Ok, I need to stop obsessing about all this old speech stuff.
Remember the Echo sample I put up? Well, it came from a whole file full of early synth samples. Some of these, I can’t even understand. And some are amazingly clear. But they must have been made to do what they did by stringing together pre-recorded words. I once played with a program like that. It could only say words out of a lexicon it had been given. What it could say, it said very well. But if it didn’t have a word in its list, it just didn’t try. Rule-based synthesis? What’s that? I was so sad when I figured that out. I wrote a big long sentence, and all it said was “was the.”
So there’s a history lesson in speech. I’m glad I found this. I think I’ll stop digging up weird robotic singing and talking stuff…for now.