AVS Forum banner
Status
Not open for further replies.

Phantom Source Perceptions in 24bit 96Khz Digital Audio

1K views 66 replies 8 participants last post by  markrubin 
#1 ·
103 AES Convention 1997, New York

I just finished reading this paper. Its conclusions are interesting to me, so I thought that I would share the summary.

Conclusions

TWO experiments have been presented to investigate into "localization blur" as a function of the sampling rate of the stimuli. The first experiment which was concerned with localization of partially coherent signals did not reveal data exact enough for a quantitative evaluation. The second
experiment to estimate MAA's in a standard stereophonic loudspeaker layout gave results stable enough for evaluation.

Analyses of the data showed that the hypothesis that localization accuracy improves with higher sampling rates above the professional 48 kHz standard has to be rejected. Nonetheless, in a subsequent experiment, a task considered to be more difficult than the conventional ABX-test, one subject proved to be able to discern HDDA from 48 kHz @ 16 bit with a reliability not expected in such a difficult task. Two futher subjects were able to discern 96 kHz @ 16 bit from 48 kHz 16 bit at more than 90 % confidence.

The differences in sound quality between the two standards as noticed by the subjects, suggest that further experiments conducted in this area should focus on the finer structures of music. Dynamic shadings (contrast), ambience retrieval and distance perception seem to be good predictors worth further research.

We can conclude by saying that the movement towards higher resolution in digital audio, at least at the recording studio level, is justified especially if we wish to prevent the finer nuances of contemporary music from being lost for future generations.
 
See less See more
  • Like
Reactions: Scotth3886
#5 · (Edited)
Great, I'm not sure I will understand all of this with almost zero technical background, but it appears to be something I'm very interested in




"One of the areas which is often cited as exposing the

deficiencies of restricted time and amplitude resolution is in

the reproduction of the spatial attributes of the auditory

event. These spatial attributes can be roughly divided into two

categories, spaciousness 1,2 and localization accuracy 3 .


Localization (accuracy) can again be divided into localization

of single sound sources, localization of multiple sound sources



radiating coherent signals and localization of multiple sound

sources radiating partially coherent or incoherent signals."


 
#9 ·
I missed the pre part of pre/post presentation. Saying it was reviewed because it was presented at a convention reads like it was reviewed at the convention. The wording was ambiguous, I thought you meant something different and misread your last sentence. My apologies. I know how peer review works, I've gone through the process many times.


Just a tip. If you don't want other forum members to respond to you in an abrasive or even personal way, try to refrain from introducing that type of tone with your posts. My only comment in this thread was a simple statement. To respond with "I knew you would put your foot in it" is inviting, maybe begging, for yet another drawn out argument that ends in a thread lock after you cry about how posters respond emotionally instead of factually. If you don't want to be involved in those types of exchanges (it seems you do), don't instigate them.
 
#12 · (Edited)
I think we want the same thing from each other... It's great that you can admit when you're wrong.

It seems, that I may have been wrong as well.

Now that the niceties are out of the way, do you have anything technical in nature to share, or are we to expect more rhetoric?
 
#16 ·
Those of us who have worked professionally in high end audio for decades certainly know of Hawksford. As I mention a year ago:
When I see some of these "scientists" Reiss selected for his 18 studies it makes me shudder. . . .
I know Hawksford is a promulgator of baloney like wires [all with the same, proper gauge and RLC characteristics] audibly mattering, but I don't know nearly as much about him as I do Oohashi.
Like so many who believe in magic he is dismissive of DBTs unless he is publishing in JAES where he is forced to respect it as being sound if he expects to be published.
 
#17 ·
Those of us who have worked professionally in high end audio for decades certainly know of Hawksford. As I mention a year ago:


Like so many who believe in magic he is dismissive of DBTs unless he is publishing in JAES where he is forced to respect it as being sound if he expects to be published.

This is pure rhetoric Zillch. Do you honestly believe that you're opinions undue these gents credentials and credibility.

Electronic devices and CODECS do sound different Zillch. It's not a big deal that they do, it's a good thing!
 
#18 · (Edited)
"Of course, with DBT it’s
easy to get a null result, so it may be a
good method if that is your agenda.
My preference is to undertake listening
tests in a completely darkened
room. The fact that the equipment you
listen to isn’t hidden and could be identified
with just a bit more light makes
it much more natural and less stressful
than being aware that the equipment
is purposefully hidden from you. Also,
being able to focus your senses purely on
sound and not be distracted by uncorrelated
visual input to the brain heightens
your auditory perception. It is very easy
to do and increases your sensitivity and
acuity, especially in spatial terms."

-Hawksford

https://linearaudio.nl/sites/linearaudio.net/files/Didden-Hawksford_Part1.pdf

Is it just me or does he barely even seem to understand what DBT means? He sounds like one of these truly misinformed people who think a DBT literally uses blindfolds over the people's eyes. :sigh:
 
#21 · (Edited)
"Of course, with DBT it’s
easy to get a null result, so it may be a
good method if that is your agenda.
My preference is to undertake listening
tests in a completely darkened
room. The fact that the equipment you
listen to isn’t hidden and could be identified
with just a bit more light makes
it much more natural and less stressful
than being aware that the equipment
is purposefully hidden from you. Also,
being able to focus your senses purely on
sound and not be distracted by uncorrelated
visual input to the brain heightens
your auditory perception. It is very easy
to do and increases your sensitivity and
acuity, especially in spatial terms."

-Hawksford

https://linearaudio.nl/sites/linearaudio.net/files/Didden-Hawksford_Part1.pdf

Is it just me or does he barely even seem to understand what DBT means? He sounds like one of these truly misinformed people who thing a DBT literally uses blindfolds over the people's eyes. :sigh:
Just to be clear you are defaming this gent correct: http://www.aes.org/events/135/presenters/?ID=1208

Malcolm O. J. Hawksford was educated at Aston University in Birmingham, UK where he received his B.Sc. with first-class honours in 1968, Ph.D. in 1972 and D.Sc. in 2008. He is now Emeritus Professor within the School of Computing Science and Electronic Engineering at Essex University, Colchester, UK. Early research embraced delta- and sigma–delta modulation (SDM) applied to colour TV coding that under the award of a BBC Research Scholarship lead to a method of luminance and chrominance multiplexing exploiting digital time-compression, a forerunner of MAC/DMAC. Principal interests include audio engineering, electronic circuit design and signal processing focused on loudspeakers, SDM, PWM linearization, spatial audio and telepresence. Malcolm is recipient of the AES Publications Award for the best contribution by an author of any age for JAES, volumes 45 and 46 and holds the AES Silver Medal for major contributions to engineering research in the advancement of audio reproduction. He is a chartered engineer and fellowHe has been formative chair of the AES Technical Committee on High-Resolution Audio and was a founding member of Acoustic Renaissance for Audio (ARA). of the AES, IET, and IOA.
 
#25 · (Edited)
Here's a little trivia for you all. RRod in another forum did some deep statistical analysis and discovered that Reiss' recent Meta study pretty much crumbles away without this very study this thread is discussing [called "Thiess" in his paper]. I.e. It has pivotal significance in letting Reiss reach his goals of not having to admit that his two administrative positions regarding the legitimacy of Hi-Res is a bit silly, since nobody can actually hear it.
 
#27 · (Edited)
Here's a little trivia for you all. RRod in another forum did some deep statistical analysis and discovered that Reiss' recent Meta study pretty much crumbles away without this very study this thread is discussing. I.e. It has pivotal significance in letting Reiss reach his goals of not having to admit that his two administrative positions regarding the legitimacy of Hi-Res is a bit silly, since nobody can actually hear it.
Doesn't your hand ever get tiered from all this waving in here. I have the entire AES Library downloaded mate. For every loose reference you point to, I can counter a 100-1, with real studies. So why don't we just stick to the two, currently presented in here?

After all, the 'nobody can hear it' part that you are claiming is under ABX. I have just noted a paper that calls the metric into serious question, and also ports some mighty fine recommendations towards its corrections.

:)
 
#26 ·
Would I be correct that nobody actually reading this thread claims to be able to hear an audible difference between [what was it again?] 48kHz and 96kHz?
 
#29 ·
Yeah, I'm sure all of AES is now scrambling away from ABX. [sarcasm]

I like how some of the critics, like Stuart, use crippled versions of it, such as AX, or whatever he calls it, which technically "can" work but in truth only lowers the listener's sensitivity to hear small impairments. Plus anyone who really understands ABX realizes that a listener is fully allowed to only listen to A and X, but never B, if they so choose.
 
#30 ·
Should anyone want to learn how ABX works, here:
 
#31 ·
Should anyone want to learn how ABX actually works, here:
 
#37 ·
Should anyone want to learn how ABX actually works, here:
https://www.youtube.com/watch?v=r5njCrqCN9k
And when you're done watching it, locate, download, and read the following:

How Conventional Statistical Analyses Can Prevent Finding Audible Differences in Listening Tests

Author: Leventhal, Les
Affiliation: University of Manitoba, Winnipeg, Manitoba, Canada

The conventional .05 significance level used to analyze typical listening tests can produce a much larger risk of concluding that audible differences are inaudible than concluding that inaudible differences are audible than concluding that inaudible differences are audible, resulting in strong systematic bias against those who believe differences are clearly audible between well designed components that are spectrally equated and not overdriven. This paper discusses ways to equalize error risks, introduces a quantitative measure of a listening test's fairness, discusses implications for literature reviewers, and presents a statistical table enabling readers to conduct equal-error analyses without calculations.:)

I will refer to it as the rest of the story...
 
#47 ·
By here, I mean the subject of argument from authrotiy. And by intellectual laziness, I mean the act and danger associated with blindly accepting statements or arguments from those with impressive credentials without examining said arguments or statements on their own merits.

Me personally, when someone of authority speaks I tend to give them the benefit of the doubt and don't fact check every sentence, at some nonzero risk of getting burned. But if something doesn't pass the sniff test, the credentials of the person it came from doesn't matter much anymore. I examine the same as if it came from someone else.

All that is a long way of saying... its OK to question a statement, position, or argument that comes even from a highly respected authority. Its how science is done. Posting a CV does nothing to advance the pursuit of truth, only attempts to obfuscate it.
 
#52 · (Edited)
A fair argument when it suits you. But not wholly a truthful one, now is it.

You are on record as directly listing your credentials and also alluding to them, as to evidence your worth, as a learned type, one worthy of intellectual respect... Your appeals to such have hardly gone unnoticed.

This being said, you are not an EE, nor do you have any published papers with the AES, nor have you made any real contributions to any audio industry as a professional (retail doesn't count).

Perhaps that's how you can so easily swat down the CV presented within this thread, as to suggest that it count for not. But out of the other side of your mouth, call for citation, when faced at every turn, with conflicting perspectives, to your biased view point. The citations attached to each of the papers presented in here, are extensive, perhaps you should roll through them, before pilling on more rhetoric.

You have simply produced another helping of rhetoric, in yet another thread. You have not posted one shred of technical information, just disinformation. If you cannot technically beat down the facts presented, attack the credibility of the author(s)... that is your MO. A terribly, intellectually dishonest MO.:)
 
#49 ·
Do you all know the funny story of why the .05 significance level is such a common one? It is hilarious.
 
#57 ·
Basically the guy who first clearly explained it all to the world wrote something along the lines of this:
"First you have to ascertain what certainty level you want to use as your criteria. Examples might be, say, 1 in 5, 1 in 10, 1 in 20, 1 in 30, 1 in 40, 1 in 50, 1 in 100, etc. It doesn't really matter; you can use whatever you want.. . . . Here, allow me to explain with a specific example where I show you the math. Randomly let's say you want to use, oh I don't know, 1 in 20 . . ."

And that, my friends, is why science so often uses the 5% level! :eek: It was a fluke!
 
#51 ·
It is funny how critics of DBTs often find all sorts of faults with it which are actually of testing in general, foolishly not realizing that their gripes would equally apply if science, um, converted to sighted studies instead.
 
#59 ·
Not this one, not today, it's 40 pages, and is somewhat complex. Its a must read twice type of publication, to get your head around it. Not for the light hearted.:)
 
Status
Not open for further replies.
You have insufficient privileges to reply here.
Top