Speaker Metadata & Export
Speaker data in Closed Caption Creator goes beyond display labels in the editor. Speakers are a structured metadata layer that influences how transcripts are organized, how AI-imported dialogue is attributed, and how formal delivery documents such as CCSLs and dialogue lists are generated. This page covers automatic speaker identification from AI imports, inserting speaker names into Event text, and the role of speaker and recipient data in exported formats.
Automatic Speaker Identification​
When you import an AI-generated transcript from services such as AssemblyAI, Deepgram, Google Speech-to-Text, or Rev AI, Closed Caption Creator automatically creates speakers from the diarization data embedded in the transcript and assigns them to the relevant Events. This means that after a transcription import, your Event list is already populated with speaker attributions without any manual assignment.
The speaker names generated by the AI service are often generic labels such as "Speaker A" or "SPEAKER_0". After import it is important to open the Speaker Manager and update each speaker's name and color to reflect the actual participants. If the AI has created multiple speaker records for the same person due to a misidentification, you can use the Merge Speakers option to consolidate them into a single record. Any Events that were tagged with either label are automatically updated to use the merged speaker.
Automatic speaker identification is also performed by Closed Caption Creator's own transcription feature when a transcript with diarization is generated directly within the application.
Inserting Speaker Names into Event Text​
In some delivery formats — particularly SDH subtitles and dialogue-oriented files — the convention is to display the speaker's name as part of the on-screen caption text. Closed Caption Creator provides an Insert Speaker Name option (available from the Insert menu) that automates this for any number of selected Events.
When you open the Insert Speaker Name dialog, you can choose whether to use the assigned speaker for each Event or specify a custom name. Additional options include a prefix character (such as a dash or bracket) that appears before the name, a suffix character that appears after it, and a toggle to render the name in uppercase. You can also choose whether the speaker name is inserted at the beginning of a new line at the top of the Event, or inline at the start of the existing text.
Once applied, the speaker name becomes part of the Event's text content and is exported as part of the caption output alongside the dialogue.
Speaker and Recipient Data in Exports​
Speaker and recipient metadata is surfaced in several export formats that support structured attribution.
In a CCSL (Continuity, Combined Spotting and Lining) export, speaker information is included in the designated speaker column. When a recipient is also assigned to an Event, both the speaker and the recipient appear in the Spotting column, clearly documenting the exchange between the two parties. This level of detail is required for broadcast compliance documents submitted to post-production facilities and broadcasters.
In a Dialogue List export, speaker attribution appears as part of the line entries, allowing the list to function as a formal record of who delivers each spoken line. This is particularly valuable for localization workflows, where translators need to understand speaker identity to maintain appropriate voice and register in the target language.
Speaker names and notes are also automatically included in any transcript export from the application. This behavior is always on and does not require any additional export configuration.