Hi, I think I am the one who is responsible for the byte switch problem. Indeed, I am using a big-endian machine. However, I thought that the PNM standard specifies the byte order. Please advise me if this is wrong. Best regards, Ulrich Deiters