| 摘要 |
The invention relates to the coding of speech at a variable bit rate, whereby the bit rates can vary from frame to frame, and more specifically to the methods and filters used for improving the quality of the decoded speech. In the solution according to the invention the weighting factors of the postfilter are not adapted on the basis of the momentary bit rate or the bit rate used in the coding of each frame, but the weighting factors are adapted according to the average bit rate calculated on the basis of a predetermined length of time. In addition to this, the weighting factors of the postfilter are also adjusted on the basis of whether the frame in question contains a voiced speech signal, an unvoiced speech signal or background noise. At frames containing an unvoiced speech signal or background noise, postfiltering is weakened so as to avoid the distortion of the signal tone because the postfiltering is adapted to a voiced signal. The weighting factors of the postfilter can also be adapted on the basis of the error rate or other parameter describing the quality of the signal or the data transfer channel. For example, postfiltering can conveniently be adjusted so that when the channel error rate and the amount of coding error increase, postfiltering is increased, whereby the effect of data transfer errors on the decoded speech signal is reduced and the tolerance of the system with regard to data transfer errors increases. |
