Re: [SlimDevices: Ripping] Obvious iTunes/AAC conversion question
RonM;435471 Wrote: > > As I understand it, the compression algorithms identify adjacent sample > points that are the same, or that are sufficiently similar given the > quality target, and recodes them as the same but without individually > describing each point. The lower the quality setting, the broader the > definition of "sufficiently similar". > Not really...but we'll go with it :) > > If this is essentially correct, then converting from one lossy format > to another may not result in significant additional data loss. If the > two formats have similar quality settings (bit-rate settings), the > conversion algorithm may not identify many opportunities for further > compression, since those opportunities have already been identified and > implemented. > You start with an original PCM file, and give it to your mp3 encoder of choice. It takes that input file, does it's magic and ends up with an mp3 file which sounds broadly similar to what you started with. Now to re-encode, you first decode back to PCM, but the PCM file is clearly different to the one you started with. Now you encode again - given a different input you get (yet another) different output. Rinse and repeat - every time you're removing more, changing more and adding more (in the form of artifacts) and getting further away from where you started. What's more, the algorithms between different formats are different, and the pre/post filters they apply are different (and destructive). Apply the transforms on top of each other and you'll get the artifacts and deficiencies of both combined. > > My knowledge here is pretty shaky, and I'm assuming parallels with > image compression. So I might well be wrong. > Take a photo. JPEG it at 80%. Convert it back to bitmap. JPEG it again at 80%. Back to bitmap. Do that a few times and see what a mess you're left with - and that's using the same algorithm every time! -- radish radish's Profile: http://forums.slimdevices.com/member.php?userid=77 View this thread: http://forums.slimdevices.com/showthread.php?t=64894 ___ ripping mailing list ripping@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/ripping
Re: [SlimDevices: Ripping] Obvious iTunes/AAC conversion question
Oh, you did... -- moley6knipe WinXP Pro SP3 > dBpoweramp Reference 13.1 > SqueezeCenter 7.4 24614 > Squeezebox 3 > ears moley6knipe's Profile: http://forums.slimdevices.com/member.php?userid=10014 View this thread: http://forums.slimdevices.com/showthread.php?t=64894 ___ ripping mailing list ripping@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/ripping
Re: [SlimDevices: Ripping] Obvious iTunes/AAC conversion question
You could be right, I'm just going on my perceived wisom. Or lack thereof! You could always post your query over at the dBpoweramp forums - the man there, Spoon, really knows his stuff. -- moley6knipe WinXP Pro SP3 > dBpoweramp Reference 13.1 > SqueezeCenter 7.4 24614 > Squeezebox 3 > ears moley6knipe's Profile: http://forums.slimdevices.com/member.php?userid=10014 View this thread: http://forums.slimdevices.com/showthread.php?t=64894 ___ ripping mailing list ripping@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/ripping
Re: [SlimDevices: Ripping] Obvious iTunes/AAC conversion question
moley6knipe;435350 Wrote: > Any lossy format will lose you bits. . . I'm not sure that this is uniformly true. As I understand it, the compression algorithms identify adjacent sample points that are the same, or that are sufficiently similar given the quality target, and recodes them as the same but without individually describing each point. The lower the quality setting, the broader the definition of "sufficiently similar". If this is essentially correct, then converting from one lossy format to another may not result in significant additional data loss. If the two formats have similar quality settings (bit-rate settings), the conversion algorithm may not identify many opportunities for further compression, since those opportunities have already been identified and implemented. I'd say that the best choice for a conversion setting (AAC to MP3, for instance) would be to choose an MP3 quality setting just above that used to create the AAC file in the first place. My knowledge here is pretty shaky, and I'm assuming parallels with image compression. So I might well be wrong. R. -- RonM RonM's Profile: http://forums.slimdevices.com/member.php?userid=17029 View this thread: http://forums.slimdevices.com/showthread.php?t=64894 ___ ripping mailing list ripping@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/ripping
Re: [SlimDevices: Ripping] Obvious iTunes/AAC conversion question
Any lossy format will lose you bits. Question is whether you can hear that loss - if not then I'd suggest it's not a problem! Try converting to 320k CBR mp3 and see if you can hear the difference. This http://www.hydrogenaudio.org/forums/index.php?act=Attach&type=post&id=4938 is a good read - contains details of a tool within Foobar for testing if you can hear the difference. The best method would be conversion to lossless, as you say, but again with formats it's never easy - FLAC is great with SqueezeBoxes but not everything plays it; apple lossless or windows media pro are your other mainstream choices. File size between the three is roughly the same I think, so it's a simple compatibility question from there. -- moley6knipe WinXP Pro SP3 > dBpoweramp Reference 13.1 > SqueezeCenter 7.4 24614 > Squeezebox 3 > ears moley6knipe's Profile: http://forums.slimdevices.com/member.php?userid=10014 View this thread: http://forums.slimdevices.com/showthread.php?t=64894 ___ ripping mailing list ripping@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/ripping
