Resemble Fill enables you to take existing recordings of speech and modify it seamlessly. Use AI Audio Inpainting for programmatic audio ads, dynamic streaming ad insertion (SAI), voice assistants, and so much more.

Table of Contents

Using Resemble Fill Through The API

  1. Find the uuid of the voice you want to use Resemble Fill with. You can obtain a voice’s uuid by querying the API for all your voices: https://docs.app.resemble.ai/docs/resource_voice/all
  2. On your end, collect all the recordings you want to modify with Resemble Fill.
  3. Use the *Create A Recording* API to upload all the recordings you collected in step 2. Make sure to set the fill attribute to true and use the voice uuid you collected in step 1.
  4. Use the Build A Voice API to train your Resemble Fill model! Make sure to set the fill attribute to true and use the voice uuid you collected in step 1.
  5. The Resemble Fill model will start training. You will be unable to use Resemble Fill until the model is ready (you will be notified via email once it’s ready).
  6. Once your model is ready, you can now synthesize using the uuid ’s of the recordings you uploaded in step 3. If you dont have the recording uuid ’s you can query the API to retrieve them using the Get All Recordings API.
  7. Use the following SSML to compose a Resemble Fill synthesis request:
<resemble:fill recording_uuid="<YOUR-RECORDING-UUID>">The modified text</resemble:fill>
  1. Using the SSML you composed in step 7, make an API call to Create A Clip.

Using Resemble Fill Through The Editor

  1. Visit the clip editor

Untitled

  1. Choose the voice you want to use Fill on, and place your caret in the text area:

Untitled