---
title: "Notes"
url: https://diatts.com/notes/
description: "The full archive of notes published on Diatts, grouped by month."
---

# Notes

Every note published here, most recent first, grouped by the month it appeared.

## July 2026

- Dialogue · July 28, 2026[What makes synthetic dialogue sound unrehearsed](https://diatts.com/notes/what-makes-synthetic-dialogue-sound-unrehearsed/)Two synthetic voices can be individually convincing and still fail the moment they have to talk to each other. The difference is almost never the timbre.

## June 2026

- Prosody · June 12, 2026[Prosody is the harder half of speech](https://diatts.com/notes/prosody-is-the-harder-half-of-speech/)Getting the phonemes right is a solved-enough problem. Deciding which syllable carries the weight of a sentence is where synthesis still shows its seams.

## May 2026

- Data & Practice · May 4, 2026[How to read a speech model card](https://diatts.com/notes/how-to-read-a-speech-model-card/)Model documentation is written to be skimmed and is most useful when it is not. A short guide to the sections that actually constrain what you can do.

## March 2026

- Evaluation & Ethics · March 19, 2026[Consent and provenance in voice cloning](https://diatts.com/notes/consent-and-provenance-in-voice-cloning/)A few seconds of reference audio is now enough to reproduce a speaking voice. The interesting questions are all about record-keeping.

## January 2026

- Data & Practice · January 27, 2026[Preparing audio datasets for speech models](https://diatts.com/notes/preparing-audio-datasets-for-speech-models/)Most of the quality difference between two speech systems was decided during data preparation, long before either was trained.

## November 2025

- Evaluation & Ethics · November 15, 2025[Evaluating speech systems without a leaderboard](https://diatts.com/notes/evaluating-speech-systems-without-a-leaderboard/)Speech quality is a perceptual property, and perceptual properties resist the single number that comparison tables want.