Hello, First of all i would like to say a big thanks for developers of this excellent product. My question is the following: I need to read raw audio data from file and remove from it noise, echo and perform some another stuff. Particularly to remove noise I'm using the "testdenoise" sample as a template. As you know it reads data from stdin. I read it from a file by chunks. Should I make speex encoding of this input data at first, or I can just clear the noise on my stream? If I should, than I can assume that after noise removal the "cleared" data needs to be deencoded, right? Thanks in advance for any help.