dafiend wrote: > I don't think this reference is relevant. The test was conducted in > January 2000. Codecs have improved since then. A LOT. > That's what I said. I've seen newer ones. The principle stays the same. > > As for the rest of your post: I don't buy it at all. I really would like > to see one of the more knowledgeable posters chime in (Julf, probedb > etc.) > You can consider "knowledgable" whoever you prefer. I would suggest you get a pair of earplugs as used by divers and do the test for yourself.
> > My laymen's take on this is as follows. Lossy audio compression will > introduce noise and distortion. The codec employs a psychoacoustic model > that tries to introduce only artifacts you cannot hear. To that end, it > makes use of spectral masking. If you introduce an artifact, say, at > 900-920 Hz, but in that same frequency band, you have signal with a > level 25 dB higher, you probably cannot hear the artifact. > > So what is the story you can tell which explains why somebody with > impaired hearing would be more capable of hearing the articaft down > there? > Ummm, sorry, there is no such thing as "different signals" in "the same frequency band". If you define a band, it's a band, there's one signal in that band. If you modify it, you modify the whole signal. What you do in an encoder is that you quantify your signal into a list of FFT quantizers. That's pretty simple for a single sine wave so that you can perfectly represent it with only a few bits. If you've got a more complex signal, though, you need more quantizers and when those cross the threshold defines by your target bitrate you simply remove some. Which ones you remove is defined by the psychoacoustic model. The model is intelligent enough to try to make sure the information loss (or more precisely: the wrong information created by the decoding process) is in the inaudible range and starts with a huge threshold but if your signal grows more complex you get closer to the audible range. This is also why studio recording usually encode just fine at even low bitrates. These days they are mastered in a way that even 128kbps mp3 doesn't really do harm to them. You can do this by mastering your signal in a way that you simply never reach heavily audible artifacts. This is much more difficult for live recordings, of course you can re-master those, too, but you will hear this, the whole recording will sound worse. There is no way a better codec can remove this effect, it's in the very principle of how the encoding works. It assumes a certain curve of frequency response in your hearing system, if you modify that curve the codec works less well. Of course, all of this gets worse with lower bitrates. > > As an aside, this whole discussion seems a bit funny. Should we conclude > from your analysis that hearing impairment is why people should spend > $$$ on lossless audio streaming? (I know you didn't say precisely that, > I'm exaggerating.) I completely don't care why people spend money on stuff, for me they can throw it in a river or spend it on golden digital cables or nonsense like that. But it's a fact that it's actually often not _superior_ hearing that allows you to identify the artifacts but actually _inferior_ hearing. I kind of suspect that some (not all) of the opposition you get from some artists towards mp3 might have to do with this. A lot of artists and DJs who've played live for a lifetime develop hearing impairments (all those loud monitors), they probably hear the result differently from how you and me perceive it. --- learn more about iPeng, the iPhone and iPad remote for the Squeezebox and Logitech UE Smart Radio as well as iPeng Party, the free Party-App, at penguinlovesmusic.com *New: iPeng 8, the Universal App for iOS 7 and iOS 8* ------------------------------------------------------------------------ pippin's Profile: http://forums.slimdevices.com/member.php?userid=13777 View this thread: http://forums.slimdevices.com/showthread.php?t=102351 _______________________________________________ plugins mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/plugins
