"RokuJoel" wrote: To do this I modified some open source text-to-speech software to output .wav files that could be played by roAudioResource.
Are you generating the .WAV files on the Roku or are you sending the text to a server and sending . WAV files back. I have Just started working on trying to generate sound effects with brightscript and would be interested in your experiences.
I was sending the text to a node.js based server script running on my mac that I modified slightly to create.wav instead of .mp3 files and use an open-source TTS command line, (built in mac TTS won't output .wav files) and send the .wav files back to the Roku. Pretty simple text to speech where on an IsListItemFocused() event I send the text of ShortDescriptionLine1 and 2 to the server, download the resulting .wav to tmp:/, create an rAaudio resource and then play it. The server caches the audio files so once they are generated making subsequent requests for the same string faster.