Thanks for the info. After a brief look at commandline piper.exe it does appear that creating a WAV file from a text file may be more straightforward than reading aloud from the same text file. Whether that would extend to mass creation of WAVs from a whole book, or even multiple books, simultaneously, is hard to predict without testing.
I think it's possible my plugin *could* be made to work with Piper instead of winsapi but unfortunately it would need a lot of changes and a time commitment my current home situation won't allow. I don't think the creation of an "official" Piper method to generate WAV would significantly reduce the effort required.
I've already suggested, in the main plugin thread, that anyone who wants to continue using the existing plugin should install a copy of calibre Portable <= v7.17, independent of their main calibre. It should continue to work in the short term unless/until MS remove the old SAPI speech engine in a Windows update.
|