In the first 128 kbps multi-format test we got a big statistical tie between AAC, MPC, Vorbis and WMA Pro, with MPC and AAC a little to the front, Vorbis and WMA Pro lagging (not significantly) a little behind, and Lame MP3 in clear second (or fifth?) place.
However, we have 4 new or updated competitors for the imminent test, seeking to oust MPC 1.14 --radio as the highest-scorer. QuickTime's AAC, which has been updated since last test, still scores the highest among the top AAC encoders. Vorbis has been showed to have a new champion among its in-house tweaked versions, aoTuVb2, noticeably outperforming the official Vorbis in the preliminaries, which still had managed to tie first place in the previous test. Lame 3.96, on its part, refuses to stay behind and sports an improved VBR setting of superior quality, according to those who have compared it to the classic 3.90.3 ABR. And let's not forget the newest challenger, Sony's Atrac3, which hasn't been as thoroughly tested as the other competitors and could give us a surprise. WMA Std 9 will probably show as our lower anchor, but we all know what Xing had to say about those prejudiced expectations. Could version 9 put WMA's terrible reputation into question?

Which do you think will be the winner(s), tiers and loser(s) of this test? This is pure SPECULATION, so feel free to cheer and show your bias; we shall see whose team is best in the end . Objective assessments of evidence are encouraged and expected, since the one(s) who guess(es) the correct and detailed final positions of the codecs are going to win a virtual cookie from the organizer of the test himself! (I haven't exactly asked him about it yet, but I don't think he will object... ) So, speak up!

Let's get ready to rumble!!

mostly like this:

[grouping]/[order]

1/1 4.55 (-0.00) musepack
1/2 4.45 (-0.10) itunes
1/3 4.40 (-0.15) vorbis
2/4 4.25 (-0.30) atrac
2/5 4.20 (-0.35) wma
3/6 3.95 (-0.60) lame

both musepack and itunes haven't changed since last test, so 1/1 and 1/2 are fairly safe bets...

i think vorbis will outperform wma this time around (aotuv improvements) and even catch up to itunes some.

atrac is completely unknown to me... so i'm just assuming it's rank. i think it will be mostly equal to wma for completely random reasons...

lame will do much better, but an obvious low rank (so it might tie for last rank or have an obvious last rank, i dunno...)

is that enough speculation? i hope so ;P


later

edit1: made the rankings actually readable...
edit2+3: finalized my rankings due to roberto's whining (below):

We gotta set up a prize to the one that guesses closer to the final results.



Well, if you guys really want to be serious about this...

"or" is not fair. Just post what place each codec will end at, in your guess.

Also, it would be more interesting of the betters guessed how what will be each codec's final score. Say, "Codec A will get 4.5 points, Codec B will get 3.2, Codec C will get 3.8..."

1. MPC (4.1)
2. AAC (3.8)
3. Vrobis (3.4)
4. ATRAC (3.2) (i'm going out on a limb here!)
5. WMA (3.1)
6. MP3 (2.9)
!

Dologan, can I buy that red teddy bear?

Keep in mind that WMA will be in its standard flavour here, not the pro that was used in the previous test, so it's reasonable to expect that it will have a worse performance.


I'm afraid you'd have to ask Frank Klemm, since that's the bear that shows in one of the pictures when the PC donation was made.

Here's my own prediction:

- MPC, AAC and Vorbis statistically tied at the first place at ~4.3 +- 0.2. MPC and AAC really close together at the front, with Vorbis a little behind.
- LAME at second (fourth) place at 3.8 +- 0.2, followed by a tie between Atrac3 and WMA at 3.2+-0.3

All right, if so, then:

1. AAC 4.2
2. MPC 4.1
3. LAME 3.9
4. Vorbis 3.7
5. WMA 3.3
6. ATRAC 3.1

Let it be...

1. Vorbis
2. MPC
3. AAC
4. LAME
5. ATRAC
6. WMA


OTOH going by the bitrate table rjmorim put up i'd say the foll
1. MPC
2. Vorbis
3. LAME / AAC
4. ATRAC
5. WMA

I've no idea of ATRAC's performance, but then again i dont like WMA much

Heh, funny.
By score:
1. MPC (4.2)
2. AAC (4.0)
3. Vorbis (3.7)
5. WMA (3.2)
4. ATRAC (3.1)
6. MP3 (2.8)

- MPC, AAC and Vorbis statistically tied.
Note, that if nothing changed, then mpc will have a bitrate ~13% greater than others so by logic it is out of comparison.
I'm unsure about atrac. It was cheaper than mp3 when I tried v2 (i think) but I unsure about current perfomance...

EDIT: small corrections