I recorded 1,180 TTS clips to find out what voices misread in LLM output
说明与用途
使用 speechready 对 LLM 输出进行预处理,将语音误读率从 32% 降至 8%,在英语和土耳其语中同样有效,且未导致控制文本回退。
问题
Voice agents often misread text generated by LLMs, especially numbers, dates, phone numbers, and code, leading to incorrect spoken output.
结果
Preprocessing with speechready reduced overall misread rate from 32% to 8%, with similar improvements in English and Turkish, and no regression in control prose.
工作流
- 1. Collect LLM-generated text containing markdown, numbers, dates, phone numbers, URLs, and code.
- 2. Run speechready to clean, verbalize, compile, and chunk the text.
- 3. Synthesize the original and processed text with each TTS voice.
- 4. Transcribe both recordings with gpt-4o-transcribe.
- 5. Score the transcriptions against the intended words, forgiving transcription conventions.
- 6. Calculate misread rates and compare before/after.
- 7. Iterate on rules and provider profiles based on held‑out tests.
证据
I recorded 1,180 TTS clips to find out what voices misread in LLM output
Result, on the same 1,180 recordings: All 32% → 8%
The scorer forgives the ways a transcription engine writes speech down: digits for spoken numbers, "21st" for "twenty-first", "NPM" for "en pi em".
Providers disagree on what they handle. Google's Chirp voices read 25°C fine; others do not.
- 记录日期: 2026-11-02T22:29:04.921Z