Originally Posted by jagorev
This thread is a real forehead-slapper.
This thread is a good example of why you need to be so careful of snake oil salesmen and flase improvements in the audiow world. Proper testing is important and it needs to be blind. It's easy to fool yourself, and it's easy to screw something up and test the wrong with.
FLAC and ALAC cannot produce different outputs, unless there's a bug en the encoder or decoder. They perform a mathematical compression of audio data (they mainly use a L-R to M-S conversion, a redundency predictor, and some bit packing like rice encoding). This process is fully revesrable and is provably losless both in the sense you can test it empiricly, and you can write a mathematical proof showing it to be.
Thus any percieved differences have nothing to do with the file formats. However, people may believe they do. There may, in fact, be no difference, just the brain playing tricks. This is a well documented phenomena in psychology for all kinds of things. If there really is an actual difference, it would be due to a problem with the test. Perhaps one decoder feeds data through an EQ, and another does not.
Well this is an example of why you need ot be so careful when testing expensive components that there's no science backing up the improvement. If you buy an expensive component, you want to believe the money was well spent, and thus you believe you hear a difference, even if there is none. If you really want to see if you arehearing what you think, you need ot do a test where your biases can't influence things, and make sure the test it setup right so there's not another factor influencing things.