Hello, 

By this video (Vidéo) & after many hours finally, i can record my voice with SFML, i change the Frequency Sampling Fs = 8000 Hz (not the default 44100 Hz), so like this i have 8000 samples/ second, but my work needs buffers of 80 samples so every 0.01 second i should read data (store it in a vector) and apply my algorithm. Something like this:

- start record 

- Every 0.01 second (read & store 80 Samples) & apply the algo

- end record 

Thanks in advance.