That is true, I wasn't addressing blatant noise issues, which comes up disturbingly frequently in computers. My comment only applies to working setups. I get noise out of my laptop if plugged into the power supply and connected to an earthed amplifier. So I just don't use that configuration.
It is true that rolloffs measure differently, but the differences are basically always inaudible. These days, oversampling lets the filter pass window move so high that it's basically a non issue.
Yes these modern DAC's do push the filter pass window very high but fundamentally as its a mathematical algorithm that represents the original signal it still introduces errors and other artifacts into the signal that were not there in the original. The argument for pushing it higher is that's its out of audible range and that the rounding errors in the calculations are much lower. However I believe this still impacts on the timing of the signal and the harmonics which go way above 20khz to which our ears are particularly sensitive.
Forgot to mention, people likely can hear difference between a NOS filterless DAC with a lot of amps since a NOS filterless DAC is a mathematically broken design, and you're inducing all sorts of intermodulation distortion in the amplifier by using a DAC like that.
That be as it may but it sounds great!! Same types of argument have been leveled for years about tube/valve amplifiers too. Although measurement is one part of the equation its also how the device sounds too.
Edited by complin - 5/19/14 at 8:35am