38 lines
1.4 KiB
Plaintext
38 lines
1.4 KiB
Plaintext
|
INTRODUCTION
|
||
|
|
============
|
||
|
|
mp3encode is a package for encoding MPEG layer I, II and III audio
|
||
|
|
files. This is the reference implementation so don't expect
|
||
|
|
spectacular performance.
|
||
|
|
|
||
|
|
Driving encode is pretty simple although it may take some practice to
|
||
|
|
work out an optimal compression rate. AIFF files are the only type of
|
||
|
|
files that seem to work with this version so you may need to install a
|
||
|
|
port like "sox" to convert your file types. I use the "nas" port to
|
||
|
|
record and edit all of my audio files but so long as you have a valid
|
||
|
|
AIFF file it shouldn't matter what you use.
|
||
|
|
|
||
|
|
Although the source code includes a decoder I have decided not to port
|
||
|
|
it because we already have a couple of top class mpeg decoders in the
|
||
|
|
ports collection.
|
||
|
|
|
||
|
|
HINTS & TIPS
|
||
|
|
============
|
||
|
|
Physcoacoustic model 2 is the only supported model in layer III
|
||
|
|
encoding. A bit-rate of 128Kbps is adequate for most forms of music but
|
||
|
|
you may need to turn down the compression to 160Kbps if the music is
|
||
|
|
particularly atmospheric.
|
||
|
|
|
||
|
|
Vocal spoken recordings are best encoded using a 32Kbps bit-rate with
|
||
|
|
layer II (layer III makes you sound like a Dalek under digital water)
|
||
|
|
with a 32kHz sample rate in mono. Using this method I have managed to
|
||
|
|
get about 5 minutes of AM Radio quality audio into around 1 MB.
|
||
|
|
|
||
|
|
MISC
|
||
|
|
====
|
||
|
|
If anyone finds any bugs in this software please let me know and
|
||
|
|
include patches to fix them.
|
||
|
|
|
||
|
|
Share and enjoy,
|
||
|
|
|
||
|
|
Joel...
|