Methods, systems, and apparatus, including computer program products, for restoring audio signals. A data sequence of samples representing an audio signal is received. Multiple filter coefficients are defined for a filter, and a current sample in the data sequence is selected to be processed. The filter coefficients are updated based on a previous sample preceding the current sample in the data sequence and a filtered value determined by the filter for the previous sample. A filtered value for the current sample is determined using the filter with the updated filter coefficients. The filtered value of the current sample is used to determine whether the current sample has been corrupted by impulsive noise, for example, a crackle.