An analysis-by-synthesis stage and additional noiseless compression (Huffman coding) tools are employed to implement the quantization and coding process in a perceptual audio transform coding system. The objective of the process is to keep the quantization noise below the masked threshold, given an available number of bits. An iterative method with two nested loops realizes the control of distortion and bit rate for the process. Preliminary analysis indicates that in some instances more than the required amount of bits, estimated by the perceptual entropy (PE), is allocated to subbands when these additional bits could be re-allocated to regions of a higher need. Moreover, quantization at a fixed bit rate neglected the bits not used in the coding process. A proposed third iteration loop addresses these two inadequacies by dynamically allocating bits according to the PE. Results indicate that there are improvements in the audio quality, particularly at low bit rates, but at a higher computational cost. Moreover with the proposed method, perceptually transparent coding is achieved at an estimated bit rate of 1.797 bits/sample-a significant improvement over past estimates of 2.1 bits/sample.:
Click to purchase paper as a non-member or you can login as an AES member to see more options.
No AES members have commented on this paper yet.
To be notified of new comments on this paper you can
subscribe to this RSS feed.
Forum users should login to see additional options.
If you are not yet an AES member and have something important to say about this paper then we urge you to join the AES today and make your voice heard. You can join online today by clicking here.
