I'll bet that some training would make some significant differences. The person who correctly identified the cymbal differences may have spent some time listening for the differences. Just as young musicians are given tonal training early on, if you know what to listen for, I imagine the test subjects scores would have improved. I would also wonder how each subjects personal histories impact their preferences. I have heard that blind people tend to have more sensitive hearing, perhaps those who mostly listen to MP3s already won't be as sensitive to compression distortion. Another slashdotted story highlighted that people are growing increasingly fond of the hiss of MP3 noise (see link below). Perhaps subjects have a unconscious bias that makes it difficult assess what is better. It would have been neat if they had run test where both were MP3s or both were lossless. http://news.slashdot.org/article.pl?sid=09/03/11/153205

Wow, already past 62,000