Text-to-Speech program progress video. Earlier this morning I finished the first program selectable for the command line interface. The program asks you for the locations of a voice file and a number of map files, then uses the map to tokenize an input string. The tokens are used as indices to the array of voice samples contained in the voice file, and the samples at the tokens' indices are extracted from the voice to a file. The user is then asked where they want the audio file saved, and the program exits upon receiving an answer. The two maps used are: a to 0, s to 1, z to 2, ð to 3; ab to 3, st to 2, zx to 1, ðd to 0. The input is thus converted to samples 3, 1, 0, and 1.
Cool. It's a good beginning. Watching on my iPhone, YouTube cuts it off at the end instead of sitting there so I have to watch on compy.
morpheusathame 1 year ago