1st step: rip from .cda/.wav ~1400 kbps full-quality source. The audio begins as “clean”, with all audio data intact for full sound quality, but you then encode. Encoding from a lossless to lossy format, results in loss of audio information, and to MP3 with 128 kbps results in a lot of data being lost.
2nd step: burn new MP3 files of 128 kbps [lossy] to a CD with ~1400 kbps bitrate. Now, while the bitrate has gone up, and we have gone from lossy to lossless, the data from the original CD that was lost while encoding HAS been lost and will not be recovered by even the best MP3 decoder. Your CD sounds exactly like the lossy MP3s you made the first time.
3rd step: rip from an audio CD that SOUNDS as bad as a 128 kbps MP3 file but has a “physical” bitrate of ~1400 kbps, to another 128 kbps MP3 file. If the encoder is good, it will be able to do an efficient encoding job and closely match the original MP3. However, more audio information will still be lost in the compression, because no MP3 encoder [that I know of] will discard the same information twice. Thus, this second ripping to MP3 causes more of the original data to be lost; so the files [I]may still be the same physical size and quality[/I] [in relation to the bitrate], but the quality of the audio [I]as you hear[/I] it is worse than the original MP3.
The issue is not just going between bitrates, but going between a lossy and lossless format.