-

abbey road d enfer said:

Well, I've measured a number of 192k converters (trying - and failing - to find one suitable for measurements up to 80-90kHz), but, apart from the Icon Cube, that has frequency response anomalies at all SR, the others have no differences whatsoever in the 20-20kHz, whatever the SR.

Click to expand...

Are you strictly talking frequency response? Or just measurable differences in the 20-20k realm? In my experience something like an IMD test with 18k and 19k tones makes it very easy to measure differences.

If you thought politics made people upset?!

There's a great Podcast from Seth Godin called Akimbo and on a recent episode he makes the case for accepting the power of placebo.  If one expects converter X or sample rate Y to sound better it will.  So go ahead and enjoy whatever it is that floats your boat, the "difference" may or may not travel outside of your studio but it probably won't be noticed in the negative either.

All that said I've spent a lot of time on this and if you care take the time to do some tests. 

On 192k I have yet to hear it sound better and I have a lot of heavyweight Grammy winning friends who swear that it does, several recording in DXD - 384k.  It doesn't seem to sound worse to me but I'm happy at 96k.

john12ax7 said:

Are you strictly talking frequency response? Or just measurable differences in the 20-20k realm? In my experience something like an IMD test with 18k and 19k tones makes it very easy to measure differences.

Click to expand...

I was interested only in frequency response, since my quest was for measurement.

80hinhiding said:

96Khz seems to be doing a decent job of preserving a signal that comes off of tape.  I figured 192Khz might do even better.. to leave fewer gaps. 
A

Click to expand...

There are no 'gaps' . - due to the reconstruction action of the DAC filtering.
There is quantisation error - reduced by increased number of bits and dependent on dither applied.
With increasing sampling frequency the filter requirements mean that there can be greater temporal accuracy in the audio band.

john12ax7 said:

Are you strictly talking frequency response? Or just measurable differences in the 20-20k realm? In my experience something like an IMD test with 18k and 19k tones makes it very easy to measure differences.

Click to expand...

My favorite test....  ;D back in the late 70s I modified my Heathkit SMPTE IMD tester to use 19kHz:20kHz  1:1 (instead of the stock 60Hz and 7kHz that was too easy). Back then I used this on analog paths (great for parsing out performance concerns in phono preamps).

While perhaps not obvious, the summed 18k and 19k tones will exhibit a rate of change equivalent to 37k. While this can occur naturally with complex music signals it rarely does.

JR

JohnRoberts said:

While perhaps not obvious, the summed 18k and 19k tones will exhibit a rate of change equivalent to 37k.

Click to expand...

Yes, two -6dB tones would conjure the same Slew Rate as a single 0dB tone in these conditions; I'm not sure it is more stress to the amplifier, but it makes detecting the non-linear by-products easier.

While this can occur naturally with complex music signals it rarely does.

Click to expand...

It does happen with cymbals.

https://xiph.org/video/vid1.shtml
https://xiph.org/video/vid2.shtml

And (at least) the latter half of https://www.youtube.com/watch?v=BYTlN6wjcvQ
As well as https://www.youtube.com/watch?v=Zvireu2SGZM

abbey road d enfer said:

Yes, two -6dB tones would conjure the same Slew Rate as a single 0dB tone in these conditions; I'm not sure it is more stress to the amplifier, but it makes detecting the non-linear by-products easier.
It does happen with cymbals.

Click to expand...

That's why I didn't say never....

JR