论文部分内容阅读
提出了一种新的主要面向交互式实时通信的低延迟高质量音频编码算法(LDX,low-delay and high-quality audio coding algorithm)。为了降低编解码算法延迟,LDX并未完全沿袭传统的感知音频编码的技术路线,而是采用了相对较短、长度固定的变换窗,从而大幅度地降低了算法延迟。同时,为了在高压缩比下获得高质量的音频,LDX对现有的心理声学模型和立体声编码算法作了改进,运用傅里叶变换和变址离散余弦变换相结合的心理声学分析方法,不仅降低了算法复杂度,同时也提高了心理声学分析的精确度。
A new low-delay and high-quality audio coding algorithm (LDX) is proposed for interactive real-time communication. In order to reduce the codec delay, LDX does not completely follow the traditional perceptual audio coding technology, but uses a relatively short, fixed-length transform window, which greatly reduces the algorithm delay. Meanwhile, in order to obtain high quality audio at high compression ratio, LDX has improved the existing psychoacoustic model and stereo encoding algorithm. The use of psychoacoustic analysis combining Fourier transform and indexed discrete cosine transform not only Reduce the complexity of the algorithm, but also improve the accuracy of psychoacoustic analysis.