Skip to main content

Do you have a style guide?

This article describes the key style preferences Sonix recommends for editing and cleaning up transcripts—especially for podcasters and radio shows—to achieve polished, reader-friendly text ready for publication or sharing.

Taylor Jennings avatar
Written by Taylor Jennings
Updated over 2 weeks ago

Important Notes

  • Sonix does not currently maintain an official, detailed style guide.

  • The points below are stylistic suggestions—how you apply them may vary depending on your content, audience, or platform.
    ​
    ​

Step-by-step Instructions

Here are the stylistic editing steps Sonix suggests when cleaning up a transcript:

  1. Break into paragraphs - Create new paragraphs (press Enter) where it feels natural—often when a new speaker begins. Include the speaker name at the start of each paragraph.

  2. Keep filler words - Retain filler or conversational words such as “kind of,” “you know,” “like”. They help preserve the natural flow and tone.

  3. Sentence beginnings - It’s okay to start sentences with “And,” “But,” or “So.” These can reflect how people speak in informal or conversational contexts.

  4. Colloquial speech - You may leave informal words like “gonna” rather than changing them to “going to.” Authentic voice can be important.

  5. Speaker labels for unidentified voices - If you can’t identify the speaker by name, use “Male Voice” or “Female Voice” as speaker labels.

  6. Editing style / verbatim level - Use clean verbatim (or non-strict verbatim) editing. That means you keep the content (what was said) but remove stutters, false starts, “ums/uhs,” background noises, etc., unless they are relevant.
    ​
    ​

Common Issues and Troubleshooting

What if I'm editing and the transcript still sounds choppy?

Break long paragraphs into smaller ones at natural speaker changes. Also, ensure you're removing stutters, false starts, or repeated filler words to smooth the flow.

Should I remove every “um,” “uh,” or filler word?

Not necessarily. The suggested style is clean verbatim, which means removing the most distracting filler content—but retaining enough conversational tone so it still sounds natural. Over-editing might lose the speaker’s voice.

Frequently Asked Questions

Is there a full, formal style guide I can download or follow?

No. Sonix currently does not have a complete formal style guide. The points listed here are the key stylistic preferences we suggest.

What is “clean verbatim” vs. “strict verbatim”?

Clean verbatim retains what is said but cleans out things like false starts, filler words, stutters, etc. Strict verbatim would include everything—even those filler words, false starts, and nonverbal noises—exactly as spoken.

Is every transcript expected to follow these suggestions?

Not necessarily. These are suggestions. Depending on your audience or use case (e.g. legal transcripts vs. podcast transcripts), you might choose a different style, stricter or looser.

For further assistance, please contact Sonix support at [email protected] or through the chatbox located in the bottom right corner of our website.

Did this answer your question?