AES Paper Digest: Differences among Several High Sampling Digital Recording Formats

[Wombat]

Yeah, but the question is whether they've measured/checked for that or not. Haven't read the paper yet; only read the abstract, so I don't know. Confounding variable means in this case that there may be other things that explain the outcome, not just the variable they were interested in (i.e. the perceptability of ultrasonic frequencies).

Lots of 'what ifs'.
Measurements will reveal the frequency interactions. What people wish to believe is another matter.

 

[Krunok]

I really don't understand the issue here: anti aliasing filter should remove anything above half of the sampling frequency during AD conversion (admittedly, this can still result in lot of ultrasonics in case of very high sampling rates) but low pas filter should do the rest in the DAC stage. Modern DACs usually have at least 5-6 of them and some start to roll of even before 20kHz.

 

[Krunok]

Lots of 'what ifs'.
Measurements will reveal the frequency interactions.

Sure it will. Frequency interaction is what we usually call "intermodulation distortion", right?

 

[Krunok]

The problem with producing high-res music is that what is captured there is frequently noise and garbage. See my analysis in these three samples:

https://www.audiosciencereview.com/...und-liaison-high-resolution-music-clips.2012/

https://www.audiosciencereview.com/...gh-res-music-naxos-jazz-at-the-pawnshop.2004/

https://www.audiosciencereview.com/...usic-the-lake-poets-live-at-the-minster.2003/

If you're right with your statement from the first link ("All I can think of is that the vocal mic/amp were at fault.") than this is an isolated case due to a fault and not something that is expected to happen regularly.

Edit: if I got t right it's the same thing for the 3rd recording as well. Seems like recording engineers simply didn't put too much effort when making these recordings.

 

[Jakob1863]

@Wombat ,

Lots of 'what ifs'.
Measurements will reveal the frequency interactions. What people wish to believe is another matter.

Sorry, but what are you arguing for?
Sure, measurements _could_ reveal "the frequency interactions" but does that help if nobody did the measurements?

Oivavoi is right; it would be great to have some more efforts to replicate these experiments wrt socalled "hypersonic brain effect" . Afair the various authors of follow-up experiments always found some evidence of an impact of ultrasonic frequency content, but - again afair - none of them measured IMD in the audio band caused by the ultrasonic content.
I remember another study that examined the case, found some related IMDs in the audio band but concluded that these were inaudible.

I´m not aware of other studies trying to replicate "hypersonic experiments" as the others were all examing something else, although more or less related.
Generally, although there will (most likely ) never be a perfect experimental research case, the matter of IMDs is an important one and should be addressed.

 

[Jakob1863]

What would be the human part that would sense the ultrasonic frequencies for our unconscious brain to "hear"? Could someone point me to scientific papers which show that these frequencies can reach the brain?

Oohashi et al. revived the interest in ultrasonic content with their AES paper from 1991 /1/, published another expanded one in 2003 /2/ and tried to explore the mechanism in an additional experiment- published in 2006 /3/- where they used sort of shielding clothes, that provides substantial attenuation of the highfrequencies, on their listeners. If the shields covered the whole body, no percepted differences were reported so the authors concluded that a new mechanism contributed to the different perception in case of added ultrasonic content. (just a short summary, so blame me for inaccuracies)

I have a weak recollection that there was some corrobation for this additional biological effect by authors from another field.

/1/ Oohashi et al., High-Frequency Sound Above the Audible Range Affects Brain Electric Activity and Sound Perception, AES convention 91, paper no. 3201
/2/ Oohashi et al., Inaudible high-frequency sounds affect brain activity; hypersonic effect, J Neurophysiol. 2000 June; 83 (6), 3548-58
/2/ Oohashi et al., The role of biological system other than auditory air-conduction in the emergence of the hypersonic effect, Brain Res. 2006 Feb 16, 1073-1074, 339-47

 

[amirm]

Doesn't the low pas filter stage after DA chip cut those?

No because these are in-band noise at high sampling rates. I am not showing what is filtered out.

 

[pirad]

I came across some non-scientific tests showing people’s preference for 320kbs over 44.1khz. I conducted my own tests. They showed a slight preference for 320kbps. Perhaps the psychoacoustic shaping of the compression codec is more important than the uncompressed samples ? For whatever this anecdotal evidence is worth I do not use 320kbps except for my car library on the smartphone. On the other hand perception does not necessarily give preference to some preconceived veracity. The story of Judgement of Paris as told in the movie “Bottle Shock” favored unknown Californian Montelena Chardonnay over top French wines. I got intrigued by the story and investigated it further including prolongued “research” in Napa and Sonoma and visits to Chateau Montelena (pic). Interestingly the winning blend came from several local vineyards. Mike Grgich was the winemaker behind the winning wine. He is not in the movie.

 

[oivavoi]

I came across some non-scientific tests showing people’s preference for 320kbs over 44.1khz. I conducted my own tests. They showed a slight preference for 320kbps. Perhaps the psychoacoustic shaping of the compression codec is more important than the uncompressed samples ? For whatever this anecdotal evidence is worth I do not use 320kbps except for my car library on the smartphone. On the other hand perception does not necessarily give preference to some preconceived veracity. The story of Judgement of Paris as told in the movie “Bottle Shock” favored unknown Californian Chardonnay over top French wines. I got intrigued by the story and investigated it further including prolongued research in Napa and Sonoma and visits to Chateau Montelena. Interestingly the winning blend came from several local vineyards. Mike Grugich was the winemaker behind the winning wine. He is not in the movie.

Interesting. There are also a couple of studies which show that young people seem to prefer YouTube audio over CD quality. Probably because that's what they are used to. Same phenomenon, IMO, that may make some people prefer tubes and vinyl. We shouldn't confuse fidelity with preference.

 

[pirad]

fidelity vs preference in subjective tests- how is it accounted for?

 

[oivavoi]

Subjective preference tests by definition can't be about fidelity. Subjective bypass or before-after tests may investigate whether a certain device is transparent or not - to those particular listeners with that particular system. That's about it, as I see it. Thumbs up for measurements.

 

[Cosmik]

fidelity vs preference in subjective tests- how is it accounted for?

Wrong question I think. Can it be accounted for? Maybe not, and that's perfectly OK. Maybe you just have to decide what you're aiming for on logical grounds and go for it. Aim for neutrality and stick to your guns, or you may never get anywhere.

And that's what happens anyway: listening tests are only conducted on mild variations of neutral. It's a pretence to think that listening tests are establishing 'preference', because no one checks whether listeners might not prefer radical DSP processes. If they did accidentally find that listeners preferred the 'Super Wide Stereo Hall' setting on an AV receiver that had inadvertently been left on rather than the $100,000 amplifier, those results would be binned!

 

[Wombat]

Subjective preference tests by definition can't be about fidelity. Subjective bypass or before-after tests may investigate whether a certain device is transparent or not - to those particular listeners with that particular system. That's about it, as I see it. Thumbs up for measurements.

Well trained and experienced listeners may do better than most in comparative subjective listening in selecting 'better quality reproduction'- better than the typical 'home' discriminator' who readily puts his/her discerning listening opinions forward as a given- degrees of subjectivism. The problem is getting a reliable standard against which to judge opinions. Measurement requires analysis but references can be struck. Objective analysis is a work-in- progress, improving all the time. Subjective analysis is hard pressed to be definitive.

 

[Cosmik]

Well trained and experienced listeners may do better than most in comparative subjective listening - better than the typical 'home' discriminator' who readily puts his/her discerning listening opinions forward as a given- degrees of subjectivism. Measurement requires analysis. Objective analysis is a work-in- progress, improving all the time.

They may do better at discriminating, but can you show that their preference is worth more as a result?

 

[Wombat]

They may do better at discriminating, but can you show that their preference is worth more as a result?

Only by statistical degree in established subjective-discriminatory test routines. The test routines are, of course, subjective in structure.

 

[Cosmik]

Only by statistical degree in established subjective-discriminatory test routines. The test routines are, of course, subjective in structure.

So perhaps they show themselves better at discriminating against known differences (who chose the differences and why?), perhaps more repeatably and reliably - less 'noise'.

But a piece of test equipment would be superior on these counts by orders of magnitude. Why don't we ask it what its preference is?

 

[Wombat]

So perhaps they show themselves better at discriminating against known differences (who chose the differences and why?), perhaps more repeatably and reliably - less 'noise'.

But a piece of test equipment would be superior on these counts by orders of magnitude. Why don't we ask it what its preference is?

Maybe.

I only take opinions on-board for consideration unless there is something more tangible and convincing(concrete) available. Always open to new information, though.

 

[pirad]

So perhaps they show themselves better at discriminating against known differences (who chose the differences and why?), perhaps more repeatably and reliably - less 'noise'.

But a piece of test equipment would be superior on these counts by orders of magnitude. Why don't we ask it what its preference is?

Who sets the test equipment preferences?
The wine tasting methodologies can be controversial too...
https://en.wikipedia.org/wiki/Judgment_of_Paris_(wine)

 

[Wombat]

Who sets the test equipment preferences?

Those who establish that the test equipment performs below the level of acceptable interaction with the test regime and can produce consistent results.

 

[pirad]

Those who establish that the test equipment performs below the level of acceptable interaction with the test regime and can produce consistent results.

I meant more like "who programmed the programmer..."