In earlier times, text-to-speech, like the cutting edge, was very imperfect. When you entered a word, it would pretty much read it, as you pronounced it ... in a monotone. Often the result would be very funny. Currently, Text-to-Speech is too smart to laugh so that you can laugh.
As a personal project, I would like to create an application that can bring this old style of text into speech, at least as a toy. In .Net, I have for me both System.Speech.dllCOM objects SpeechLib. (Microsoft Speech Object Library) Both seem to use the OS built into Text-to-Speech, which is again too smart. Is there a way to configure them to disable everything that makes it intelligent?
I tried several different "SayAs" parameters, I tried to set the culture to invariant (exception!), And now I'm looking at SSML. Looks like I'll have to find the old technology, but I don’t even know where to start.
As an example of the chaos that I hope to see, there are a few Moonbase Alpha here for you: http://www.youtube.com/watch?v=Hv6RbEOlqRo (Make sure you wear headphones!)
Con flab these newfangled text-to-phoneme converters, and normalizers, and tubeless phones, and ...
source
share