Generate subtitles from the audio track of your videos now with the Azure AI Video Indexer service. Learn how to enable and use the service from this page.

In this article:

Configure Azure AI Video Indexer with RMS

Assign the Contributor role in the Video Indexer account

Register Azure AI Video Indexer in RMS

Generate Closed Captions

AudioAnalyzer Preset

Create Transform

Validation

Before you begin

The Video Indexer account is not created during the RMS deployment. Create a Video Indexer account or use an existing Video Indexer account (not created during the RMS deployment). The Video Indexer account must exist outside of the RMS managed resource group. 

Configure Azure AI Video Indexer with RMS

Assign the Contributor role in the Video Indexer account

1. Go to the RMS managed resource group and locate the Managed Identity section. Find the RMS user-assigned managed identity. Its name is id-rms-{unique-suffix}. Copy for later.

2. Open your Azure AI Video Indexer page.

3. Navigate to Access Control (IAM) tab.

4. Click Add > Add role assignment.

5. Switch the tab to Privileged administrator roles and click on the Contributor role.

6. Click on Managed identity and Select members.

7. In the side panel, set the Managed identity as User-assigned managed identity, and in the Select field, select the RMS managed identity you copied in the step 1.8. Click Select > Review+assign.

You have now assigned the RMS managed identity to the Video Indexer, and RMS can now access the account.

Register Azure AI Video Indexer account in RMS 

1. You will need the following data to register the Video Indexer:

• Video Indexer Account name
• Resource group where the Video Indexer has been created
• Subscription ID

2. Go to the RMS Console > Settings > Service Connections > Video Indexer Account.

3. Select the + Add Account.

4. Enter the Video Indexer account details and click Add video indexer account.

This completes the setup process for Azure AI Video Indexer in RMS.

Generate Closed Captions

AudioAnalyzer Preset

The AudioAnalyzer preset is a JSON configuration that tells Azure AI Video Indexer how to process the audio track of your video. It supports multiple languages to generate VTT (WebVTT) subtitle files.

AudioAnalyzer preset JSON sample:

{
"audioLanguage": "en-US",
"mode": "Standard",
"@odata.type": "#Microsoft.Media.AudioAnalyzerPreset"
}

The audioLanguage parameter accepts standard language codes. Here are some commonly used options:

For the complete list of supported languages, visit videoindexer.ai > Model customizations > Language ID tab.

Create Transform

1. Open RMS Console > Transforms and Jobs.

2. Click on Create Transform.

3. Select Custom JSON preset tab.

4. In the Custom JSON preset editor, paste the AudioAnalyzer preset (shown above).

4.1. Optionally, modify the audioLanguage to match your video's language.

5. Set the transform name (e.g., "AudioAnalysis-Captions-EN").

6. Click Create Transform.

Validation

Follow these steps to create an encoding job that generates captions:

1. Upload your media into the asset.

2. Click Create Job.

3. Submit the job using the Transform you just created.

4. Navigate to Assets in the RMS Console after the job is complete.

5. Select the created output asset.

6. Go to the Asset blob section of the asset.

7. See the generated .VTT (WebVTT) file, along with transcript and metadata files.

This completes the setup and usage process for generating closed captions using Azure AI Video Indexer in RMS.