To provide a method and a device for extending a compressed audio signal at high speed, and electronic equipment.

In this method for extending the compressed audio signal, a sample extracted from the audio signal is formed into frame by utilizing predictive coding for extension, this sample is compared with a previous sample and a prediction error is detected. Frames frn, frn-1 and frn-2 of the compressed audio signal are stored and on the basis of the stored frames frn, frn-1 and frn-2, the compressed audio signal is extended by utilizing predictive decoding. Besides, at least one of memory pointers P1 and P2 is used for showing the storage positions of the frames frn, frn-1 and frn-2, and this is used for instructing the storage position of the frame frn-1 preceding to the frame frn under processing at present and instructing the storage position of the frame frn-2 preceding to the previous frame frn-1.