Please forgive me if this is in the wrong forum.

This is a question I ran across last year when Apple first debuted the iTunes Music Store. Steve Jobs stated (I'm paraphrasing something I heard last year) that because they took the original masters of some albums and converted those directly to 128 kbs AAC, it is infact better sounding than the CD.

Now, we all know that Steve was greatly fudging by stating that lossy is better than lossless, but my question is regarding a comparison between the two 128 kb's AAC files. Would the original master turned 128 kb's AAC sound better than the file that was created when the consumer CD was ripped to 128 kb's AAC?

Hmm.. how what do YOU think?

Ah, I think I see where this may be going.

So what you are saying is that it is much like any person deciding what they can hear. In other words, the master converted into 128 kb's AAC would infact carry more data (trying to think of a way to say this) than the consumer disk 128 kb's AAC?

I'm basically looking for a more scientific "yes or no" type of answer, not being dependent on hearing ability or perception. The reason why I ask it in more scientific terms is because very few would be able to get to those masters.

Any "increased fidelity" would come from a higher sampling rate or greater bit depth. But since they're selling 44kHz, 16-bit audio, that's impossible.

Another possibility is that they have access to masters that haven't been highly compressed dynamically. But that is pretty unlikely.

They're not selling 16bit audio. Audio data is represented as floats inside lossy compressed bitstreams.

...and that explains the claim of higher quality. They are, supposedly, using 24bit masters for some of the albums in their store. Of course, that doesn't apply to all albums.

Did I act too quick to state that Jobs was incorrect with his statement that those files would sound better than the consumer CD?

Can someone provide testsamples for an ABX-test? I'd like to test it, but I don't have any "master"-recordings.

IMO when the encoded file have 128 bit-depth it can't be better than rip from a CD to a higher quality setting although the master has 24bit.
When we are comparing two 128 AAC's I think that the one from the 24bit master should be better...

128kbps doesn't mean 128bits-depth. It means that overall the file is using 128kbps per second, but doesn't tell anything regarding bit depth.

Most lossy files have simply no bit-depth in the usual meaning.

Thanks for clarification Gabriel.
In the other hand I don't know what does it mean and what to think about it...

Yeah seems like it

I would say it is "unlikely" Steve Jobs is right, but I wouldn't tell it's imposible.
128Kbit is a rather low bitrate, but on some samples it could be transparent, so this samples may sound better, if created directly from 24/48 or 24/96 master...
Would be interesting to try.

But aren't the files still decoded to 16-bit accuracy?

Only if you ask the decoder to decode it to 16bits

Until any listening test is conduced, it's impossible to say if he acted too quick or not.

Knowing Jobs, I would go for the "that's just marketing" approach, at least until we have good proof.