On Nov 3, 2006, at 6:29 PM, Solar Power Studios wrote: > Leslie Bell wrote: > >> They'd play the same performance, and >> have people guess which was 192k. Not everyone could tell the >> difference, but a majority of the pros there could. > > Interesting. I'm curious, do you know the details of how the two > sources > were recorded? > I once heard a 192 vs 44.1 demo where the low res version was just > the high > res, sample converted down...which of course didn't prove very much. I don't know. It could actually say a lot, depending of course on the quality of the conversion. What was the reaction of the people listening? The paper at the link that Jay Asher posted contents that the electronics doing the sampling basically become so inaccurate at the high sampling rates that the fidelity is actually damaged. So it is quite likely if that is the case that when golden ears hear a difference between high sample rate play back and 44.1 or 48k play back what they are really hearing is the difference between a undistorted low sample rate file and slightly distorted high sample rate file. OTOH in a test where a file was recorded at 192 and then a copy was down sampled to 48k assuming that the down sampling was perfect if the listeners could still identify a difference then it would in fact be conclusive proof that people were perceiving *something* over 22k. Of course assuming the down sampling is perfect is a big *if* too.