Narrator, my fun text-to-speech app that allows mixing multiple voices in a document and exporting to iTunes, has been updated to version 2.0.3.
This release includes some important bug fixes, and is a recommended update for everyone:
Fixed an issue with the Export to iTunes feature, where the importing-into-iTunes phase could fail with a "parameter error" if exporting a lot of text.
The lyrics value in the Export to iTunes is now limited to a length of 20,000 characters, to avoid overloading the export mechanism.
Fixed an occasional crasher if the word highlighting is out-of-sync with the spoken text for some reason.
Disabled automatic emphasis in speech for non-Apple voices, to work around a bug in the Cepstral voices, where it gives a spurious exception regarding an unexpected emphasis close tag, when there is nothing wrong with the text.