Recently we released speech recognition for live streaming in Nimble Streamer to process incoming streams and generate WebVTT subtitles from the audio within the stream.
Now we’ve added one more subtitles format: CEA-708. Once Nimble Streamer gets the incoming data and processes it through the speech recognition engine, the output may get the subtitles in the content. Since subtitles are carried via NAL units, any protocol may be a transport and the playback is only limited to the player. Namely, CEA-708 can be delivered via SLDP low latency protocol and played in SLDP HTML5 web player.
Let’s see how CEA-708 subtitles are set up.
First, please follow Automated voice recognition setup article to enable speech recognition, this is required for subtitles’ generation.
The second step is to enable CEA-708. Go to Nimble Streamer / Live streams settings menu. Here you can enable this on either server’s level or on each application’s level.
Click on Generate CEA-708 subtitles checkbox and save settings. They will be applied with Nimble instance within a few seconds.
At the moment, there will be around 2-3 seconds delay of subtitles output compared to the HLS.
Notice that Nimble supports only a limited character sets for CEA-708 captions. These are:
- BNA – Basic North American character set
- SNA – Special North American character set
- WES – Extended Western European character set : Extended Spanish/Miscellaneous
- WEF – Extended Western European character set : Extended French
- WEP – Extended Western European character set : Portuguese
- WEG – Extended Western European character set : German/Danish
Feel free to contact our team if you have any questions or suggestions regarding this feature set.