Greetings.
I'm a new in audio signal processing, and I have just gotten some new information of all kinds of audio formats.

1.AAC is a lossy compression, isn't it?The quality can't be better than FLAC, can it?
Is it better than mp3s or ogg vorbis?
What's the difference between AAC and AC3?

2.I have seen the book list previously.
the book list:http://www.hydrogenaudio.org/forums/index.php?showtopic=7645&hl=books
Just like I said, I am a new, and actually, I know nothing in audio formats, so I'm searching for a book which is suitable for me: a full formats guide for new.
I am in Taiwan, and there aren't many computer English books here; don't even mention to audio-related books.
Due to this, would you tell me which book is suitable for me needs?

I hope my English is still readable, and wasn't causing you problems.

Yes it is.No, the quality from a lossy codec can not be equal or better than a lossless codec if its encoded from the same source.The Multiformat 128kbps public Listening Test ranks AAC files encoded with iTunes as #3. Behind Vorbis aoTuV and MusePack, but in front of Lame.
Be aware that this might not be the same at other bitrates. Also remember that there are other encoders for MP3 and AAC out there, that might perform differently.
In short terms AAC is a codec for stereo signal, when AC3 is 5 channel (Dolby Digital 5.1). They are both lossy codecs though.
No problems!

AAC can perfectly handle 5.1

AC-3 can perfectly handle stereo as well.

@Gabriel & kjoonlee. Ooops, I did not know that.

actually, itunes is tied at number 2 with lame, and MPC/outuv(vorbis) are at #1.

They are 2 different lossy audio codecs.
(This was not mentioned by the others)

I'm nitpicking, but it certainly can be equal...

Then we have different meanings of quality loss!
It is not called lossy vs. lossless for a random reason, right?

Of cource there are cases where you dont hear any difference, but that does not mean its no quality loss!?
If you code music and lose some data, meaning that it can not be decoded to its original state, I would say it got a quality loss!
You mean differently I guess?

Garf is just nitpicking:
For some specific signals, the result of lossy and lossless encoders can be equal. Silence is an obvious example...

Thanks for your replys about AAC and AC3.

But I still have a problem here:books for a new.

Actually, I don't know what book is suitable for me as a new.
I should study acoustics first or direct study for formats of musics?

I really don't have an idea.
Would you give me some opions?

(I've seen "interesting books" article, but I still don't know which one to choose.)

Yes, the one in iTunes 4.5 is overall much better than the 4.2 codec that was tested.

No one has done a proper ABX of it though

WTF! If noone did a proper ABX, how the heck can you claim 4.5 is "overall much better"?

Bleh.

hence the " "

Actually 4.5 is considered worse or at least no better by many, which is why it was not used.

I depends on your purpose. Do you want to encode some CDs of yours, or do you want to study the psychoacoustics theory ?

There are some infos in the hydrogenaudio knowledge base (link on the portal), but it is down for the time being.
There are some in-depht informations at http://www.mp3-tech.org/ and at http://www.david.robinson.org/papers.html

Just for beginning you can simply search for standart/codec names(aka AC3, MP3, AAC, Vorbis, OGG and so on) in Wikipedia.
It can give very general, but still useful knowledge...
http://en.wikipedia.org/wiki/Main_Page

Actually, I want to study encoding first.
I have some plans to make good encoded files from cds and dvds(dvd-a or audio from dvd-v).

So i think I'll start with knowing how encoding works and comparisons between all kinds of audio formats, and then start studying in the psychoacoustics theory.

Are there books writing detailed description of audio formats and comparisons between those formats?
Or the information is just spread all over the internet?

thx a lot

I don't think studying encoding will significantly improve your encoding ability (though it would obviously help you to *write* an encoder). Many smart people have worked on a range of encoders, and folks such as rjamorim have conducted tests to show which is 'better'. So pick the top two or three and give them a listen for yourself.

Cheers, Paul