Vowel space (F1 × F2)
The x axis is F2 (lower = further back, so the right side is back), the y axis is F1 (lower = higher tongue). One dot per vowel — female voice only (male clips are too short to track reliably; see the method section).
Word list
Click a word to hear it whole.
or use the + on each cell
In a sentence
How the word sounds in a sentence. Tap the line to hear it all, tap the highlighted word to hear just that part.
Measured values
Median over the middle 50% of the voiced span.
| Kana | Romaji | Speaker | Duration | F1 | F2 | F0 |
|---|---|---|---|---|---|---|
| あ | a | female | 180 ms | 940 | 1570 | 246 |
| い | i | female | 360 ms | 342 | 2316 | 229 |
| う | u | female | 150 ms | 378 | 1721 | 190 |
| え | e | female | 190 ms | 595 | 1876 | 298 |
| お | o | female | 170 ms | 478 | 957 | 207 |
How to practise
- Listen to あ → い → う: the lips stay flat for う, not rounded.
- Say English 'oo', then Japanese う — feel the lips unround.
- Alternate あ and お: Japanese お is less rounded than English 'oh'.
Where these numbers come from
Synthesised (ja-JP-NanamiNeural / ja-JP-KeitaNeural), peak-normalised, speech span cut at -28 dB below the file's loudest frame, then the median F1/F2 over the middle 50% of the voiced span via parselmouth. Computed by 53_ja_tts_measure.py — never typed in.
Two voices are not a survey of Japanese. Formants depend on vocal tract length, so compare relative positions, not absolute values.