We've seen this test before, but why not give it a go again.
I used my office computer system which consists of:
Turtle Beach Santa Cruz sound card
H/K 380i receiver
Angstrom Legato speakers
Here's the test:
http://www.klippel-listeningtest.de/lt/default.html With the furnace clicking on and off while i was doing the test, i managed a -27dB range on average, just ahead of the distribution curve. I figure the hump at the top end of the curve is from people who have extracted the sound data and viewed a response graph of some kind to ascertain which sample had distortion and chose the correct answer. The end is too skewed otherwise.
At the bottom end the hump is likely from people who started the test, found it to be long and just clicked randomly to finish it off and see what happens at the end skewing the results downward.
Those are my thoughts on it.
