整源谱、演奏、个体适用性或学习效果验证。
库存时长合计562.070秒；同哈希MP4容器合计562.026秒，差0.044秒原因未明。二者是记录口径，不是实际听看时长。
相关阅读
有限来源范围
原创固定例题检查
Ask what moves: an inner line, an approach tone or an approach chord
When you see a semitone, do not immediately rename the whole passage. Circle its voice, then ask: do the other tones stay? What is the target, and when is it reached? Moving one tone and shifting a whole chord can involve the same key but remain different tasks.
This is a complete paper task requiring no piano, singing or playback. There is no optional playing route or assignment of hands, fingering or pedal. Do I and A first, then C’s whole-group comparison. Three voices mean simultaneous notated lines, not three hands or automatically three staves.
G marks scoped basic theory, O our original design, and T2 the located author explanations below. All exercise sequences and schedules are original. Source lesson 51 teaches whole-chord approach; our single-tone A is an original prerequisite contrast, not evidence that the source shifts only one tone.
Separate pitch names, voices and targets
G: a semitone is the distance between adjacent standard-piano keys, including black keys. A sharp raises one semitone and a flat lowers one; natural C–D–E–F–G–A–B–C steps are 2, 2, 1, 2, 2, 2, 1. Middle C is C4, with octave numbers changing across B→C. The 88 keys start at A0=1 and C4=40; here D4=42, E4=44, F4=45, F♯4=46, A4=49.
G: a voice is a traceable line. An inner voice lies between lower and higher tones at that moment; a moving line can also be at the top or in the bass. Track one line before interpreting one changed tone as a shift of the whole chord.
G/O: we stipulate D4–F♯4–A4 as the target collection. D names the chord and is called its root; this is a D-major triad with root distances 0, 4, 7 semitones. Counting beats from zero, the target window is fixed at offsets 4–6, the first two beats of bar 2. Equal collections do not guarantee equal harmonic functions; we infer no function from sound, form or a real piece.
G/O: I/A/C are version IDs; I is not a Roman-numeral tonic here, and C is not a C-chord name. I changes a middle tone between two held outer voices. A presents one neighboring tone before the target collection. C presents the entire target collection shifted by one semitone, then returns all three tones to the target. Approach describes a relation to a target, not guaranteed appeal, stylistic fit or permission to add it to any score.
G/O: C’s preceding group is C♯4–E♯4–G♯4, keys 41, 45, 48: a C♯-major triad one semitone below this D-major collection. Each tone then rises one key to 42, 46, 49. Count letters from C: C is 1, D is 2 and E is 3, so the third keeps E; E♯4 shares piano key 45 with F4 here. In the grid and data, # is the same sharp sign as ♯, raising the letter’s natural pitch by one semitone. A’s single note is spelled F4. Sharing a key establishes neither the same letter role nor a complete chord from one tone.
Original comparison: circle what actually changes
Hide Check result and track each version’s lower, middle and upper voice. For I, circle middle-tone changes; for A, the isolated tone before the target; for C, each line’s semitone. Predict whether all versions contain D–F♯–A in the target window and whether all reattack at its start. The complete schedule follows.
| Version | Check result |
|---|---|
| I | Outer voices hold 0–6; the inner line changes and reaches the target at 3, holding across 4. |
| A | Only F4 at 3–4; all three target tones appear together at 4–6. |
| C | Whole C♯-major triad at 3–4; each tone rises one key to D major at 4–6. |
G/O: I’s middle line F♯→F→E→F♯ uses keys 46→45→44→46 and steps −1, −1, +2, so not every step is a semitone. With D and A held, F natural gives the different vertical collection D–F–A; E gives D–E–A. A longer-span D reference does not mean every instant is still a D-major triad. This exercise is inspired by single-line motion, not a claim that every such line is a line cliché.
Check and correct
I fixes the outer voices while changing the inner line. A has only F4 at 3–4; do not invent accompaniment. C simultaneously has C♯, E♯, G♯ at 3–4. A and C use the same middle key, but C has two additional tones. All contain the target at 4–6. I reaches it at 3 and holds without reattacking at 4; A and C attack all target tones at 4.
If you shifted I’s outer voices too, restore the specified held D4 and A4. If you call A a whole-group approach, count the tones actually written at 3–4: only one. If you respell C’s E♯ as F, the key remains right but the stipulated major-triad letter structure changes. Correct the specific error, then recheck the entire timeline.
Equal target windows can have different attack times
Fix 4/4 and quarter note=60 BPM: four beats per bar, one second per beat. Each version has two bars; offset 0 starts bar 1 and offset 4 starts bar 2. Each cell is one beat: a pitch marks a new attack, = continues the same note without reattack, and . is rest in that voice. Predict total seconds, note-event counts and the number of distinct times at which a new note begins (count simultaneous notes once) before reading the answers.
Read the three voices of one version simultaneously; compare I/A/C separately rather than overlaying them. Holding across a barline creates no automatic reattack. Every voice releases at offset 6 and rests to 8. Release here is a symbolic endpoint, not measured disappearance of piano sound; no pedal or audio implementation is supplied.
I
Lower [D4 = = =] [= = . .]
Middle [F#4 F4 E4 F#4] [= = . .]
Upper [A4 = = =] [= = . .]A
Lower [. . . .] [D4 = . .]
Middle [. . . F4] [F#4 = . .]
Upper [. . . .] [A4 = . .]C
Lower [. . . C#4] [D4 = . .]
Middle [. . . E#4] [F#4 = . .]
Upper [. . . G#4] [A4 = . .]Check time, edit and retest
Each version lasts 8 beats=8 seconds; do not add simultaneous voices to make 24 seconds. I has 6 note events at four attack time points 0, 1, 2, 3. A has 4 events and C has 6, both at time points 3 and 4. A’s lower/upper voices rest 0–4 and the middle voice rests from 0 to 3; all C voices rest 0–3. Every version ends with rest 6–8 in each voice.
To change only A into C, respell F4 at 3–4 as E♯4 for the stipulated chord structure, keeping key 45. Add one-beat C♯4/G♯4 in the lower/upper voices and shorten those initial rests from 0–4 to 0–3. Keep the target 4–6 and rests 6–8 unchanged. Restore A by removing those two added tones, restoring rests 0–4 and spelling the middle tone F4 again. Do not move the target or delete the vacated beat.
Later hide the answers, redraw A/C at 3–4 and 4–6, and explain I’s lack of reattack at 4. This is a retention retest, not evidence of transfer to an unfamiliar score. For a real score, first verify edition, key signature, all voices and original rhythm, marking the target and joins before deciding whether such a separate comparison is useful. Do not add tones or shift the score merely to enrich it.
Sources and limits
Original candidate awaiting independent review, unpublished. Text, selected stills and original arithmetic checks are not original-audio, movement, complete-course, actual-performance, application or learning-outcome acceptance.
B51-S01 · T2
Episode 48 shows one moving voice within a longer harmonic span and labels one passage at different analytical levels. The earlier example moves the top voice; a later one has an inner line. We retain tracking what moves, not unchanged instantaneous membership from the author’s broad Em reading or universal emotional/minor-key applicability.
ASR 14–17, 19–22, 28–39, 56; selected positions/accidentals at 75/87s, alternative labels at 108/118s, and selected inner-line placement at 234/246s. Uncertain solfege 23–27 is not used to establish pitch. At 156/169s only the longer-example layout was inspected, not every note.
B51-S02 · T2
Episode 51 shifts the whole target chord up or down a semitone before returning to it, rather than shifting only one tone. We retain that operational distinction and selected labels, not complete verification of the complex source voicings, rhythm, joins or effects. Our single-tone A and D-major comparison are original designs.
ASR 16–18, 29–39, 42–43, 45–53, 69–70; method text at 58s, baseline labels at 75s, upper-neighbor labels at 102s and lower-neighbor labels at 133/144s. The blurred 88s frame supplies no symbol evidence; 165/187s support layout only.
B51-S03 · T2
The author also cautions that repeated use can sound contrived. This is stylistic advice; this package includes no audition or audience experiment validating that effect. Real pieces still require melody, bass, voice-leading, metrical-placement and style checks; this package makes no such piece-level determination.
Author opinion in ASR 55–63; the three-version layout at 165s locates the discussion without proving an audience effect.
All 84+86=170 existing ASR segments were read, with 16 selected stills/16 distinct image hashes inspected. 57 segments are retained as candidate support for concise attribution; 113 are unused, including uncertain solfege, unused narrative, promotion, repetition and unvalidated emotional/applicability generalizations. Unused does not mean confirmed wrong, and retained does not mean every sentence correct. No new ASR, audition, continuous-motion, complete-source-score, performance, individual-suitability or learning-outcome validation occurred.
Inventory durations total 562.070 seconds; the same-hash MP4 containers total 562.026 seconds. The 0.044-second difference is unexplained. These are recorded duration measures, not actual watched/heard time.
Related reading
Register, spacing and chord identity
Retention retests and transfer
Bounded source scope
Original fixed-fixture checks