Luis 
Hi, I am trying to work with audio in a Mini2440 without success. Using record tools in Qtopia, I am able to record my voice without problems, but it doesn't work in my program. I am using last linux image from FriendlyARM. Program records in 16 kHz (sampling) 16 bits. Is there a problem with this sampling rate. It is a speech recognition program that can work with OSS and ALSA. In a previous kernel, audio was distorced, low and high frequency was the same. Frequency in 3 kHz, appears in 5 kHz, exactly the same. Same for 2kHz that will reappear in 6 KHz, and so on. Any ideas? Am I doing something wrong with a configuration file? Luis