Audiophile PC Music Player Must read

[jwc]

I've been reading Tom's Hardware since the beginning. I'm a computer geek really and never really seen A PC review site (of any credibility) venture into the true audio quality of PC sound systems w/o focusing on 3d gaming or Movie DTS typical BS.

I'm not agreeing with his conclusions.....but I feel a great read for everyone here.

http://www.tomshardware.com/reviews/high-end-pc-audio,3733.html

[Sancho Panza]

Thanks, will give it a read.

[WMcD]

I really do think that ABX tests are valid. People who flunk them argue that long term listening does allow themselves to discern subtle differences when they are not engaged in first-time listening to different hardware. But then somehow there is never a follow-up when they have listened long term. So the excuses are never re-examined.

But I do have a thought on the ABX tests -- they are not a fair comparison as they are run -- to my understanding.

It is my understanding that the listener is given a recording which is running. So the listener flipping (for example) between what is on the time display seconds 1-5 (as A), then 5-10 seconds (as , then 10-15 (as X) then 15 - 20, etc, Of course the listener is not constrained to 5 second samples in my theory but I throw these out as an example.

The musical content has changed quite a bit, in these three samples. So the listen is not really presented with the same music sound in A and B and X. So if asked to detect whether the hardware in X is more like A, or B, he can't possibly tell because the test signal has changed. E.g. does the first five seconds of a song sound like the second five, or the third five?. Just for tonight I'm thinking about Sgt Pepper and Lovely Rita Meter Maid. Perhaps four "bars" of music correspond to 2 seconds.

A better test would be to take a five second sample and then use that consistently as the A and B and X, over and over again. This could be a cymbal crash or a bass drum hit. Something which people could agree to as being challenging to the dynamic range of signal reproduction in the system of recording and playback and/or hardware. Or even very low level signals to test the low end of sampling and reconstruction, noise floor, and dithering.

For what it is worth: It was probably the second pilgrimage to Klipsch, the first one to Indy where they gave people the opportunity to ABX wires. Monster Wire was supposed to be one of them. But apparently that got pulled off the list. So I was left alone with an ABX of zip cord and something else. I had brought along a test CD with pink noise. In the ABX test, I could not hear a difference in a A-B test, much less ABX.

So, why is not pink noise used? I have not seen it. This removes the issue of the music being different described above.

WMcD