A hearing apparatus that processes an audio signal by adaptive filtering operates efficiently. The hearing apparatus has a first stage transformation unit and at least two second stage transformation units. Each of the second stage transformation units splits the signal of one of the first stage channels through transformation onto sub-channels of the channel. A filter unit in each channel that has a second stage transformation unit filters the signal of the channel depending on weighting factors that are determined in each case for the sub-channels. The filter units are configured to exchange weighting factors among themselves, which means that only some of the weighting factors actually need to be calculated.