If you halve the samplerate from 44.1khz to 22.05khz, you of course get lessQuote:>I don't know the exact corelation, but kbps is kilo-bits per second. And
khz
>is the playback frequency. I think at lower kbps rates you have to use
lower
>playback rates.
>Can someone clear up the relationship?
raw input material you need to encode into the mp3 bitstream. But on the
other hand it sounds like crap, because you effectively cut out all higher
frequencies than 11 khz straight out. It is much better to let the encoder
work with the entire spectrum using 44.1khz samplerate, and then let the
encoder choose which frequencies to throw away. If you then lower the
bit-speed the encoder can work with, it needs to cut out more frequencies
and introduce more noise, but chances are high that it does this in a way
that sounds better overall than just cutting the entire frequency spectrum.
Also, there is more hearable information in the lower half of the spectrum,
so cutting the samplerate in half doesn't mean you can halve the bit-rate
and get something which upon decoding sounds as well like the original 22
khz sampling as a decode of a 44 khz sampling with twice the bitrate sounds
(if you get my point :). I've heard some 56 kbit/s 22 khz mp3's and they
sound like shit.
You should stay with 128 kbit/s for a stereo 44.1khz signal, or 160kbit/s if
you want it to sound really good.
Or, more specifically to answer your question, if you really need to lower
the bitrate, try making two mp3's at the same bitrate but with different
sample-rates and listen for yourself which sound better. In the lower
sample-rate case you will probably hear more "narrow band radio" distortions
and in the higher sample-rate case you'll hear more "psychoacoustical noise
shaping" distortions. It's just a matter of which sound best to you :)
/Bjorn