3 ms·
Suzanne Ciani created the Xenon pinball soundtrack (1980)
- karmakaze 4y agoPretty neat how they did data compression for voice synthesis. Regarding that comment about a higher frequency female voice needing more data, the MC3177L chip shown uses CVSD modulation[0]. [0] https://en.wikipedia.org/wiki/Continuously_variable_slope_delta_modulation https://en.wikipedia.org/wiki/Continuously_variable_slope_de...
- peteforde 4y agoI was, in fact, wondering about that and thank you for the link. Sadly, I'm not connecting the dots; can you walk me down the path a bit? Is the basic idea that the masculine voice was set as a baseline, and higher frequencies would require more data to represent the offset?
- karmakaze 4y agoI think the way it works from the way I read it, is that there's only a single bit of output for each step of the waveform being shaped that can specify up or down. For a female/higher frequency waveform the crests/troughs occur more rapidly so the bits have to be output faster thus needing more bits per unit time. There might be more to it than that as there might also have been something there about variable time steps, but that was the gist of it that I got.