AUDIO CODING ON THE BASE OF PICTURES WITH ADDITIONAL FILTER BANK TO ATTACH SPECTRAL INTERFERENCE IN THE LIMITS OF THE PICTURES.

摘要

A method for signal processing, comprising: receiving an input signal comprising a sequence of input frames, a respective input frame comprising a plurality of encoded blocks that carry encoded audio information, in which a first input frame it has a final coded block that follows a first intermediate coded block within the first input box and a second input box that immediately follows the first input box has a start coded block that precedes a second intermediate coded box within the second input box, in which the coded start block conducts audio information filtered by a first group of analysis filters, the final coded block conducts audio information filtered by a third group of analysis filters, and the first and second intermediate coded blocks conduct audio information filtered by a second group of filters d and analysis; receiving a control signal (323) that identifies a boundary between said first input box and said second input box; generating a first synthesis block of synthesized sample blocks that immediately precedes a second synthesis block of synthesized sample blocks, the first synthesis table having a synthesized sample block that follows a first intermediate synthesized sample block within the first synthesis table, and the second synthesis table having a synthesized starting sample block that precedes a second intermediate synthesized sample block within the second synthesis table, in which the starting block of synthesized samples is generated by applying a first group of synthesis filters (326) to the starter coded block, the final synthesized sample block is generated by applying a third group of synthesis filters (328) to the final coded block, the first intermediate synthesized sample block is generated by applying a second set of synthesis filters (327) to the first block intermediate coding and the second intermediate synthesized sample block is generated by applying the second group of synthesis filters to the second intermediate coding block, and in that, within a frequency range, in Hz, from approximately 50,000 divided by the number of samples in the synthesized starter sample block up to approximately 150,000 divided by the number of samples in the synthesized starter sample block, the frequency response (341) of the first group of synthesis filters, together with the first group of analysis filters , attenuate the spectral energy more than the frequency response (342) of the second group of synthesis filters together with the second group of analysis filters, and within a frequency range, in Hz, from approximately 50,000 divided by the number of samples in the final synthesized sample block up to approximately 150,000 divided by the number of samples in the final synthesized sample block, the frequency response (341) of the third group of synthesis filters, together with the third group of analysis filters, attenuate the spectral energy more than the frequency response of the second group of synthesis filters together with the second group of analysis filters; and generating an output signal (330) by overlapping adjacent blocks of synthesized samples and adding corresponding overlapping samples within the blocks of overlapping synthesized samples.