Is it possible to extract timing information or segment a single piece of text within Swift with a view to captionning - this would also provide a solution to the query re Daisy books someone else has requested.
I'm currently using Cepstral as an end-user option to replace poor quality speech (Festival/FestVox) in an open-source PHP eFolio app being developed for people with intellectual disabilities.
Also wondering what the implications for licensing are [this app will be used initially by small speciailist schools/colleges in the UK]. It is running server-side and caches mp3 audio of any text on pages when they are created - will eventually run as a single cron job over night due to the time required for text2wav/swift and lame to process longer runs.
Thanks,
Simon