Phonology of modern Lhasa Tibetan

The following summarizes the sound system of the dialect of Tibetan spoken in Lhasa, which is the most influential variety of the spoken language

Vowels
Tournadre and Sangda Dorje describe eight vowels in the standard language:

Three additional vowels are sometimes described as significantly distinct: or, which is normally an allophone of ; , which is normally an allophone of ; and  (an unrounded, centralised, mid front vowel), which is normally an allophone of. These sounds normally occur in closed syllables; because Tibetan does not allow geminated consonants, there are cases where one syllable ends with the same sound as the one following it, with the result that the first is pronounced as an open syllable but retains the vowel typical of a closed syllable. For instance, zhabs (foot) is pronounced and pad (borrowing from Sanskrit padma, lotus) is pronounced, but the compound word, zhabs pad is pronounced. This process can result in minimal pairs involving sounds that are otherwise allophones.

Sources vary on whether the phone (resulting from  in a closed syllable) and the  phone (resulting from  through the i-mutation) are distinct or basically identical.

Phonemic vowel length exists in Lhasa Tibetan, but appears in a restricted set of circumstances. Assimilation of Classical Tibetan's suffixed vowels—normally ‘i (འི་)—at the end of a word produces a long vowel in Lhasa Tibetan; this feature is sometimes omitted in phonetic transcriptions. In normal spoken pronunciation, a lengthening of the vowel is also frequently substituted for the sounds and  when they occur at the end of a syllable.

The vowels, , , , and each have nasalized forms: , , , , and , respectively. Historically, this results from a syllable-final, such as , , etc. In some unusual cases, the vowels , , and may also be nasalised.

Tones
The Lhasa dialect is usually described as having two tones: high and low. However, in monosyllabic words, each tone can occur with two distinct contours. The high tone can be pronounced with either a flat or a falling contour, while the low tone can be pronounced with either a flat or rising-falling contour, the latter being a tone that rises to a medium level before falling again. It is normally safe to distinguish only between the two tones, because there are very few minimal pairs which differ only because of contour. The difference only occurs in certain words ending in the sounds [m] or [ŋ]; for instance, the word kham (, "piece") is pronounced with a high flat tone, while the word Khams (, "the Kham region") is pronounced  with a high falling tone.

In polysyllabic words, tone is only important in the first syllable.

Consonants
Notes:
 * 1) The unaspirated stops, , , and  typically become voiced in the low tone, being pronounced , , , and , respectively. These sounds are regarded as allophones. By a similar process, the aspirated stops , , , and  are typically lightly aspirated in the low tone.  The dialect of upper social strata in Lhasa does not use voiced stops in the low tone.
 * 2) The alveolar trill  is in complementary distribution of the alveolar approximant ; therefore, they are treated as one phoneme.
 * 3) The voiceless alveolar lateral approximant   resembles the voiceless alveolar lateral fricative  found in languages such as Welsh and Zulu and is sometimes transcribed $\langle\rangle$.
 * 4) The consonants, , , , , and  may appear in syllable-final positions. The Classical Tibetan final  is still present, but its modern pronunciation is normally realized as a nasalisation of the preceding vowel, rather than as a discrete consonant (see above). Note that  is not pronounced in the final position of a word, except in highly formal speech. Also, syllable-final  and  are often not clearly pronounced, but instead realized as a lengthening of the preceding vowel. The phonemic glottal stop  appears only at the end of words in place of an , , or  which were pronounced in Classical Tibetan but have since been elided. For instance, the word for Tibet itself was Bod in Classical Tibetan and is now pronounced  in the Lhasa dialect.