Transcripts and Captions allow you to streamline the process of repurposing and reviewing content, adhere to accessibility regulations/expectations, and ensure your video content is available to a larger audience.
This article provides information about closed captions and instructions on generating and exporting a captions file of your Socialive recordings and clips.
- What are transcriptions?
- What are closed captions (CC)?
- What is a captions file?
- Which videos are supported for closed captions?
- What caption file formats are supported?
- What languages are supported?
- Generate and export a captions file
- Things to Remember
What are transcriptions?
Transcription refers to the process by which audio is converted into written text and made available as a document. It may or may not include time codes.
In Socialive, this will be made available in (.txt) format and will include time codes. The producer’s audio will also be included as it is often helpful to have their questions/guidance on the transcription as well.
What are closed captions (CC)?
Closed Captions (CC) are a text version of a video's spoken part and atmospherics (i.e., music and sound effects).
What is a captions file?
Captioning is similar to transcription, but captions always include time codes and adhere to standard file formats for use in video players. A captions file is a plain-text file containing the text of what is said in the video, along with time codes for each line of text. In Socialive, caption files will be made available in (.srt) or (.vtt) format.
Most popular video platforms support captioning by uploading one or more captions files to a particular video.
Which videos are supported for closed captions?
You can generate transcription and caption files for composite videos, ISOs, and clips. This feature won’t entail live captioning or imprinting any captions on video files; it is a feature that provides the necessary files.
What caption file formats are supported?
- WebVTT (.vtt) – Web Video Text Track- machine-readable, suitable to use as a caption track, supports metadata and formatting aimed at many/most web players
- SubRip (.srt)- machine-readable, suitable to use as a caption track, wider compatibility but fewer formatting metadata options.
- Transcript (.txt)- human readable, with time stamps
What languages are supported?
Socialive is using Amazon Transcribe as the back-end service to generate the transcripts and captions. So the languages supported are determined by this service, which may expand in the future. The current languages supported are:
- Arabic (Gulf, Modern Standard)
- Chinese (Simplified, Traditional)
- English (Australian, British, Indian, Irish, New Zealand, Scottish, South African, US, and Welsh)
- French (FR-French, and Canadian)
- German (DE-German, and Swiss)
- Portuguese (PT-Portuguese, and Brazilian)
- Spanish (Spain and US)
For the most up-to-date information, check out Amazon Transcribe’s support article here.
Generate and export a captions file
- In the Socialive Recordings, select the video or session you want to generate captions and transcript files for.
- For the clip or recording you want to generate the captions file, click the Download button.
- Select the option to Export Captions. The file may take a few minutes to generate, depending on the length of the video. While generating, the option will be grayed out as shown below:
- When the file is ready, you’ll receive an email from Socialive with a direct link to download the file in VTT, SRT, or TXT format.
- You (or any other admin in the account) can also go back to the Recordings Library, click Download, and select Download VTT, Download SRT, or Download TXT, depending on your desired file format. Once you have generated the caption and transcript files, they will always be available in the Socialive library to download.
Things to Remember:
The text for the transcript and caption files are being produced by the Amazon Transcribe service. Some things to be aware of:
1. Audio quality will impact transcription - if the speaker is speaking clearly at normal volume and in a generally quiet environment, the transcript should be accurate.
2. Sometimes, Amazon Transcribe will produce accurate results for some speakers but not others. It is always important to read over the files before adding them directly to your video in an external editing tool.
3. Amazon Transcribe struggles with accuracy if there are multiple languages detected in a recording. It will usually pick up the first language spoken.
4. There will be issues with files where no audio is detected; the email will indicate that the processing failed.
There will be an error when exporting captions for a video file over 2GB in size. An error email will be sent with directions for troubleshooting. If there is an error:
- Navigate back to the recording that you would like to generate captions for
- Click to export to audio (MP3), and wait for the success email to be received
- Lastly, return to the video to generate a caption on and click to export captions again