Automatic Subtitle Repositioning with On-Screen Text Detection

MediaCAT now detects on-screen texts and adjusts subtitle placement accordingly, ensuring subtitles and on-screen elements remain clearly distinguishable. To further refine the positioning, users can customize the sensitivity level for different types of content, allowing for optimal detection and positioning. This new feature makes it easier for users to identify subtitle and on-screen text overlaps during QC.

To use this feature, enable on-screen text detection during the creation of a transcription project with video files.

If you missed enabling the on-screen text detection during the project creation process, you can still enable it by accessing the Review Reposition Conflicts modal from the Subtitle Position menu. Once enabled, the system will detect on-screen texts throughout the video and display a list of potential positioning conflicts.

In the Review Reposition Conflicts modal, you can click on the detected items to review how and where on-screen texts appear in the video. The engine will automatically suggest repositioning to avoid overlap, though manual adjustments can be made if needed. Additionally, you can adjust the sensitivity setting—default is moderate, but we recommend using low sensitivity for content such as sports broadcasts, where background text and banners are common. This ensures optimal subtitle placement without manual intervention.