lrossouw wrote: > Consequently if you have a compression algorithm that assumes the ear > is a "linear system" and then removes things a linear system can't > "hear" you'd be removing things the ear can actually hear. Perceptual coding is not like zipping a file. No mp3 encoder left out the real behaviour with testing. It is not like one simple algorythm leaves out information. There are many choices an encoder has to do and all of them are fine-tuned by real listeners on a huge sample corpus. So even if the FFT that may be used as part of evaluating what to leave out another choice of manualy changing the masking behaviour makes the encoder a good one or a bad one. The news that some aspect of human hearing behaves different as some FFT changes nothing here.
After reading some things on the net i see several audiophile claims now take parts of that paper as prove for everything that makes no sense. Totaly unfounded and out of context. You may read some posts here from J. D. Johnston on Hydrogenaudio: http://www.hydrogenaudio.org/forums/index.php?showtopic=99371&st=25 He is one real authority on the subject and knows everything behind it. I even may quote his statement about the thread title over there. The thread is called "Human hearing beats FFT" Quote: "So, the headline is just confused, it's comparing an apple, an orange, and a crate full of bowling balls, and concluding that apples are orange-colored and weigh 12 lbs." ------------------------------------------------------------------------ Wombat's Profile: http://forums.slimdevices.com/member.php?userid=4113 View this thread: http://forums.slimdevices.com/showthread.php?t=98124 _______________________________________________ discuss mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/discuss
