When encoding 44.1 kHz audio to a 128kbit/sec mp3, lame by default cuts
off the high end with a transition band of 15115 Hz - 15648 Hz. (32 kHz
audio to 128kbit/sec mp3 has a slightly lower cut-off by default with a
transition band of 14065 Hz - 14452 Hz.)
If someone was encoding 128kbit/sec mp3s from their soundcard (ie from an
analog source like a mixer instead of a CD rip) and was okay with these
default low pass filter frequencies, should they probably use 32kHz as the
sample rate instead of 44.1 kHz? Seems to me about the same frequencies
would be represented ( ~20 Hz up to around 15 or 16 kHz) but the
compression ratio wouldn't be as big for 32kHz. Would that lower
compression ratio result in better sound quality? Or does more samples a
second result in better sound quality at all frequencies?
Are there soundcards out there that don't support playback at 32kHz but do
support 44.1kHz? Shoutcast's dsp plugin text file says:
[search for] the LARGEST number in the Hz column that your speed can
support that is a MULTIPLE OF 11 --
choosing 44, 22, or 11 kHz will assure that the majority of listeners
can hear your broadcast.
Thanks,
Sidi
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
