Which live stream workflow do I use best for broadcasting?

Introduction:

Broadcasters often require full control over their video and audio signals when using third-party systems like Deep Live Hub. Typically, broadcasters send a video stream to Deep Live Hub and only receive a text stream (subtitles and translations) back. The broadcaster then merges the video and subtitle streams before broadcasting.

Deep Live Hub provides the text stream via HLS or HLS-realtime in regular or partials mode. Since broadcasting generally allows for some delay, you can focus on subtitle quality by using longer processing times.

In some cases, broadcasters may want to ensure that subtitles are error-free, particularly for important events like political broadcasts. In such cases, activating the Live Editor for human intervention may be a good option, though it introduces an additional 15-second delay.

Creating a Live Stream Workflow for Broadcasting:

• Log in to Deep Live Hub:

  • Click the red "+ Workflow" button and select "Live Stream" from the top menu.

• Choose a Workflow Template:

  • Select the "Custom" configuration workflow template.
  • Enter a name for your workflow and click "Finish."

• Configure the Workflow:

  • Navigate to the Process tab.
  • Select your translation languages from the dropdown menu.
  • If you’ve created dictionaries, select one to improve word recognition.
  • Confirm your settings by clicking the red "Save" button.

• Set ASR Configuration:

  • Choose the ASR processing time that suits your needs. For broadcasting, we recommend a longer processing time (10-15 seconds) for better subtitle and translation quality.
  • Partials mode delivers transcripts quickly, but subtitles may require further corrections. Partials currently do not affect translations.
  • Confirm your settings by clicking "Save."

• Enable the Live Editor (Optional):

  • If human editing is needed for final corrections before broadcast, enable the Live Editor by toggling the slider in the Process tab.
  • This will introduce an additional 15-second delay for human review.
  • Confirm your settings by clicking the red "Save" button.

• Set Up Outputs:

  • Navigate to the Outputs tab.
  • By default, an SRT Configuration and Video Output Pull Configuration are provided.
  • If using partials, create a new configuration by selecting "Text configuration."

• Collect Input URLs:

  • In the Overview tab, gather the RTMP or WebRTC URLs and your stream key for input.
  • If the Live Editor is enabled, find the editor’s URL and PIN under the Processing tab.

• Manage Streams:

  • View the log of all past and current streams under the Streams tab.
  • To disable or delete a workflow, go to the Settings tab. Stopping the stream in your encoder will deactivate the workflow.