I am trying to understand if I would lose any perceivable sound quality (subjective, I know, but relatively speaking) under the following 2 scenarios:

1. 'Reconverting' an AAC file that was originally ripped at 192kbs to 128kbs. Obviously, the quality would change between the 192 and 128, but I am wondering if the 'reconverting' from the already compressed file would make the 'new' 128 file any less quality than if I were to have ripped it straight to 128 from the CD. Make sense?


2. Downloading files from iTunes (at 128kbs obviously) and then burning them to a CD to re-import them again into iTunes. I am doing this to remove the iTunes file protection so I can use AACGain to normalize the volume. Does re-importing them from the burned 128kbs made CD cause any distinguishable/significant audio quality loss from the original downloaded iTunes files?


Thanks in advance for any answers/assistnace!

The answer to both questions is "yes".

However, the question of whether or not you hear the difference would require some ABX tests with your sample files.

-brendan

Even though I wouldn't hear the difference, a reconversion (in any way) would "psychologicaly" disturb me

The process you describe in both questions is called "transcoding". No matter whether you directly transcode a lossy 192 kbit/s to a 128 kbit/s file or create an audio CD out of lossy files and import the resulting tracks to iTunes, both scenarios are in fact the same and could lead to an audible loss of quality. That's because a lossy encoder irreversibly removes major (mainly inaudible) parts of the original music and also alters the audio data during every encoding process, i.e. transcoding could cause possible distortions and further quality losses. Hence, it's generally recommended to encode your music either directly from the original CD or from a lossless source, but never from a lossy one.

If you really need to transcode your AAC files, as described in the second question, use high bitrates to prevent them from suffering audible quality losses. The definition of "high" is up to your ears, personally I wouldn't choose any average bitrate below 192 kbit/s. Or, even better, go for ALAC, since transcoding to a lossless format won't cause any further changes to the audio data.

For more detailed information about the transcoding process, read here: http://wiki.hydrogenaudio.org/index.php?title=Transcoding

That's just one of the 'funny' things you run into when buying 'protected' DRM files. Absolutely not recommended as there are legal stores which sell unprotected files, plain old redbook CDs or you might be interested in CC-licensed music and such.

And yes, there are losses. Especially if you go for lower bitrates like 128 reencoded to 128, it will probably be audible. However, for storage on my mobile, I do transcode from higher bit vorbis or mp3 to q .25 HE-AAC for more storage space. On the move, this is near transparent for me and the reencoding loss wouldn't matter much at this rate anyway.

Thanks a ton guys (or gals)! This helps tremendously.

loophole's little transcoding tests might also be of interest concerning this topic: http://www.hydrogenaudio.org/forums/index....showtopic=48999

Despite what you would expect in theory, practice shows that transcoding isn't always as bad as its reputation. However, it should be mentioned that people usually don't transcode from a high bitrate like ~192 kbit/s to the same high one using the same codec, except as a workaround to get rid of DRM. Commonly they transcode their files either to lower bitrates to save some space on their portable devices or to other file formats for compatibility reasons. And in these two cases the results might look a little different, although I must admit that I haven't done any serious testing so far. The only listening test concerning transcoding seems to be the one done by guruboolez a while ago.