This paper presents a method to reduce area and timing overhead due to the implementation of standard single symbol correcting codes to provide ML Flash memories with error correction ability. In particular, the proposed method is based on the manipulation of the parity check matrix which defines a code, which allows to minimize the matrix weight and the maximum row weight. Furthermore, we will show that a minimal increase in the redundancy, with respect to the standard case, allows a further considerable reduction of the impact on the memory access time, as well as on the area overhed due to the error correction circuitry.