Hi All! All the posts in this forum are about the MPEG 1 standard, which is pretty old. I've noticed a few Mp3's around encoded in the MPEG 2 layer 3 standard (you can easily see that with encspot). I tried out lame (3.95) in encoding mpeg2 or 2.5 mp3's files, And read about it a little- mpeg2 is said to provide the same quality at almost half(!) the bit rate. I don't realy get it - why i don't i see great headlines and everybody moving to mpeg2? It's a smaller file with better quality, isn't it? Why no one is using the mpeg2 settings? what are the recommended settings for mpeg 2? I've read also that mp3's encoded under the mpeg2 standard are encoded at sampling rate of 22.5 khz, and mpeg2.5 at 11.25 (!) khz. Isn't that audible? if not how come they didn't reduce sampling rate in the first place? Wow.. so many question.... In other words - FAQ about the mpeg2 capability of LAME and tooLAME ... Is it really the next generation of encoding? Assaf
MPEG 2/2.5 are just extensions to be able to encode mp3 (MPEG x layer 3) at lower sampling rates. The maximum sound frequency that can be stored is 1/2 * sampling frequency (Nyquist limit). Because of that a lower sampling frequency is equal to lowpassing the signal. Depending on your ears and the type of music you encode, a lowpass starts to become audible at ~ 16kHz. If you e.g. encode at 16kHz *sampling* frequency, the resulting 8kHz lowpass will make everything sound like played back through a telephone. For encoding speech this can be acceptable, though to get very small files with acceptable quality. Same quality @ half bitrate is simply not true. Example: Resampling from 44.1 to 32 kHz (= using MPEG 2 instead of 1) makes sense approx. in the 80-100kbps range. The quality gain (compared to using 44.1kHz + 15-16kHz lowpass) will be equal to a bitrate increased by ~ 5-10kbps.
