Voice changer -- 2

Overview: The Voice changer for Asterisk represents DSP software solution that is able to change voice of a participants in two way voice communication. This is an Asterisk module that implements an asterisk application that can be used from dialplan. It should be able to change human voice of a speakers in a quality way. It should be able to manipulate voice attributes like: chiefly frequency, harmonic structure, intensity... It also must be capable to identify and change formant of human voice. Resultant real-time voice must have the following characteristics: it is clear, it keeps intimacy of a speaker and it is natural. DSP should be able to produce at least 10 different resultant voices from a single speaker. It should be possible to set all parameters that define voice attributes when called from a dialplan. Every channels should have its own DSP component. There is an existing Voice Changer asterisk application which can be used as starting point for this project. This solution yet must be much more sophisticated. Advanced fast pitch-timbre changing algorithm that produces natural and real-time voice output must be implemented. Functional demands: It’s an Asterisk module that implements voice changer asterisk application. Change human voice in quality way. It manipulate voice attributes like: chiefly frequency, harmonic structure, intensity... It also must be capable to identify and change formant of human voice. Resultant real-time voice is clear, it keeps intimacy of a speaker and it is natural. Separate configuration for every channel. Every channel must have its own DSP processor. It must be capable to apply changes for incoming, and outgoing channels separately. Start/stop the DSP with the application. To use voice changer one should call the application with start parameter (i.e. DSPVoiceChanger(<start|stop>, <in|out|both>), from which moment channel’s DSP is started, until stopped, or channel being released. When DSP is stopped, asterisk must not be any different as if there were no DSP module installed. It uses channel variables for controlling DSP at any time. From the moment when application has started, DSP should change voice attributes defined by local variables. If variables are changed at any moment (I.E. by AMI), DSP should respond with correction in changing attributes in time frame of 5 seconds. There can be any number of variables for control, but no more then 10 or 15 per channel in one way audio.