I'm looking into AAC, because it has more potential than MP3.

My baseline is the latest LAME 3.97 v2 etc etc

How does AAC compare? Can smaller size with very comparable quality be achieved? What's the best encoding method? How does itunes compare to NERO?

1. Use search.
2. See the latest listening tests. Also, see guruboolez personal tests comparing itunes and nero (showed very similiar perfomance with itunes to be slightly better).

Anyway, the practice tell us that all modern codecs can not be distinguished from original on bitrates about 128 Kbit.
For example, 5 min ago I tried to ABX Roxette Tourism (1992) first track encoded by nero aacenc at 0.34 (resulting at bitrate about 100KBit) and failed.

As for best encoding method, it is a matter of personal preference, I think. I use foobar and still happy.

okay, just did a search. Of course, summaries would be nice instead of wading through all of that.

So My quick questions are

What AAC settings are most comparable to LAME mp3 3.97 in quality?

How does AAC work in rockbox? (if rockbox has good compatability, I think I'd seriously consider reencoding my entire collection to AAC)

I have a great many cds encoded in monkeys audio. What's the fastest way to encode them in MP3? Is it possible to get foobar to encode in the itunes AAC? Could AAC perhaps achieve, say, similar quality to mp3 v2, with a 20% file size reduction?

hmm, so I just did a test encode. iTunes AAC vbr 192. Stuck it on rockbox. It works, but it'll pause for reloading every 5 seconds or so. Unlistenable. of course, I am considering a nano purchase in the future...

file sizes are comparable between 3.97 v2 and itunes aac vbr 192. I can't really tell a difference between the two, well, except the aac seems a bit more muffled than the mp3. hmm, actually, the mp3 does a better job.

At 192k everything will sound about the same. You need to go down to 128k or so before theres any difference, and even then, the last multiformat listening test was pretty much all tied. I think AAC is probably better at 100k and below, and that somewhere between 100 and 140k the difference vanishes, but I'm just speculating.

Also, Rockbox only supports AAC on the Ipod (and only just barely). I don't recommend AAC if you want to use Rockbox. Ogg will perform as well, and is MUCH better supported (in Rockbox).

In order to play AAC without pauses on a Rockboxed iPod, you'll have to disable the peak meter, and maybe some other tweaks.

Also in lieu of what what said, the Rockbox developers I conversed with claim Vorbis decodes faster then MP3 on iPod targets.

OGG looks nice maybe, but it's compatability is limited. It's an iPod world, so the only format that I would consider to replace MP3 is AAC, because so many people have ipods it would be easy to move it back and forth.

Oh yeah, tangent, I heard something about LAME skipping in certain iPods, due to power conservation settings and jumps in bitrate.... has that been fixed?

I don't get it. If I were interested in AAC, I would just use ipod firmware, and if I were interested in Rockbox firmware, I would just use Ogg vorbis.

And for your tangent, yes it's been fixed. I saved you a two minute search, your welcome

Please provide ABX results for this. I realise it's just a passing comment but considering all the tests I've seen done i find it hard to believe that "muffled" would be a compression artifact for AAC at 192kbps.

I agree with loophole. And if you're gonna test anyway you might as well start out ABX-ing AAC@128VBR versus MP3@192VBR. That'll save you a few tests...