Originally Posted by drecar
Yet, are there not test subjects out there that could not hear a difference in abx sound test between the original sound and degraded mp3? Is it possible the test fooled their brains into thinking the sounds are the same? Have such test been done?
I don't believe that to be the case. I have statistically verified myself that a difference, in fact, and improvement could be 100% heard in a blind A/B test between a 128kbps mp3 (EAC + LAME) and a .wav file (EAC). This test was done in foobar2k, it was level matched, and the sample size was 100 trials broken into 5 groups of 20 trials over the course of the day.
Many people say "there is no difference", if the test could have fooled my brain, it would have esp. with all the preconceived notions that 128kbps = cd quality.
I repeated the test with 256kbps and scored 95 out of 100.
I could not score better than 70/100 on a test with 320kbps, so thats debatable.
Flac/Monkey I could not tell the difference period.
Level matching to 0.1mV takes away most of the bias that would be present during the test. Being blind to what device is operating takes away the rest of the bias.
I'd almost want to say that if you can't score >90/100 on a 320kbps EAC+LAME vs. .wav in a controlled level matched double blind A/B test, you would not have a chance of hearing slight variances between cdp if they do exist. If you can, this 'fact' would not change in controlled environment since it would be inherent in your senses to be able to detect this no matter what other people say, and other preconceived notions. In other words, you would "JUST KNOW" and that's that. If the test is not well-controlled, you cannot make any conclusions about the experiment. This is just the scientific-method at work. Unless all variables are accounted for, the results are not significant.