Spelling test audio: make your word list read aloud
A spelling test needs someone to read every word clearly, in the same way, at the same pace. Children practising at home, adults learning English and teachers preparing tests for a class all run into the same problem: a new list every week and nobody free to read it.
This guide shows how to turn a word list into spelling test audio with the free SuperClonica engine: the word, a sentence and the word again, one file per item, with pauses for writing.
The usual format
Each item gives the number, the word, a sentence that uses it and the word once more: “Number one. Necessary. It is necessary to wear a coat today. Necessary.” Then comes a pause long enough to write. The sentence matters most for words that sound alike, such as there and their or night and knight, and for words that read differently in different meanings, such as read or lead. For a spelling bee drill, say the word, a short definition and the sentence.
Prepare the list
Put the list in a table with two columns, word and sentence, one item per row. Keep the sentences short and natural, and check that each sentence makes the meaning clear. Write numbers inside sentences as words if a child must not see the digit. If a word is a name or a rare term, listen to it once before the test: the voice reads letters by their usual sounds.
Voice every item
Install the free engine with python -m pip install superclonica-cpu and, in Python, after import csv, from superclonica import Superclonica and tts = Superclonica(), then for n, row in enumerate(csv.DictReader(open("words.csv", encoding="utf-8")), 1): text = f"Number {n}. {row['word']}. {row['sentence']} {row['word']}."; audio, rate = tts.synthesize(text, voice="voice.json", lang="en", seed=2); tts.save(f"{n:02d}.wav", audio). You get one file per item. The engine trims the silence at the start and end of each file, so add the writing pause yourself: join the files in any audio editor with 8 to 15 seconds of silence between them, or play them one by one and pause the player.
Voice and pace
Try voices free on their pages with a whole item, 30 to 100 characters, no account needed; a calm narrator keeps every word clear. For younger children or learners, slow the voice with --speed 0.9 or down to 0.7, the slowest setting. Use the same voice and seed for every list so the test always sounds the same.
Spelling lists in other languages
The same method works for vocabulary dictations in German, French, Spanish, Russian and other languages, and for learners of English in Japan, Korea or anywhere else: the engine speaks 31 languages; Chinese is not supported yet. Choose the language of the words, not of the learner, so every word is pronounced as it should be.