I've noticed many are interested in how wma9 performs..
I wrote earlier that on 4 samples at 128kbps, I couldn't
really differentiate vorbis, lame, wma9, and aac (although I could at times differentiate from original).

I don't like doing tests all that much.. but I performed
a very quick test now between ogg vorbis 1.0 and
wma9 at 64kbps.. the results surprised me.. the testing
was done on 5 randomly picked samples that ff123 used
for his big 64kbps test a while back.

blackwater
wma9 kicks vorbis' ass..
Both handle the first half well, but when
the drums and guitar kick in in the second,
vorbis starts squeeling really bad. Wma9,
I think, more than acceptable for casual
hearing on portable or whatever.

Beautyslept
there were things that made me think maybe
I would differ on more thorough testing, but
sounded quite the same on this quick test.. not
quite the same as the original, but more than
good enough for portable use and other casual
hearing.

MidnightVoyage
this is interesting.. as this is no thorough
test and I'm just listening for the most obvious
differences.. well.. the hissing at the beginning
is sharpened quite much by vorbis.. wma9 is much
better there.. otherwise they don't sound the same
as the original, but i'm not really testing thorough
enough to be able to tell if i can differentiate
wma9 and vorbis on the other stuff.. but they sound
same-ish.. but the hissing sounds way different, and
well, it is enough that regardless of what other
differences might be here wma9 kicks vorbis' ass on
this one.

lisztbminor
boring piano.. it sounded like i perhaps could've
differentiated the codecs if i had done some more
thorough testing.. but, like.. they sounded the
same-ish, and well, i don't care more than that.
Good enough for casual hearing.. well actually no
it's darn boring, but.. you get my point.

waiting
wow.. like.. there's so much stuff going on in the
background of the vorbis.. swooshing, garbling.. well
i don't have a clue what to call it all.. but it's really
bad.. now i didn't really bother to compare the wma that
much to the original.. it could've been a little poor too,
but compared to the vorbis, it sounded like heaven.. blah you get my point.

Well.. I did participate in ff123's 64kbps test.. there, wma8 and aac were far behind the winners, for me,
ogg vorbis 1 and mp3pro.. I was most impressed by
vorbis, I truly was..

Now, the testing I did back then was much more thourough. But this test here.. well.. on 2 of the
samples they were same-ish, on one vorbis F*ed
up the hissing in the beginning, and on the two
others wma9 majorly kicked vorbis' ass. Hmm..
maybe it is luck, poor testing, whatever.. but to
me it seems wma9 really has improved quite a
lot.. in this test it was clearly quite superior to
vorbis.. i don't know what would happen if i tested
on 10 samples instead of 5.. but.. i am seriously
starting to think that wma9 is the best format
for streaming at 64kbps right now, probably even
better than mp3pro.. because vorbis and mp3pro
scored very much alike for me back on ff123's
test.

I am very glad I did this test.. now I'm really
convinced that wma9 surely must be superior
to mp3, at the very least (i recently chose it
for portable use). So maybe it's faulty logic,
but since I cannot differentiate at 128/160 this
is all i can go by.. wma9 might be the best format
for me at 64kbps, and lightyears ahead of mp3 at
that bitrate.. (mp3 at 64, the few times i've tried
it, is HORRIBLE).

Anyway, I would seriously have liked to see the
results of a new 64kbps test.. I realize that probably
ain't happening, but.. I have this feeling that wma9
could actually have won! of course these are only
the results of my quick testing on 5 samples.. but..
I am very impressed how much better wma9 performed
than vorbis.. especially considering vorbis was my
winner back then on the ff123 64kbps test.

What more can I say? I realize many people for many good reasons won't use be using wma regardless of how good the audio quality may be. But many who most likely haven't even tested wma9 diss it for its quality already.. it would be very ironic if they too tested it out and actually found out that the quality is quite decent.

I don't know.. but anyway.. i hope i encouraged more to test out wma9.. i would be very interested in seeing if others find it to be great too at 64kbps.. and you know, considering wma is here and NOW on portables.. it surely WOULD be a good thing if the quality is actually good, right? well, many would probably answer no to that, but.. I try not to mix in politics too much into audio encoding and whatever.

Sorry for my english, as i said before, already have two other native languages to shuffle.

Hmm i just realized i forgot something that might
be very important.. what settings i used for the
encoders.. it is interesting..

i used wma9 at 64 kilobits per second CONSTANT
BIT RATE.. now, vorbis doesn't really seem to have
any constant bit rate?? I used -b 72 -m 72 -M 72..
why 72? well, when I tried those settings with 64,
i kept getting 59 and stuff all the time.. with 72,
I got a lot closer to 64.. which makes it more
fair, right?

Remember.. 64kbps is meant for streaming purposes..
you cannot just simply use -q0 for streaming purposes..

so maybe vorbis performed so bad this time because of the commandline parameters.. ? But.. I can hardly believe that it matters all that much.

Anyway.. is there a cbr mode i am missing in vorbis? or was what I used a cbr mode?? but it doesn't quite sound like cbr if it never gave me 64kbps but 59 and stuff etc when I used -b 64.. or perhaps it calculated for some overhead or whatever.. dunno. All i Know is wma at least had a perfect bitrate of 64.. although there was some formatoverhead of course.

Vorbis is an inherently VBR codec. You can force it to be pseudo-CBR, but only by degrading the sound quality, in some cases quite significantly.

The Vorbis settings you chose seem on first sight to be absolutely horribly bad. Just use -q 0, which is the VBR 'around 64kbit/s' setting, and see what you get. If you want to constrain, then use ABR, with --managed -b 64. Setting an absolute maximum/minimum bitrate is something that is *very* rarely needed.

Yeah John is correct. And you can use VBR for streaming. CBR is not a must. Keep in mind that all players buffer by default. So for easy passages the buffer will contain a quite long segment. As the music gets more complex the bitrate needed gets larger and the buffer holds less time wise. In practice if you have a good pipe your buffer should keep you playing smoothly. If you have a crappy pipe no ammount of buffering or even CBR is gonna keep you playing smoothly. It will be quite a rare thing to have a track that will choke your pipe and buffer if you have a good connection.