What is the National Library at Kolkata scheme?

The National Library at Kolkata (formerly Calcutta) romanisation is the diacritic-based transliteration standard used in academic and library cataloguing for Indic scripts. It belongs to the same family as ISO 15919 and IAST, using marks like ā, ī, ṛ, ś, ṣ, ṅ, and ṇ to distinguish sounds.

How is the inherent vowel handled?

Bengali is an abugida where a bare consonant carries an inherent vowel a. The tool adds a after a consonant unless a vowel sign (kar) replaces it or a hasanta (্) suppresses it, matching how the script is actually read.

How are conjuncts transliterated?

When a hasanta joins two consonants, the inherent vowel of the first is dropped and the consonants are romanised back to back. So ক্ষ becomes kṣ and স্ক becomes sk, reflecting the cluster pronunciation.

Is transliteration reversible?

Not perfectly. Several Bengali distinctions collapse in pronunciation (for example শ, ষ, and স), and the scheme preserves spelling distinctions with diacritics but does not capture every contextual sound change. Treat the output as a faithful script romanisation, not a phonetic transcription.

Does it work on mixed text?

Yes. Latin letters, digits, spaces, and punctuation pass through unchanged, so you can romanise Bengali embedded in otherwise English text without breaking the rest.

Bengali to Latin Transliteration

Email me this result

Get this tool's output sent to your inbox, plus one useful tool a week. No spam, unsubscribe any time.

Romanising Bengali means representing each letter of the script with Latin characters and diacritics. This tool uses the National Library at Kolkata (Calcutta) scheme — the diacritic standard used in academic and library cataloguing — handling the inherent vowel, conjuncts, and vowel signs correctly.

How it works

Bengali is an abugida, so the algorithm tracks the vowel that follows each consonant:

A consonant is romanised to its base Latin form (ক → k, শ → ś).
If a dependent vowel sign (kar) follows, that vowel is emitted (কা → kā).
If a hasanta (্) follows, the inherent vowel is suppressed and the next consonant joins directly, giving a conjunct (ক্ষ → kṣ).
Otherwise the consonant carries its inherent a (ক → ka).

Independent vowels (অ → a, ঈ → ī) and signs like anusvara (ং → ṃ) and visarga (ঃ → ḥ) map directly. Latin text and punctuation pass through.

Example

নমস্কার romanises as namaskāra: ন → na, ম → ma, then স্ক (with the hasanta) → sk, and ার → āra. The inherent vowels and the conjunct are all resolved automatically.

Notes

The output is a faithful script romanisation, preserving spelling distinctions (শ/ষ/স as ś/ṣ/s) with diacritics. It is not a phonetic transcription, so it does not model every contextual sound change in spoken Bengali.
Use it for citations, transliterated bibliographies, and cataloguing where the NLC/ISO 15919 family is expected.