Skip to content
Nuvexa AI
Menu
Voice · PRACTICAL TUTORIAL

Prepare a two-speaker script without mixed dialogue

Speaker labels assign each line before synthesis. A voice choice alone cannot tell the tool where one character ends and another begins.

Published by Nuvexa AI · Published · 2 min read

Start here

This example uses the sample details below. Replace them with your own facts before applying the instructions.

  • Host asks a question; guest answers; two selected voices.

Start with the facts

Host asks a question; guest answers; two selected voices. Read the script aloud once before generating audio. Keep speaker names and punctuation consistent so the intended delivery is clear.

Spot the problem

The wording or action to avoid is: “Generate the whole conversation as one unlabelled paragraph.” Speaker labels assign each line before synthesis. A voice choice alone cannot tell the tool where one character ends and another begins.

Make the correction

Host: What arrives in this purchase? Guest: A PDF pattern; the finished toy is not included.

Check before using the result

Play the exchange and confirm that the question and answer use the intended voices in the correct order.

Worked example and reusable input

Use the stated inputs to reproduce checked local outputs. Replace sample facts when applying a brief to your own work.

TASK
Prepare a two-speaker script without mixed dialogue

SAMPLE INPUT — replace with your actual facts
Host asks a question; guest answers; two selected voices.

AVOID
Generate the whole conversation as one unlabelled paragraph.

CORRECTED APPROACH
Host: What arrives in this purchase? Guest: A PDF pattern; the finished toy is not included.

ACCEPTANCE CHECK
Play the exchange and confirm that the question and answer use the intended voices in the correct order.

Key takeaways

  • Speaker labels assign each line before synthesis. A voice choice alone cannot tell the tool where one character ends and another begins.
  • Play the exchange and confirm that the question and answer use the intended voices in the correct order.