After collecting the sound of the electronic pipe organ, the audio needs to be denoised, and the imported audio file needs to be parsed to obtain the audio data information. Based on the timbre parameters obtained in the timbre parameter editing interface, calculate the frequency domain information of the synthesized timbre of each frame, and then perform the inverse Fourier transform to obtain the time domain waveform of each frame connect the time domain waveforms of different frames by the cross-average method to obtain the time-domain waveform of the synthesized tone (that is, the audio data of the synthesized tone). The harmonic peak method based on improved confidence is used to identify the pitch, obtain the fundamental tone of the tone, and calculate its multiplier. The audio time domain information (that is, the audio data obtained by file analysis) is framed and windowed, and fast Fourier transform (FFT) is performed on each frame to obtain the frequency domain information of each frame. The timbre synthesis module realizes the timbre synthesis of the electronic pipe organ according to the current timbre parameters. This research mainly discusses the construction of a database of electronic pipe organ tone recognition based on artificial intelligence. In the past few decades, artificial intelligence technology has experienced rapid development, and its application in modern industrial systems has grown rapidly.
0 Comments
Leave a Reply. |