Hi to every reader. I have been given a task to optimally format the long speech recording (170 minutes long - some guy philoshophing about many things in life). The recording was made in studio, so it is studio quality (no noise, reverb and other stuff). I have received it on an DVD disc as one large WAV file (the disc is recorded in "data" mode -UDF). My task is to optimize the recording format and to provide the result in three variations: WAV, MP3 and some other audio compression format which is best suited for speech. So there are three questions: 1) I will definitely strip the recording to one channel (mono). That will reduce the file size by 50%. But I wonder would it be OK to downsample it so 22.2KHz? I know that difference between 22.2KHz and 44.4KHz is noticable with songs, but I am not sure does it matter with speeches? Or if not that, what about 32KHz? Is that also a standard sample rate? Does MP3 support it? 2) What bitrate would be just enough to use for MP3 version? I thought to use 64 kilobits per second since the file is mono. But would even the lower value be fine for speech? 3) The most important question: Which audio compression format is the best for speech? Is there some format which is specialized for voice recordings?
First, I have forgot to write one important information: the WAV file I have received is in 44KHz-16bit-stereo format. I do not understand what do you mean with 24K? 24 kilobits per second? That is quite low, so the sample rate was probably 11KHz or at most 22KHz? All three questions from the first post are still open for other respones. I need as much suggestions as possible.
