Losseless Compression: compress file, decompress file. Compare
original file to decompressed file and you find they're identical.

Lossy: Compress, decompress and compare them and you'll have 2
different files.

Then, according to your definition, CDA is lossy. If you compare the
two
waveforms, you'll have different results. *

MZ, by my definition, it's not lossy.

Your definition is stated above. You base it on whether or not you have
"different files". Well, considering that the discussion is based on the
conversion from original program (not CD) to CDA and mp3 formats, in both
cases you'll have "different files" from the original program. This is due
simply to quantization and anti-alias filtering, among other things. I'm
not saying these differences are significant (in fact, it's my opinion that
they're not). Rather, I'm pointing out that there are indeed differences.
This is important because I think it illustrates why your definition is
faulty and perhaps even misleading.

I've done this test. Have you?
Obviously not.

No, you clearly haven't.

Do yourself a favor. Record something on your PC, or be lazy and
extract a song from a CD, it doesn't matter.

Burn that wav file to a CD.

You're missing the point, genius. I suggest you go back and reread the
discussion. The point was the conversion from program to CDA requires a
change in waveform. Read it again.


Extract it from the CD using EAC or CDEX or EZCDDA Extractor (all are
either free or have free trials of full versions).

Compare the files. They are identical.

Of course.

You CANNOT extract a soudn file, compress it with MP3, OGG Vorb, AAC or
WMV and decompress it to the original wave.

Yeah. So what's your point? Nobody said that the mp3 format wasn't lossy.