This article describes how to change how AI pronounces text in voice reading. There are two methods.

Additionally, corrections are unified across the entire team and will be reflected when reading other SOPs.

① Correct During SOP Viewing

1. Click the "Report" icon
   
2. Select the relevant section (in this example, MΩ)
   

3. Enter the correct reading and click "Register". When you open it again, it will display the specified content.
   

   
   ※ By default, users below Group Administrator level submit a "Report", and when a Team Administrator or higher reviews and approves the contents, it will be applied.

② Set in Advance from Dictionary / Templates

You can also set it from the sidebar by opening "Dictionary / Templates".

③ Bulk Import with Excel / CSV

When you want to register or update many pronunciations, you can import from an Excel or CSV file using the "Pronunciation Dictionary" tab's "Bulk import" button. This is convenient when you want to quickly apply a terminology glossary or your company's furigana list.

1. Open "Dictionary / Templates" from the sidebar and switch to the "Pronunciation Dictionary" tab
2. Click "Bulk import" in the upper right
3. Click "Excel template" in the upper right to download the template. If there are already registered pronunciations, they will be populated in the download, so you can edit and re-upload as-is
4. Drag and drop the template (or your own CSV / Excel with "notation / pronunciation / language (optional)" columns), or paste the contents
5. A diff preview is displayed, allowing you to check the contents in four categories: Add / Update / Skip / Error
6. If everything looks good, click "Register N items" to confirm

• Supported file formats: CSV, TSV, Excel (.xlsx / .xlsm). Character encoding is auto-detected as UTF-8 or Shift-JIS
• Language column: If omitted, it will be registered as Japanese (ja)
• Japanese pronunciation must be entered in hiragana or katakana. Any other characters (only kanji, alphabetic characters, symbols, etc.) will cause an error
• Pronunciations that are the same as existing ones are automatically skipped
• After registration, it may take several minutes for the changes to be reflected in existing audio (re-creation)
• When a user below Group Administrator level performs a bulk import, an "Report" is sent as an approval request to the Team Administrator or higher, just like a normal addition

Related Dictionary Features

"Dictionary / Templates" has the following dictionaries as well. All of them support bulk import via Excel / CSV.

• Notation Unification Dictionary: Replacement rules for text within the team (reference: Standardize Notation Variations (Notation Unification Dictionary))
• Terminology Glossary: Pre-specify translation terms for automatic translation (reference: Terminology Glossary)