Configure MediaCopilot transcription language and subtitle export

Avatar
Cires21
Follow

Introduction

Use this guide to configure transcription language and subtitle export settings in MediaCopilot for multilingual live events. It explains how to choose the best source language, select subtitle output formats, and validate your workflow before going live.

Before you start

Before configuring a live transcription workflow, confirm the following:

  • You have access to the MediaCopilot project or workspace for the event.
  • The live audio source is available and correctly routed.
  • You know which language or languages will be spoken during the event.
  • Your export destination supports the subtitle or caption format you need.
  • You understand whether the event should be processed as one workflow or as separate jobs per language.

Configure transcription language

  1. Open the MediaCopilot job or workflow for the live event.
  2. Select the source language that best matches the primary live audio.
  3. If the event includes multiple spoken languages, decide whether to process each language in a separate job or use a single multilingual workflow.
  4. If available, choose the transcription model that best fits the content type, accent profile, and audio conditions.
  5. Save the configuration before starting transcription.

Enable subtitle and caption export

  1. Open the export settings for the transcription job.
  2. Enable the subtitle or caption output formats required for your workflow.
  3. Confirm the export destination supports the selected subtitle format.
  4. Review any delivery rules, naming rules, or destination permissions that may affect the export.
  5. Start the live job and verify that subtitle output begins as expected.

Multilingual live content recommendations

For multilingual events, the most reliable approach depends on how the audio is produced and how the subtitles will be delivered. Use the following recommendations as a starting point:

  • Use separate jobs when each language requires its own transcription, review, or export path.
  • Use a single workflow when the event is primarily one language with limited multilingual segments.
  • Confirm that the subtitle destination can handle the required format, timing, and language metadata.
  • Test the workflow with a short live segment before production use.

If subtitles are delayed or incomplete

If subtitle output is delayed, missing, or incomplete, check the most common causes first:

  • Input audio quality is too low, inconsistent, or distorted.
  • The selected transcription model does not match the language or content type.
  • Post-processing rules for segmentation or line length are making subtitles appear late or truncated.
  • The export destination is not configured to receive the selected subtitle format.

Recommended checks

  1. Verify the live audio feed is clean and stable.
  2. Confirm the source language is set correctly.
  3. Review the transcription model selection.
  4. Inspect segmentation, reading speed, and line length rules.
  5. Run a short test segment and compare the output against the live source.

Best practices

  • Always validate language selection before starting a live job.
  • Use a short test segment to confirm timing and export behavior.
  • Keep subtitle formatting rules consistent across events when possible.
  • For complex multilingual productions, document whether each language is handled separately or in one workflow.

Next steps

If you continue to experience transcription or subtitle export issues, review your workflow configuration, confirm destination compatibility, and contact support with the job details, source language, subtitle format, and a sample of the affected output.

Additional information

For related product information, visit MediaCopilot at https://cires21.com/mediacopilot or contact the Cires21 team at https://www.cires21.com/contact.

Comments