Igbo-TTS-Dataset

Language

Igbo (also rendered Ibo; native name: Asụsụ Igbo or Ụsụ Igbo) is a Niger-Congo language of the Volta-Niger branch, spoken primarily in southeastern Nigeria. It is one of the three major official languages of Nigeria alongside Yoruba and Hausa, and is spoken as a first language by approximately 25 to 30 million people. Igbo speakers are concentrated in the states of Abia, Anambra, Ebonyi, Enugu, and Imo, with significant communities also found in Delta, Rivers, and Bayelsa states. There are also substantial Igbo diaspora communities in the United Kingdom, United States, and other parts of the world.

Igbo is one of the most linguistically studied languages of sub-Saharan Africa and has a substantial literary tradition in both print and oral form. The language plays a central role in the cultural identity of the Igbo people, one of the largest ethnic groups in Africa. It is widely used in education, broadcasting, religious practice, and popular culture across southeastern Nigeria.

Igbo is a tonal language with two underlying phonemic tones — high and low — which interact to produce a downstep pattern and a range of phonologically complex surface tone sequences. Tone in Igbo is morphologically and grammatically significant, distinguishing lexical items as well as grammatical categories such as tense, aspect, and focus.

Variants

Igbo is a dialect continuum spanning a wide geographic area. While the dialects differ considerably in phonology, vocabulary, and some grammatical features, a standard variety — Igbo Izugbe (Standard Igbo) — was developed in the 1970s through a coordinated academic and governmental effort and is used in formal writing, broadcasting, and national education. Major regional dialect groups include:

Central Igbo dialects:

• Owerri Igbo — spoken in Imo State; phonologically and lexically central; forms a key reference variety for Standard Igbo
• Umuahia Igbo — spoken in Abia State; closely related to Owerri; a prestige spoken variety in modern Igbo

Northern Igbo dialects:

• Onitsha Igbo — spoken in Anambra State; historically influential due to the role of Onitsha as a major publishing and commercial centre; considered conservative in some morphological features
• Awka Igbo — spoken in Anambra State; phonologically close to central varieties; widely represented in academic and religious literature

Southeastern Igbo dialects:

• Ngwa Igbo — spoken in Abia State; notable for distinct phonological features and lexical items
• Mbaise Igbo — spoken in Imo State; transitional between central and southern varieties

Riverine and peripheral varieties:

• Ikwerre — spoken in Rivers State; transitional and sometimes classified separately
• Ika and Ukwuani — spoken in Delta State; Igboid varieties that differ substantially from central Igbo

The variety represented in this dataset reflects the standard spoken variety of Igbo as used in contemporary media, education, and public discourse in southeastern Nigeria. The content draws on the common register widely intelligible across Igbo-speaking communities.

Writing System

Standard Igbo uses a Latin-based alphabet established through the Onwụ Orthography, which was formalized by a committee of scholars and adopted in 1961. It is the basis of the official orthography used in school curricula, government publications, and print media. The Onwụ orthography uses an augmented Latin alphabet of 36 letters, incorporating eight vowels (a, e, i, ị, o, ọ, u, ụ), consonants, and the digraphs ch, gb, gh, gw, kp, kw, nw, ny, sh:

a, b, ch, d, e, f, g, gb, gh, gw, h, i, ị, j, k, kp, kw, l, m, n, nw, ny, o, ọ, p, r, s, sh, t, u, ụ, v, w, y, z

Key features of the Onwụ orthography include:

• Sub-dotted vowels (ị, ọ, ụ) to represent distinct phonemes: ị for the high front unrounded vowel /ɪ/, ọ for the open-mid back rounded vowel /ɔ/, and ụ for the near-close near-back rounded vowel /ʊ/
• Tone marks (acute accent for high tone, grave accent for low tone, macron for downstep) are used in linguistic and educational materials but are frequently omitted in everyday writing, journalism, and informal texts
• The digraph ny for the palatal nasal /ɲ/, and nw for the labial-velar nasal /ŋʷ/

An earlier missionary orthography, associated with the translation of the Bible into Igbo (completed in 1913) and subsequent colonial-era publications, was in use before the Onwụ standardization. This older orthography used simpler Latin letters without sub-dotted characters and represented the sounds of Igbo imprecisely.

Orthographic representation in this dataset: The transcriptions in this dataset are written primarily in a convention that employs the standard Igbo dot-below characters (ọ, ị, ụ) consistent with the Onwụ orthography. A subset of the recordings — particularly those drawn from older source texts — uses an alternative notation in which the diaeresis character is substituted for the dot-below (e.g., ö in place of ọ, ü in place of ụ). This alternative notation reflects the typographic conventions of some earlier Igbo print materials in which the dot-below characters were unavailable or impractical. Both notations are present in the dataset, and users working with the transcriptions should be aware of this variation.

Grammar and Linguistic Features

Igbo is an isolating language with an SVO (Subject-Verb-Object) basic word order. Key grammatical features include:

Tonal system:

• Two underlying tones: high (H) and low (L), interacting to produce phonological processes including downstep (a lowered high tone following a low tone), tone spreading, and tone deletion in specific morphological environments
• Tone is a primary grammatical marker in Igbo, encoding tense, aspect, focus, and negation in conjunction with segmental morphology

Tense-Aspect system:

• Igbo marks tense and aspect through a combination of tonal alternations and preverbal particles rather than inflectional verb endings
• "na-" — marks progressive/habitual aspect (e.g., "Ọ na-eri nri" = "He/she is eating")
• "rị/rụ/la" (suffix variants) — marks perfective aspect
• "ga-" — marks future tense (e.g., "Ọ ga-aga" = "He/she will go")

Pronominal system:

• Igbo pronouns encode number but not grammatical gender: "ọ" (he/she/it), "ha" (they), "anyị" (we), "ụnụ" (you plural), "m/a" (I)
• A distinction is made between inclusive and exclusive first-person plural in some varieties

Noun class and concord:

• Igbo lacks the noun class system typical of Bantu languages but uses a system of number marking through tone change and some suppletive forms
• Definiteness is expressed through word order and context rather than articles

Serial verb constructions:

• Serial verbs are common, as in other Volta-Niger languages, allowing sequences of verbs sharing a single subject without overt connectives

Negation:

• "a-" or "e-" verbal prefix in combination with tonal modification marks negation in many tense-aspect constructions
• "ọ bụghị" and "ọ naghị" function as predicative negative copulas

Source

The textual material in this dataset originates from a variety of Igbo written and transcribed sources. These include:

• Oral tradition narratives and cultural commentary, including accounts of Igbo musical heritage (ekpili music tradition), the role of griots in West African oral culture, and the cultural significance of the kola nut (ọjị) in Igbo society
• Folklore and didactic narrative (akụkọ ifo), including the well-known cycle of tales featuring Mbe the Tortoise
• News reporting and current affairs material drawn from Igbo-language journalism, reflecting contemporary Nigerian public discourse on topics such as healthcare, government policy, road safety, and social affairs
• Biographical and social commentary content

The texts were segmented into short utterances suitable for read speech and used as prompts for audio recording sessions.

Domain

This dataset is derived from prompted read speech. The speaker read aloud pre-written Igbo texts drawn from oral tradition, narrative, cultural commentary, and contemporary news sources. The content covers a range of registers and everyday topics typical of modern spoken Igbo, including civic commentary, social observation, cultural heritage, folklore, and general public discourse.

The dataset has been structured as segmented, read-style speech suitable for speech synthesis tasks.

Size

The dataset is composed of 17 folders containing audio clips and corresponding mapping files.

Each folder contains between 39 and 175 audio files. Individual audio clips typically range from 2 to 45 seconds in duration.

Folder-level durations range from approximately 5 minutes to just under 40 minutes of audio.

The dataset represents a total of 1,335 audio files with a combined duration of approximately 4 hours 38 minutes and 2 seconds of segmented Igbo speech.

A detailed breakdown of durations and file counts per folder is provided below.

Structure

Each folder in the dataset contains:

• A collection of audio files in MP3 format
• A tab-separated mapping file linking each audio file to its transcription

Each line in the mapping file follows the format:

audio_filename.mp3 key sentence attempts

The dataset is designed for TTS pipelines requiring paired audio-text data.

Sample

1. dc957b154bda2c43b239c57b9858219d.mp3 | Ekenekwaa m unu ndi nwem. Eji m aka na ihe n'agarai unu nke öma.
2. df8edbf7ea8138b39f2f80104984d8af.mp3 | Egwu ekpili so na otu egwu kara ezi nka n'ala Igbo.
3. f2fa99265321f8612b293a097854580e.mp3 | Ọjị dị nnukwu mkpa n'ọnọdụ onye Igbo.
4. d50767247ed690b98f0c96ef39998600.mp3 | Onye wetara Oji wetara ndu.
5. 958931cb6a336a138a95dd6ee632a33b.mp3 | Ndị dọkịta Naịjirịa akagbuola abụbọọrụ ha gbara.
6. c8e2ae9c98931136735c77490d1d899c.mp3 | Agụụ mee gị ihe o mere Mbe, i gaghị ata ya ụta n'ije aghụghọ maọbụ arụrụala ọbụla o jere.
7. efb1b1476ff7ef1512abda9afd9d1c1b.mp3 | Ihe mere m ji lụnye Aboy nwunye - Eze Chibuzor Gift Chinyere.
8. 19f560e5e3a5a0cf7dbecf09842c7522.mp3 | Akụkọ na-eru anyị ntị ugbua bụ na ọkụ amabeghị ihe butere ya enwurula n'ụlọụka Notre Dame Cathedral dị na Paris, mba France.