Standard Japanese has five vocalic phonemes /a i u e o/.

The phoneme /a/ is an open prevelar or velar vowel, between [a] and [A] and very different for each person. We transcribe [a].

The phoneme /i/ is the cardinal [i] : unroundend high palatal vowel.

The phoneme /u/ is an unrounded high velar [µ]. It is a little less back that [u] (but this is common for unrounded vowels). It as an even more front allophone [µ+] after [s z ts dz] . We always transcribe [µ].

The phoneme /e/ is a unrounded palatal vowel between [e] and [E]. We transcribe [E].

The phoneme /o/ is a rounded velar vowel, between [o] ed []. We transcribe [O]. Notice that it is the only rounded japanese vowel.

Table of vowels: graphemes/phonemes/phones

phoneme	example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
a	方	kata	kata	/kata/	[kata]
i	木	ki	ki	/ki/	[ki]
u	上	ue	ue	/ue/	[µE]
e	絵	e	e	/e/	[E]
o	男	otoko	otoko	/otoko/	[OtOkO]

Short vowels in the end of a word (before pause), especially when emphatic, may be followed by a glottal stop, written in Japanese {っ} or, in western characters (ローマ字), with the sign of a short vowel.

Es. (from Vance):

やまとがきえたぞっ
Yamato ga kieta zo(!
[jamatO Na kiEta dzO/]

Long vowels

Long and short vowels has in japanese the same timbre.

long vowel	example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
/aa/ > [aù]	お母さん	okaasan	okaasan	/okaasaN/	[Okaùsaµ)]
/ii/ > [iù]	お兄さん	oniisan	oniisan	/oniisaN/	[OnJiùsaµ)]
/uu/ > [µù]	夫婦	fuufu	huuhu	/huuhu/	[¸µù¸µ8]
/ee/ > [Eù]	お姉さん	oneesan	oneesan	/oneesaN/	[OnEsaùµ)]
/oo/ > [Où]	お父さん	otoosan	otoosan	/otoosaN/	[OtOùsaµ)]

Within the same morpheme /e i / gave [Eù] at the end of XVIII century.

Nowadays:

{えい} : /e i / > [Eù], ex.g. 景気 /keeki/ > [kEùki] (the times)

{えい} : /e i / > [Ei] when pronounced slowly and 'with accuracy', as to teach a new word to children.

{ええ} o {エー}: /ee/ > [Eù] always:

es.

ケーキ: /keeki/ > /keeki/ > [kEùki] (cake)

お姉さん : /OneesaN/ > /oneesaN/ > [OnEsaùµ)]

Within the same morpheme

{おお}: /oo/ > [Où]

e {おう}: /oo/ > [Où]

Between morhemes both {えい} and {おう} are pronounced with two separates phonemes even in rapid speech.

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
毛色	/keiro/	/keiro/	/keiro/	[kEiRO]
子牛	/koushi/	/kousi/	/kousi/	[kOµþi]
争う	/arasou/	/arasou/	/arasou/	[aRasOµ]

The same, even within the same morpheme, in modern foreign loans.

Ex.g.

スペインン /supeiN/ > [sµ8pEiµ)]

ソウル (= Seoul) /souru/ > [sOµRµ]

Voiceless vowels

/i/ and /u/ has voiceless allophones [i8] and [µ8], when they are between two voiceless consonants, mostly in unstressed syllable. /u/ is usally voiceless after a voiceless consonant even before pause.

In rapid speech they may completely disappear (and other vowels between voiceless consonants may became voiceless).

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
口	kuchi	kuti	/kuti/	[kµ8tþi]
文学	bungaku	bungaku	/buNgaku/	[buNùNakµ]

Table of the Japanese consonants:

The consonants

OCCLUSIVES

Japanese occlusives are not aspirated.

/p b/ are bilabial occlusive, voiceless and voiced [p b].

The phoneme /b/ is sometimes bilabial approximant [B] when between vowels.

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
勉強	/benkyò/	/benkyoo/	/beNkyoo/	[bENùkjOù]
日本	/nippon/	/nippon/	/nippoN/	[nJiùppOµ)]

/t d/ are dental occlusive, voiceless and voiced [t d].

/t/ has two important allophones more: dental affricate [ts] before /u/ and prepalatal affricate [tþ] before /i, y/.

/d/ before /i, y, u/ is pronounced like the phoneme /z/: [dz, z, dü] (see the phoneme /z/).

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
方	/kata/	/kata/	/kata/	[kata]
どなた	/donata/	/donata/	/donata/	[dOnata]
待つ	/matsu/	/matu/	/matu/	[matsµ8]
待ちます	/machimasu/	/matimasu/	/matimasu/	[matþimas]

/k g/ are velar occlusive, voiceless e voiced; before a /i, y/ they ahve prevelar allophones. We always transcribe [k g].

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
聞く	/kiku/	/kiku/	/kiku/	[ki8kµ]
元気	/genki/	/genki/	/geNki/	[gENùki]

The phoneme /g/ is nasalized to [N] at the beginning of word. Instead, young people use often the not nasalized velar fricative [Ä].

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
右	/migi/	/migi/	/migi/	[miNi] o [miÄi]

The pronunciation of di /g/ with [N] presents some problems and it is not clear whether [N] is to be considered an allophone of /g/ or a real phoneme /N/. Actually there are some cases in which /g/ > [g] even not at the beginning of a word (from Vance):

1 - in compounds /g/ is often [g] especially if it is at the beginning of a second element which exists as an indipendent word;

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
日本銀行	/nipponginkò/	/nipponginkoo/	/nippoNgiNkoo/	[nJiùppOµ)giNùkùO]

2 - /g/ > [g] after the honorific prefixes /o/ e /go;

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
お元気	/o-genki/	/o-genki/	/ogeNki/	[OgENùki]

3 - the morpheme /go/ (= "five") is always [gO], and we may have opposition between [g] and [N] with a phonologization of /N/;

meaning	example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
fifteen	十五	/jþgo/	/zyuugo/	/zyuugo/	[düµùgO]
rear services	銃後	/jþgo/	/zyuugo/	/zyuuNo/	[düµùNO]

4 - in omomatopeici doubled adverbs.

example	Hepburn transcription	Kunrei-shiki transcription	Phonemic transcription	Phonetic transcription
がらがら	/garagara/	/garagara/	/garagara/	[gaRagaRa]