Transcoding of subtitles is the process of reading subtitle data from a signal or data stream and convert those subtitles on the fly to one or more other transmission formats. Re-timing with specific delays can be configured between 1 minute and 23 houres.


Formats in and out, transcoded by the Cavena Subtitle Transmission Unit, STU, DVB Subtitling can be an input format.


A typical example of an STU transcode application is to read US Line-21 captions from an NTSC signal and transcode it into EBU Teletext for PAL output. Another example is to read EBU teletext from the VBI of an SDI signal played out and transcode this into DVB Subtitling to a DVB multiplexer. It can also be incoming DVB Subtitling retime stamped and output as DVB Subtitling. Scaling fromS D to HD DVB Subtitles or vice-versa can take place.
The Cavena transcode solution is modular and flexible. Capacity is dependent on input format being an SDI signal or IP and number of languages to be transcoded. The start module being a 1RU solution, based on an HP server or similar.
The Cavena transcode solution can be standalone or automation controlled.
The Cavena transcode solution can be upgraded to cater for file based play out or a combination of transcode and file based play out.



Captioning or subtitles for hearing impaired has become a mandatory requirement for most TV stations and the service is highly appreciated. Most events can be prepared days before and the subtitles are timecoded with a good synchronisation between audio and subtitles. But for live events, where subtitles are written simultaneously, either by keyboard or by voice recognition, there is a delay between the subtitles and the TV service. This delay can be more than 8 seconds and will severely degrade the TV experience for the hearing impaired. At the same time, there is a video delay induced by the MPEG encoders. This delay can, depending on video standard, be 2-6 seconds and can be used to gain a higher synchronicity between subtitling and audio during Live subtitling. Cavena has therefore implemented a, per session, variable delay. For timecoded subtitles or subtitles that are cued the delay of the subtitles corresponds to the delay of the video encoder, but during a live session the translator can adjust the delay to gain a couple of seconds. This functionality is included in the Cavena live protocol and is currently being implemented by DR.