The four sets
Click a cell: A = written form, B = actual reading.
Minimal pairs in sentences
Two sentences side by side, differing by one sound in the same slot. Tap a card to hear the whole line, tap the highlighted part to hear just that.
Homophones and correlation
Near 1 = the waveforms nearly coincide; clearly lower = different sound. The へ row is the control: read alone it is he, so it must be low — which shows the measure can tell same from different.
| Written | Read as | Dur A | Dur B | Correlation |
|---|---|---|---|---|
| を | お | 150 ms | 150 ms | 1.00 |
| へ | え | 190 ms | 170 ms | 0.02 |
| ぢ | じ | 220 ms | 220 ms | 1.00 |
| づ | ず | 260 ms | 260 ms | 1.00 |
How to practise
- を only appears as a particle and is always o, never wo.
- へ as a particle is e (学校へ行く); inside a word it is he (部屋 heya).
- ぢ / づ survive mostly in rendaku and nasalised contexts (はなぢ).
Where these numbers come from
Both kana synthesised in one voice, spans aligned at onset, zero-lag Pearson correlation. A correlation near 1.00 could also mean the same file twice, so the page carries a control pair that should differ (を vs え).