Re: NVDASAPI: convey SAPI speech through NVDA
@18, as the first post says, it uses the NVDAControllerClient. The DLL doesn't have any of those abilities, you call the speech function in teh DLL and it sends that speech to the screenreader, and that's it. It would break terribly in games like Manamon with cutscenes, which obviously in Manamon isn't an issue since it has native screen-reader support but any game that depends on the TTS waiting. Either way, it seems pretty broken and could never be fully functional because of the way SAPI works. Within a functional SAPI voice, there should be a way to speak to an audio stream, which NVDA could not do. As I said earlier, the speech is queued internally in NVDA and there is no supported way, especially in the controller, to speak a string from your NVDA synth to an audio stream. The best we could hope for would be for someone to use the new speech refactor changes in 2019.3 to build a function into the controller that would be able to discern whether the screen-reader itself is speaking, but at no point will the SAPI voice see an audio stream of this speech. So this will probably break in terrible ways in any application that uses those methods to do any processing with the SAPI voice.
My research here is based on the front-facing interface for interacting with SAPI voices from other applications, because I've never been able to find a description of the process required to make a SAPI voice myself.
-- Audiogames-reflector mailing list Audiogames-reflector@sabahattin-gucukoglu.com https://sabahattin-gucukoglu.com/cgi-bin/mailman/listinfo/audiogames-reflector
