question is simple.

if compatibility with software and hardware players doesn't matter, but only the best quality/size, what should I choose?

gennerally AAC or MPC :-)

the choices were vorbis and lame!!!

Vorbis, and never look back.

well, what do you want? For transparency, vorbis 1.01 is not the way to go. Garf's GT3b1 would be preferable. LAME 3.90.3 with --alt-preset standard is very good at transparency, with probably more man hours of tuning and testing going into that than GT3b1.

If you want the best quality at lower bitrates, Ogg 1.01 is the best choice.

IMO, Vorbis is the future, but we still have to wait a long long time (at least GT3b1 added).

Now, my NEX II only play MP3 ...

To 1.0.1? I'm afraid not, IIRC Garf's tunings will only be incorporated in 1.1 onwards.

I've tried GT3b1 at -q6 with some of my CDs. And generally the oggs produced were very slightly bigger than aps mp3s.

make both for afew files listen to them and ...it's up to you

...but do it blindly to avoid the placebo!

i have encoded the same song (Linkin Park - Numb) with Vorbis 1.01 (q6) and Lame 3.93.1 (my own compile) (--preset standard) and my opinion is that vorbis sounds better.

I compiled vorbis myself (optimized for athlon xp / sse), source code acquired from cvs.

Please support your opinion with fact's.

Read this, it is a part of the forum rules.

You should use Lame 3.90.3 --alt-preset standard

sorry for that, i'll read the rules.

i think that vorbis sounds better because there are some parts of music in the background that can't be heard with mp3, and the vorbis sounds somehow clearer.

and for Latexxx: --preset standard in 3.93.1 is the same as --alt-preset standard in older versions. the differences between 3.93.1 and 3.90.3 are mainly in speed.

Actually, that's not correct either. 3.90.3 is the recommended version here, for quite a few reasons, and none of them are speed related.

i didn't say that 3.90.3 is recommended because of the "speed" reasons.

the reason that 3.90.3 is recommended is that it has been thoroughly tested, and 3.93.1 was not.

check the lame history log, and you'll see what has changed since 3.90.3.

i use 3.93.1 because it is faster than 3.90.3.

...anyway if ogg sounds better to you , end of story .

Oh, my god !

If you're looking for quality/size q6 is a bit wrong: for optimal filesize/quality go Vorbis q2 or q3 (90 or 112 kbps).

If you're looking for quality, don't waste your time and go lossless (FLAC or MA)

It is not just because of the quality testing, 3.90.3 also adds -Z to APS by default.

Edit: http://www.hydrogenaudio.org/forums/index....=ST&f=16&t=9301

Do you suggest that no nuances existes between 100 kbps encodings and lossless ones? Are mpc, --alt-preset, hybrid encoders... totally useless? Or do you simply suggest that vorbis -b6 or mpc -q7 aren't good enough?
I don't understand.

stipe,

In order for you to prove your claim that codec a at setting x sounds better than codec b at setting y you need to use an abx tool (which is explained in the link I provided). Try a google search for winabx to find said tool and learn how to use that.

Once you've satisfied yourself that you can distinguish between codec a and codec b using this tool please let us know your findings. Explain what the differences were and if possible provide a sample.

Any feedback is appreciated.

Let me rephrase it a little more correctly:

"If you're looking for quality, don't waste your time and HD space and go Musepack (-quality standard -xlevel) or mp3 (--alt-preset standard), or really anything at that level of transparency, unless you have proven by ABX that you have excellent hearing and have the monetary and technical capacities to procure and operate a 250 GB HDD or larger, and even a RAID array, in which case the phrase in the quote box is meant for you."

Just a question
If the guy used ABX and it showed that mp3 is better than ogg.. but he thinks ogg sounds better what should he do ?

i used abx, and i was able to tell the difference between the mp3 and ogg almost every time. ogg sounded better.

Then it seems, you have your answer.

ABXed against what? Each other? IMHO, it doesn't matter if you can tell the difference between them. Which one do you feel is closer to the original source?

sorry, i said it wrong.

i ABXed them against the source wav ripped from cd.

i think that vorbis is closer to the source.

ABX (in itself) doesn't tell you which sounds better, it only reveals variances between two test samples, if there are any real variances.

Differentiation is objective. Once a difference is established, quality determination is subjective. To do both, I'd recommend using ABC/HR to first ABX each encoded sample from the source WAV, then use the rating system to describe which sample sounded better and to describe why.

Even if you can establish a difference, one may sound "better" than the other, but not always. Sometimes they're just "different".

I realised my mistake in not mentioning this shortly after my last post, but I get the impression I'm waisting my time explaining anything here. Especially after seeing ABX claims after just 15 odd minutes with these hq settings and again, no kind of proof to back up any claims.

i abxed the samples before i even posted here. i didn't abx it in 15 minutes.

Exactly my thoughts

There are still noise problems with Vorbis even at high bitrate - mp3 doesn't have this problem. Therefore, some quiet and "simple" tracks might be transparent with lame ~160 kbps and ABXable with Vobis at > 250 kbps. This may affect classical mostly.

One reason why Vorbis wins over MP3 ( non-quality related, so ABX not needed ).

A much improved, proper, modern tagging system that is universally supported across players. ID3v1, while widespread, has its limitations, ID3v2 has its problems.. But Vorbis comment tags rock

You are always free to use APEv2 tags on MP3s, which are at least as flexible as Vorbis Comments.
Vorbis tends to bloat the bitrate less extremely on punkrock/hardcore music and noise problems are not significant/notable on this music.

dev0

APEv2 tags, good as they are, aren't a universal standard. My hardware player won't read them ( the main reason I'm using MP3 now anyway, until iRiver finally puts out the Vorbis firmware ), and not all players support them ( Winamp? Windows Media Player? QCD? )

One area I am quite certain vorbis tends to win over mp3 at similar bitrates is preecho. If you've got lots of electronic music, or music with sharp percussion, it could be a pretty big win to go with vorbis.

yes, I do have a lot of electronic music, and I think that vorbis sounds better with them.