TTS Narrator output wav



We are currently still using the old Pipeline to generate TTS Daisy 2.02's.

Is it possible to leave the audio uncompressed? So instead of MP3, get WAV as output in the Daisy 2.02.

This could be a workaround for us to split single volumes into multi volumes, using the DTB Audio Encoder Splitter (which we already use for real narrated audiobooks).

Thanks in advance,

I dont think you can get WAV files from Pipeline 1 Narrator unless the scripts are modified significantly.

You may use the newer tool - Tobi to generate speech for your DTBook XML files. In Tobi you can keep the audio files in WAV format. However, note that Tobi will not create a DAISY 2.02 book, it will export only in DAISY 3 (Z3986) format.

If you want to know more about Tobi, you may visit Tobi page on DAISYpedia which contains step-by-step tutorials and videos.


As the audio is generated you can follow the creation of the temporary WAV files in the pipeline-temp directory (that is from memory). So you can potentially have access to the raw WAV files , note that they are deleted after the MP3 is created.
The hairy bit then would be to match the individual WAV files to the different entries in the *.smil files which are then referenced in the content.html file.

Potentially, there are generated in order so you might be lucky and have each span class="sentence" element that would correspond sequentially to the next WAV file.

Loads of fun I'm sure !!

best regards / Colm