NHK is using a new media file extraction and re-formatting system to add audio description tracks to archive programmes.
The work is an extension of the Japanese public broadcaster's project to make its library of material on P2 solid state digital video cards compliant with loudness regulations. This began in late 2012 using UK developer Emotion Systems' eFF (Emotion File Finish) system to bring the data-based programmes in line with the TR-B32 standard.
Audio description is now a major requirement in TV broadcasting to give blind and sight impaired people a sense of what is happening on screen. While this is added to new programmes during post-production, archive material has to be re-worked to include descriptor tracks.
Raphael Samad, business development director of Emotion Systems, says NHK was having to load P2 material from the storage area network (SAN) into a video editing suite, where the soundtrack was saved into a separate file before it could be worked on in a sound studio. Once audio description was added the file went back through the process to save it into the SAN.
"The whole procedure could take up to eight hours" he adds. Samad explains that NHK engineers asked if the work could be streamlined, along similar lines to the file-based monitoring and correction of loudness levels. Emotion Systems produced eRAP, which extracts the audio content of a file without it having to be played out through a video suite before being transferred into a WAV format container.
Once in WAV form the audio description track is added to the programme, which is then re-wrapped into its original container. "Extracting an hour show takes several minutes," comments Samad, "The audio descriptor is put on it and the file is reformatted, which again takes only minutes.
This is allowing NHK to work on a programme during the day and broadcast it the same evening." The eRAP system was implemented at NHK in September as part of an ongoing project to put audio description on all archive material.