No reading tracking with some speech synthesis

[pat-emmabuntus]

When I use the Svox Pico or Piper speech synthesizers during reading, the word read is not highlighted and the reading does not change pages automatically.

If I use eSpeak-ng when reading, the text read is highlighted and when the reading reaches the bottom of the page there is a page change, see screenshot below.

I used Calibre 6.13 on Debian 12.

Is there any customization or option from Calibre to enable the playback tracking feature?

Moderator Notice
Please read the guidelines about large inline images and use the attachment method

[kovidgoyal]

You will get word tracking if the engine you use provides that data to calibre, there is nothing calibre side that can be done about it.

[JSWolf]

You should upgrade to calibre 7 to get the latest features.

[pat-emmabuntus]

Originally Posted by kovidgoyal

You will get word tracking if the engine you use provides that data to calibre, there is nothing calibre side that can be done about it.

Thank you Kovid Goyal for your feedback.

Indeed this mode of operation comes from the synthesizer engine and in the case of speak-ng it is this option "EspeakIndexing" at 1 in espeak-ng.conf which activates word tracking.

[pat-emmabuntus]

@kovidgoyal It's not just the problem of tracked words during voice reading, there's also the fact that Calibre doesn't change pages when voice reading switches to another page. On the other hand, it works when there is a change of chapter.

This function is very important for us because it must be used by blind people and if during voice reading there is only following on the chapters, these people will have to read an entire chapter otherwise when reopening Calibre the software will reread a passage already read.

Is there an option to have Calibre change pages during voice playback or a function to tell Calibre to remember the voice reading position in the document when closing the document viewer?

Thank you in advance for your advice.

[DNSB]

As Kovid Goyal said, unless the software supplies the reading position data to calibre, calibre has no way of knowing where your reader software is reading from.

From your original post, it sounds as if eSpeak-ng supplies this information and Svox Pico or Piper speech synthesizers do not. I would suggest checking with their respective tech support to see if such support can be added to their products.

[pat-emmabuntus]

FYI, Firefox's narrate mode tracks sentences when playing audio, see this video : http://share.emmabuntus.org/Accessib..._svox_pico.mp4

More information on this function : https://support.mozilla.org/en-US/kb...free-web-pages