Assume two cases:

1) A great setup, wonderful stereo microphones, a live concert, and electronics that capture the sound and convert it to 320 CBR MP3's.

2) A great setup, a perfectly created CD of that same concert, and the same electronics ripping the CD and converting the WAV to 320 CBR MP3's.

How much worse case 2 sound? Is it possible to quantify what has been lost in some way, perhaps as a percentage?

This is an odd question. Generally speaking, they would be the same.

Howevever, if you are recording @ 24bit / 48kHz, then would suffer a very slight (but likely inaudible) quality loss with #2, because you would have to downsample to 16bit / 44.1kHz

----------

How could they be the same. Certainly, the CD is not the same as the original sounds.

Again, assuming that you are recording at 24bit / 48kHz (48kHz being MP3's highest available samplerate), you would gain -96dB of noise and lose frequency info above about 21kHz when downsampling to CD levels.

---------
Oh, ok, you are assuming digitization of even the live recording, followed by encoding at 320 CBR.

I should have been more precise. Assume instead that the recording of the live concert captures perfectly THE ENTIRE WAVEFORMS.

But then, to my knowledge, you couldn't burn an audio CD of the concert. You would have to make a WAV file that fit audio CD specifications. Technically it would be better to go straight to 320kbps mp3 and skip out the audio CD part as stated previously, there would be a very slight drop in audio quality when going down to 44.1KHz and 16-bit depth.

However, one cannot assume that the recording perfectly captures the entire waveform as this never happens. Even in the studio, this doesn't ever happen due to the nature of recording (even modern-day digital recording doesn't capture absolutely everything perfectly). So that is not a safe assumption to make. One should look at the situation from a real world perspective in that some information will be lost during the recording phase. You won't be able to hear the information that is lost if your equipment is good though.

Perhaps I am misunderstanding the original question.
Is the question not:

Live Concert -> Perfect Microphones -> Perfect Recording -> MP3
VS
Live Concert -> Perfect Microphones -> Perfect Recording -> CD -> MP3
?

If so, I think benski nailed the answer out of the gate.


with his later clarification


Perhaps I misunderstand the question?
You asked specifically how much worse case two would sound.
Let us make the (reasonably safe) assumption that MP3 is transparent to your ears @ 320 CBR (the bitrate in your question.)
So the difference is that case one will sound to you like 24bit / 48kHz and case two will sound like 16bit / 44.1kHz.
Can you tell a difference? Likely not. But this too can be tested w/o needing to go to the lengths described in your original question.

I think the original poster is inquiring about the differences between a live concert and a concert CD. An example would be MP3's that were ripped from . . . say . . . the Statler Bros. Farewell Concert CD vs. MP3's that were encoded from recordings made during the actual concert.

It's a bit difficult to make an mp3 without digitising the sound waveform first!

Cheers,
David.

MP3 is a compression technique applied to digital recordings.

If a live performance is miked and represented at analog line-level as "high-quality", then a PCM recording can digitize that analog waveform with a given sampling rate and precision. CD is one common PCM-based technology that will give you (best case) 22kHz of bandwidth and 96dB of dynamic range. Other practical PCM-based technologies can give you 100kHz of bandwidth and 10 or 20dB more of dynamic range.

Once an analog signal is digitized, mp3 can only degrade the signal. A CD converted to mp3 will sound equal or worse than the CD (or "more different from the original").

-k