An initial version of PHISYX needs high CPU load to calculate complicated computing, and to do real-time data processing. In order to improve in many aspects of PHISYX (not only for speeding up its data processing time), several trials have been investigated. The followings are those outcomes.
PHISYX2 processes audio data seven times as fast as its initial version. Its algorithm had been re-built, so that all information processing has been more optimized and can be possibly processed in real time.
PHISYX2 controls formant frequency which is characteristics of person's voice. The formant frequency can be moved as required, so that original formant can be kept, ensuring that the data is always heard in its original form.
The first PHISYX version produces unreal human voices after the data are processed, because formant frequency is accurately scaled in a same way as other frequency elements are done. PHISYX has been successfully upgraded in this point.
The recent PHISYX version allows data processing to be more accurate. The re-constructed algorithm makes phase synchronizing optimized in PHISYX2, so that sound phases can precisely keep their originality.
 |
 |
||||||||||||
|
|||||||||||||
