Media Analysis & Transcription preferences

Last updated on Jun 23, 2025

Customize transcription settings in Adobe Premiere Pro to improve accuracy, language support, and speaker labeling with Speech to Text.

The Preferences dialog is open to the Media Analysis & Transcription panel, showing options for caching analysis results, enabling visual analysis, automatic clip transcription, speaker labeling, and setting the default language.
Control how media is processed for speech-to-text and visual analysis.

Configure Visual Analysis and Transcription settings in the Media Analysis & Transcription preferences to customize your media processing.

UI option

Function

Cache analysis results for re-use

Lets you choose where media analysis and transcription will be cached:

  • Next to the media as a sidecar file: Caches media analysis and transcriptions in sidecar files next to the media. If the media is imported into another project on any computer, it won't require reanalysis.
  • In the Media Cache: Caches media analysis and transcriptions in the Media Cache folder. If the media is imported into another project on the computer, it won't require reanalysis.
  • Do not cache: Saves media analysis and transcriptions only in the project file. If the media is imported into another project on any computer, it will require reanalysis.

Analyze all imported media to visually search your projects

Enables automatic visual analysis of every clip you import into your project.

Automatically transcribe clips

Auto-generates speech-to-text transcripts for imported clips or all clips in your sequence.

Transcription preferences

Lets you choose which media is to be transcribed:

  • Select Auto-transcribe all imported clips to transcribe all existing and newly imported clips using the selected language.
  • Select Auto-transcribe only clips in sequence to transcribe all the clips in your sequence using the selected language.

Speaker Labeling

Lets you choose speakers to be labeled:

  • Select Yes, separate speakers to have labeling for separate speakers.
  • Select No, don't separate speakers to disable speaker labeling and treat all dialogue as coming from a single speaker.

Enable language auto-detection

Allows automatic detection of the spoken language in a clip during transcription. When enabled, the default language will be used for locale selection or as a fallback if the language can’t be detected.

Default language

Lets you choose the desired language to transcribe clips.