Computer narrated epub3 with word-level text-audio-synchronization


Hello daisy ;-),

if we produce computer narrated epub3 books with the dp2 we get this with a well sentence-level text-audio-synchronization. Now our desire is that we also want the flavor of word-level synchronization. How is this feasible? Is it configurable?

Another question is, if it`s possible to increase the worker threads count during the audio generation in the TTS-adapter? At the moment I think there are 3 per job.

Much thanks in advance,

Hello Lars,

It looks like the granularity of synchronisation is not configurable at the moment. How feasible it is I can't say. We will look into it and try to get an idea of what needs to be changed in the code to achieve this. How important is it for you?

Regarding the number of threads for audio generation, can you try one of the following properties?

- threads.number
- threads.encoding.number
- threads.speaking.number

You can specify properties in a configuration file that you pass as an input to the script.

A config file looks like this:

<property key="threads.number" value="5"/>

Great Bert.
Thank you for your notes.

I will try to set the number of threads via config file.

The ability to configure the granularity of synchronisation is not urgent, but it would be nice to have this feature in the future.

Best Lars