Re-Encoding MP3s

Let's say you burn an MP3 onto a CD, at which point the MP3 is un-compressed into a .wav file. Then, you want to create an MP3 from the CD (what happened to the original MP3? Let's pretend it was eaten by lions). Assuming you know the original bitrate that the mp3 was created at (192 kbps, for instance), when you re-encode it at the same bit rate, are you "losing" anything? Or should you attempt to encode it at a higher bit-rate, and it's just a matter of diminishing returns?

-J.

... Let's pretend it was eaten by lions). ...

Well, then let's pretend it sounds better than it does...   

Are you sure that you un-compressed it when you burned it to a CD? .mp3 files can be burned to a CD as well, and most modern CD players will play them quite nicely.

Let's just put it this way; The quality of the music ain't going to get any  better with all this back-and-forth compressing and un-compressing. It will be what it will be.

-Phil

If the disk is an audio CD, the data will be uncompressed. Many DVD players play both CD and mp3 (and other formats) but few home CD players play anything but CD. Windows Explorer will tell you what you have in an instant.

There is no loss in decoding from mp3 to wav. Every time a wav is encoded to mp3 it goes through the same perceptual encoding process: the bit depth of various parts is reduced still further and more noise is added. However, the result will not necessarily be audible for the first re-encoding. I recommend using the LAME encoder with its V2 preset, which produces mp3, most of the time, that people cannot distinguish from the wav input.

With a computer at hand, simply trying the re-encoding, and listening to the results, is easy. Since it is so easy to fool yourself into hearing whatever you believe you might hear, you might want to acquaint yourself with bind ABX testing, also very easy when the audio is on computer. PCABX and WinABX are two of the freeware tools for performing the tests.

I don't know, if I were stlf I think I'd be looking through whatever lines the lions' cages for droppings...  Andy's answer is excellent and perfectly reasonable IMHO but double-compression is such a dirty term around here (justifiably) that we should just say it for the sake of saying it: double-compression should be avoided at all costs.

While a bit tedious to do, ABX tests are the way to get answers. The software is freeware and easy enough to use.

I understand the basic argument against re-encoding lossy compressing audio, but I’ve had occasion to do a fair amount of it because I wanted some kind of editing of the data. This has all been with voice only audio (with maybe a bit of music background here and there), so I don’t really know if music would be much different, but in ABX tests, almost all of the 2nd version mp3 I’ve done has be undistinguishable (for me) from the first encoding. I don't believe this would continue to hold through multiple re-encodings, but I have had not reason to test that.