Abstract: The present invention may be used to improve the quality of music output from audio systems by simulating the effect of low frequency signals in the human ear. This thus allows listeners to perceive the lower frequencies signals, even though the speakers may be incapable of providing such low frequency outputs. The present invention provides simple methods for processing enhancing bass effect in audio signals. The simple methods also result in the bass enhancement being computationally less intensive. The bass effect enhancement techniques described in a first and second embodiment are based on the response of sine and cosine transfer functions and on the directional independence of low frequency components. The human ear is unable to resolve directions from low frequency components. The bass effect enhancement technique described in a third embodiment is based on response of a exponential transfer function.

Abstract: An MPEG-1 layer 3 audio encoder, including a scalefactor generator for determining first scalefactors for encoding a block of audio data if a temporal masking transient is not detected in said block of audio data; and for selecting the maximum of said scalefactors for encoding said block of audio data it a temporal masking transient is detected in said block of audio data to enable greater compression of said audio data. Increases in quantization error, due to use of the maximum scalefactor are pre-masked or post-masked by the temporal masking transient. In cases where the last portion of a block includes a temporal masking transient that masks the preceding portions of the block, the maximum scalefactor is only used to encode the block if the resulting increase in quantization error is less than 30% of the quantization error for the block.