Creating and editing captions
Caption metadata
The ability to add caption metadata allows you to add information to your captions beyond what is being said at that point in time. CaptionHub currently supports two types of metadata:
- Speaker identification
- New paragraph flags
If you've enabled Speaker detection, our speech recognition engine will attempt to identify who's speaking and assign a label to their captions.
CaptionHub labels the first speaker as S1, followed by S2, S3 and so on. These labels provide a useful starting point, but you can update them to reflect the speaker's name.
To assign or edit a speaker:
- Click the speaker icon next to the caption number in the spreadsheet view on the Edit page. This opens a pop-up window. The icon will appear differently depending on whether a speaker has already been assigned:
- No speaker assigned
- Speaker assigned
- Select a speaker from the existing list, or click Add new speaker to create one.
- Alternatively, enter the speaker's name directly in the text field and click Add and assign.
- To edit an existing speaker's name, click the pencil icon.
- To remove a speaker from a caption, click Remove.
- To assign a different speaker, click the speaker icon and select another speaker from the list.
Speakers are automatically colour-coded in the Editor view to make them easier to identify throughout the caption set. These colours are for visual reference only and do not change the colour of the caption text.
If you’d like the caption text for different speakers to be displayed in different colours, you can map speakers to colours by selecting Map speaker colours… from the cog menu.
This opens a modal where you can select a colour for each speaker.
Make your selections, then click Apply colours to confirm.
The caption text for each speaker will now be displayed in the selected colour in the video preview and in exports that support colour, making it easier to distinguish between speakers.
Note: The current transcript export format that supports and reflects changes made to speaker identification is DOCX (Word).
Transcribed text can be difficult to read when presented as a single block of text. For supported export formats, you can use paragraph flags to indicate where a new paragraph should begin.
To add a new paragraph, click the paragraph button. This marks the caption as the start of a new paragraph in your transcript.
To remove the paragraph flag, click the button again.
