74ls248（74ls248）.docVIP

74ls248（74ls248） Http:// The frame of real-time voice transform technology and based on DSP Chen Kai 1, Wang Jia 2 1.2 School of information engineering Beijing University of Posts and Telecommunications, Beijing Haidian District (100876) E-mail:cyre@163.com Abstract: in this method, many voice transform method and audio effects are discussed, and gives the Matlab language description. Integrated voice transformation and a part of the audio effects processing technology, design a kind of widely used in film Real time voice changer multi domain audio and voice security and entertainment, the first is simulated with Matlab program, adjust, finally Using ADIs BF535 implementation. Keywords: voice conversion, audio effects, Matlab, DSP, BF535 CLC number: TP391.4 1. introduction Voice conversion (VC, voice conversion or VT voice transformation) refers to the change of a speaker (source The speaker, source speaker) speech personality characteristics, so that it has another speaker (target speaker, target Speaker) speech personality. [1] Speech has a wide application range, for example: (1) the conversion in the text (TTS, text-to-speech) system application. (2) can be used in film dubbing. (3) speech can be applied to restore the voice, voice channel speaker to help improve the damaged Intelligibility. (4) can be used for personalized voice in secure communication in disguise. (5) as the preprocessing of speech recognition, to reduce the impact of speaker differences. The personalized features of speech can be divided into the following three categories: The segmental feature: describe the features of speech sound. The characteristic parameters including pitch frequency, resonance peak The resonance peak bandwidth, pitch frequency, energy etc.. The suprasegmental features: description of the prosodic features of speech. The main factors include the feature length, pitch The change in frequency (pitch), energy etc.. The language features: including idioms, dialects and acc