hi all
I spent today still working on the mdct based on the fft, and it's now
30% faster than the LAME one, but it results in an overall speedup of 1% in the
full code :) Oh well. at least I learnt some things about optimizing. It's also
a bit easier to read than the other mdct. patch to come soon.
The freshmeat entry for LAME should be updated (mark?)
The docs *are* a little deficient (as one slashdot reader pointed out).
The html docs appear to be "borrowed" from the mp3enc documentation. I suggest
we start again from scratch.
Probably just a detailed "usage" document. For more technical info they
can hit the website.
-m quick description of the adv/disadv of each mode
in terms of quality and speed.
-s allowable sample frequencies. mpeg1/2 layerIII freqs.
(i have mistakenly tried putting mpeg2.5 frequencies in.
anyone wanna try for 2.5 support?)
-b some detailed examples of the VBR mode would be good, as
well as some "ballpark" indicators of equivalent quality
for different settings. See the XingEncoder documentation
linked on my website under "vbr tests" for an indication
of what I mean.
-e does anyone use this?
-f
-k this one is a mystery to me :)
-d is this a "userland" option? or more of a developers thing?
does it have noticeable effect on time/quality?
later
mike
www.cryogen.com/mikecheng
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
