[nick_charles]

Quote:

Originally Posted by TwoTrack /img/forum/go_quote.gif Nick, I appreciate your comments but I think a statistically significant finding in test 1 certainly reaches the goal of a demonstrable audible sound difference. Some here may not know but several commonly available mic reach up to 40khz. The increasing adoption of hirez in the pro world has created demand and mic technology continues to improve. Several omnis and ribbon mics have these stats.

Yes, for two subjects there was a detectable difference, that this was conscously audible is not certain as neither could detect a 22K signal at very high levels so how could they detect such music at lower levels (10db lower and with masking frequencies present) when presented separately through a supertweeter. One clue may be found in Lawton's:

Damage to human hearing by airborne sound of very high frequency or ultrasonic frequency: HSE report 343/2001.

Lawton reviews 40+ years of research on the effects of high frequency sounds which while not consciously detectable can cause discomfort, headaches, nausea and vomiting. Since the tests where the 2 subjects detected a difference had the longest exposure they would be most likely to generate these "discomforts" while the shorter exposures would not. This would explain the paradox where the shorter test was less sensitive than the longer test which is not normally true.

 

[SP Wild]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif For test 1 they had 13 subjects and each subject listened to 32 AB pairs. Two of the 13 subjects managed over 70% correct, this is statistically significant, though I would say it is far from stellar, neither managed 75%. The other 11 were at or about random chance. So let’s recap. In one experiment 2/13 subjects could marginally detect the presence of > 21K frequencies using an atypical Japanese instrument sample and a mic with a massive treble lift and absurd energy levels, with a randomizing pattern as yet undisclosed and not (apparently) adjusted for bias and very high levels of induced distortion after 2 minutes of exposure, but not when it was reduced to 20s and had to have the sounds at > 90 db to detect a pure 22K tone, which is granted impressive. Will this have me racing out to buy a high-res player ?

Like I've always suspected, good is never good enough when it comes to DBT procedures or reults. Can't blame those two, their hearing was only as good as yours Nick.

 

[SP Wild]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Lawton reviews 40+ years of research on the effects of high frequency sounds which while not consciously detectable can cause discomfort, headaches, nausea and vomiting. Since the tests where the 2 subjects detected a difference had the longest exposure they would be most likely to generate these "discomforts" while the shorter exposures would not. This would explain the paradox where the shorter test was less sensitive than the longer test which is not normally true.

I suspected as much after 2 months ownership of the Cambridge Dacmagic, switching from their steep filter (attenuates all frequencies after 22khz) and their other upsampling filters (that do not).

Didn't take me 40+ years of research or a fancy psychology degree, just confidence in my judgement.

 

[nick_charles]

Quote:

Originally Posted by SP Wild /img/forum/go_quote.gif Like I've always suspected, good is never good enough when it comes to DBT procedures or reults. Can't blame those two, their hearing was only as good as yours Nick.

Actually their hearing is much better than mine, but I am not jealous

As I said the paper was a pretty good paper and apart from some minor quibbles it was a very well conducted study.

Let's say that those two subjects really did reliably hear a consciously audible difference. What are the implications, that 15.38% of listeners can hear the presence of added ultrasonic content. Okay I'll accept that, as part of the other 84.62% I won't lose any sleep, I know that my hearing at 51 is not that good any more shrug.

However, in the 2nd test the same "good ears" were unable to detect the difference even when using segments with the highest levels of ultrasonics, this you must agree is puzzling. The only diffference was that the exposure was shorter.

So we can hypothesize that the long exposure had a different effect, but we also know that the listeners could not detect ultrasonic components on their own even at very high levels, yet the added ultrasonics while very high energy were at least 10db quieter than the pure tones.

So we are back to not being able to consciously detect the material but being able to detect it in a harder test. This seems more consistent with neurophysiological artifacts than audible ones, if we discount the extra distortion created by the ultrasonics.

The later Meyer and Moran paper (2007) did the reverse they filtered the ultrasonic content out by doing a 16/44.1 post-analog stage so any high content which was embedded in the original sources was stripped out of the redbook version and here not one of their subjects managed to detect the difference.

 

[nick_charles]

Quote:

Originally Posted by SP Wild /img/forum/go_quote.gif I suspected as much after 2 months ownership of the Cambridge Dacmagic, switching from their steep filter (attenuates all frequencies after 22khz) and their other upsampling filters (that do not). Didn't take me 40+ years of research or a fancy psychology degree, just confidence in my judgement.

did it give you headaches ?

 

[SP Wild]

Gave me headache alright - only after a little while, not straight away. Not just a headache but somewhat of a sickening sensation - mild seasickness is what it felt like. I didn't realise it was the music doing this - just thought I wasn't feeling well lately. I move to change digital input on the dac but pushed the wrong button and the steep filter was selected instead (22khz roll off) - immediatey the headache lifted and gradually the slight nausea eased. Carried out testing with my other phone and conclude only the steep filter did not sicken me. This was the only filter that attenuated the upsample frequencies above 22khz.

Your post was the only post or suggestion that matched my experiences. I kept it to myself as whoever I ask just skimed over the question and others will think I am making it up. But I have confidence in my judgement.

Oddly enough this occurred with my first amp only, the Musical Fidelity X-can V3 and Dacmagic upsampling. I didn't notice any ill effects with other subsequent amps. I now use a Bryston Dac with no such problems with its upsampling - needless to say I compared it to its non upsampling feature.

Perhaps I imagined this, still, your a psychologist you know I am telling the truth as I see it.