How to Type Dakuten and Handakuten

With Romaji input, type the voiced sound directly. Use g for the k-row with dakuten, z for the s-row, d for the t-row, b for the h-row, and p for the h-row with handakuten. For example, ga → が, ba → ば, and pa → ぱ.

What the marks do

Dakuten is the two-stroke mark seen in が. It changes selected unvoiced consonants into voiced sounds. Handakuten is the small circle seen in ぱ and creates the p-row.

Handakuten changes the h-row to ぱ ぴ ぷ ぺ ぽ, typed pa pi pu pe po.

Type the sound, not the punctuation mark

In ordinary Romaji typing, enter go for ご rather than typing こ and adding a mark. The IME produces the precomposed Kana naturally.

Direct Kana layouts may provide a dakuten/handakuten modifier key. A virtual keyboard can instead present every voiced character as its own button, which is clearer for occasional use.

Special readings in the d-row

The characters ぢ and づ exist, but their modern pronunciation overlaps with じ and ず in standard Japanese. IMEs may use di and du to distinguish them, while ji and zu normally create じ and ず.

Use the spelling required by the actual Japanese word. Do not choose ぢ merely because its sound resembles ji.

Katakana and foreign sounds

The same principles create ガ, バ, and パ. Katakana also uses dakuten with ウ to form ヴ, which appears in some foreign-name transliterations. Conventional spelling still varies by word.

Unicode detail

Japanese text can represent a voiced Kana as a precomposed character or, in some technical contexts, as a base plus combining mark. Normal input methods produce the expected text form. For everyday typing, use the complete Kana output and avoid manually assembling combining marks unless a technical task requires them.

Test every voiced row directly in the Hiragana keyboard or Katakana keyboard.