diff options
author | 2015-08-23 17:01:12 -0400 | |
---|---|---|
committer | 2015-08-23 17:01:12 -0400 | |
commit | 882edf70ba2c0e849aa2c3a678776b8b88695d3e (patch) | |
tree | 252518bf9d63ebcb731cc211817ff8fb1c33cb65 /3rdparty/jsoncpp/src/lib_json/json_value.cpp | |
parent | dbf39ba66747a8e0d8fcd061bac95ab76e5c57ac (diff) |
tms51xx: removed target_x variables in favor of reloading from coefficient ROM during the generation loop (as the real device does).
Implemented proper ZPAR and unvoiced ZPAR logic for zeroing parameters during idle and unvoiced frames.
Changed the pitch zeroing logic during frame inhibit to last an entire interpolation period rather than one sample.
Replaced the speech-is-synthesizing state machine with the original patent TALK, TALKD and SPEN bits.
Redid the talk_status logic to use (TALKD || SPEN) as the real device does. [Lord Nightmare]
Diffstat (limited to '3rdparty/jsoncpp/src/lib_json/json_value.cpp')
0 files changed, 0 insertions, 0 deletions