Does High Resolution Audio sound better

[The Computer Audiophile]

USB cables do not color the signal unless the cable is defective.

Lets not turn this into a USB cable discussion. Please start another thread if this discussion is desired.

[wakibaki]

Thank you for telling us. We never would have guessed that from your seven previous, consecutive posts. FWIW = not very much.

So you don't hear a difference. Fine. Save your money and don't buy hi res downloads. Others most certainly have, do, and will continue to hear a difference. Do you really think that you are bringing some brilliant new insights that have not been seen before? FYI, we already have members - it's not difficult to identify them - who have raised and continue to raise the same, tired, boring arguments regarding hi res time and again.

 

You are adding nothing new, and you most certainly are not going to convince anyone who enjoys the benefits of hi res and disagrees with you.

 

What you have here is a comparatively small number of hi-res enthusiasts congregating locally and drowning out any rational discussion in threads like this.

 

You have to remember that statistically you are only entitled to a tiny fraction of the bandwidth. Your views are not representative of the population as a whole. As long as it is an open forum, you will get people coming along from time to time and trying to reconnect you to reality. And you wonder why you have a problem attracting membership?

[lamode]

Others most certainly have, do, and will continue to hear a difference.

 

Yet they can't pass a proper blind ABX test.

 

I wonder how many of the self-appointed golden ears have performed such tests. The power of expectation bias is not to be ignored. From my experience, such people will flat out refuse to test their beliefs for fear of being proven wrong.

 

If someone feels more comfortable knowing that his files are high res without being able to hear a difference, then that's fine - it's his money - but he should be honest that that is the reason he pursues hi-res. I haven't seen any studies able to show a significant pass rate comparing 16/44 with higher-res formats.

 

Example: http://drewdaniels.com/audible.pdf

[wakibaki]

Do you really think it is possible to discuss this issue against a background of people discussing their anecdotes as though the matter has been settled in their favour?

 

This is a poor substitute for evidence with provenance, specially when all the evidence shows this not to be the case. It doesn't even compete with rational discussion of the balance of probabilities.

[The Computer Audiophile]

Yet they can't pass a proper blind ABX test.

 

I wonder how many of the self-appointed golden ears have performed such tests. The power of expectation bias is not to be ignored. From my experience, such people will flat out refuse to test their beliefs for fear of being proven wrong.

 

If someone feels more comfortable knowing that his files are high res without being able to hear a difference, then that's fine - it's his money - but he should be honest that that is the reason he pursues hi-res. I haven't seen any studies able to show a significant pass rate comparing 16/44 with higher-res formats.

 

Example: http://drewdaniels.com/audible.pdf

Hi lamode - You may be interested in the debunking of the Meyer & Moran paper you cited. I can't find the links right now (on my mobile), but some are very interesting.

[mmerrill99]

Yet they can't pass a proper blind ABX test.

There have been a number of proper positive ABX test results published showing that a high-res file resampled to 16/44 is audibly discernible but ask Waki what he means by "evidence with provenance" for your get-out clause which questions the honesty of the tester.

 

Furthermore, the AES paper by Robert Stuart further shows that there is a discernible difference between high-res & RB due to "The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System

[wgb113]

Again, the point of hi-rez to me is that it takes the entire argument about better out of the equation, it's one less thing to "worry" about. If an album was mastered in 24/96 and I can buy that quality file I have exactly what I want - the highest possible quality source material available for that recording.

 

Bill

[firedog]

Yet no-one has ever passed a 16/44 v HD audio ABX test by a statistically meaningful margin.

 

Simply not true. What is true is that every time someone does, people who refuse to believe it reject the test.

[lamode]

Simply not true. What is true is that every time someone does, people who refuse to believe it reject the test.

 

Please share these studies.

[mmerrill99]

Simply not true. What is true is that every time someone does, people who refuse to believe it reject the test.

Indeed, & the latest & greatest excuse is that anybody now doing an ABX test needs to have some "trustworthy" individual administer/oversee the test. You can see where this leads - how do you define trustworthy - it's the perfect get-out excuse for not accepting results that are unpalatable.

 

It's interesting how, for more than 15 years, negative results have been accepted without question by this same group - no accusations of dishonesty, no calls for proctoring.

[Tony Lauck]

...but the quote you selected didn't mention dithering. And the author is right that 96dB is the usual quoted figure.

 

If you do not dither 16 bit audio, there will be horrible distortion on reverb tails as notes die down. In addition to this obvious distortion, the echo will suddenly die out. Quoting noise figures for an improper realization of the format is incorrect. And yes, most people use the number of 96 dB, but they are wrong.

 

 

Evidence.

 

Noise shaped dither at 44 kHz is marginal, because the noise remains within the range of human hearing and the range of playback gear. Some people perceive "inaudible" high frequency noise as harshness, particularly if it varies with the musical signal. This may be a function of individual listeners and individual playback systems. The noise is still there, it's just been swept under the rug. (The situation is different when noise-shaping is used with oversampled systems such as DSD, because the bulk of the shaped noise is filtered out in the playback process by a low pass filter.)

 

What do you mean by "quiet"? The point of the article still stands - the DR of a CD exceeds the requirements of domestic listening rooms.

 

Speak for yourself. The dynamic range of 16 bit audio is adequate for mid-fi reproduction. It is not adequate for high fidelity reproduction because when music is played at live concert levels the peak SPLs are at least at 105 dB and human hearing is sensitive down to 0 dB at some frequencies. To get sufficient quiet I do have to turn off all my appliances. I live in a rural area, with no road traffic closer than 300 feet.

[Allan F]

Hi lamode - You may be interested in the debunking of the Meyer & Moran paper you cited. I can't find the links right now (on my mobile), but some are very interesting.

 

Chris, here are links to two CA threads related to this topic:

http://www.computeraudiophile.com/f8-general-forum/meyer-and-moran-debunked-meridians-robert-stuart-22105/

http://www.computeraudiophile.com/f8-general-forum/was-meyer-and-moran-debunked-robert-stuart-22675/

[Allan F]

Allan,

 

Give a chance to a newbie to understand that in CA the only format that can be bashed is 16/44:)

 

Dennis,

 

Hi-res is a religion and if you don't surrender you are called troll

 

Of course, alfe, the forum is in dire need of more anti-hi res evangelism. The enjoyment people get from listening to hi res is clearly diminished if they can't prove the difference to others.

[lamode]

Chris, here are links to two CA threads related to this topic:

http://www.computeraudiophile.com/f8-general-forum/meyer-and-moran-debunked-meridians-robert-stuart-22105/

http://www.computeraudiophile.com/f8-general-forum/was-meyer-and-moran-debunked-robert-stuart-22675/

 

It's hard to take Bob Stuart seriously as an objective bystander as his company is pushing high-res digital as much as anyone (right now they are trying to launch their new MQA format, for example).

 

But I'll still keep an open mind and read it if I can find a copy.

[sdolezalek]

Yet they can't pass a proper blind ABX test. Example: http://drewdaniels.com/audible.pdf

 

If, in 2015, you are going to post that paper, then you also need to post the following:

 

The BAS Study Revisited | Real HD-Audio

 

BAS Experiment Explanation page - Oct 2007

 

Conclusive "Proof" that higher resolution audio sounds different

 

AES E-Library » Sampling Rate Discrimination: 44.1 kHz vs. 88.2 kHz

 

https://www.jstage.jst.go.jp/article/ast/26/1/26_1_50/_pdf

 

"P14-3 The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System—Helen M. Jackson, Meridian Audio Ltd. - Huntingdon, UK; Michael D. Capp, Meridian Audio Ltd. - Huntingdon, UK; J. Robert Stuart, Meridian Audio Ltd. - Huntingdon, UK

This paper describes listening tests investigating the audibility of various filters applied in high-resolution wideband digital playback systems. Discrimination between filtered and unfiltered signals was compared directly in the same subjects using a double-blind psychophysical test. Filter responses tested were representative of anti-alias filters used in A/D (analog-to-digital) converters or mastering processes. Further tests probed the audibility of 16-bit quantization with or without a rectangular dither. Results suggest that listeners are sensitive to the small signal alterations introduced by these filters and quantization. Two main conclusions are offered: first, there exist audible signals that cannot be encoded transparently by a standard CD; and second, an audio chain used for such experiments must be capable of high-fidelity reproduction.

Convention Paper 9174"

 

And this: AVS and AIX Records performed a series of experiments last year comparing 24/96 and 16/44 music. 56 people took the test, and the results were very interesting. Of those with non High Resolution Audio systems (ie: 16/44 reproduction quality) the correct identifications of the high resolution tracks had a distinct downwards trend with increasing identifications, and showed no clear ability to identify the high resolution tracks. However, those with High Resolution Audio systems (better than 16/44 quality) showed a contrasting upwards trend with increasing identification.

 

Having said all of that, I put myself in the camp of not necessarily being able to distinguish between 16/44 and 24/192 when both are upsampled and then run through a D/A conversion process at 24/384 so as to allow for better filtering. I don't belueve this is about the audibility of frequencies above 20kHz, I believe this is all about the A/D and D/A and other processes to which the music is subjected between microphone and my ears and I do believe that when those processes are conducted on higher resolution material they are less destructive and therefore have a better chance of delivering the best result to my ears!

[wakibaki]

I'm having difficulty understanding why I'm being told to reconsider scientific orthodoxies established for 150 years or so, but I get pilloried for questioning new orthodoxies introduced only yesterday.

 

The idea than frequencies above 20k are inaudible has a lot more support than just Meyer and Moran.

 

The last set of ISO standards showed equal loudness contours to ~16k.

[wakibaki]

There have been a number of proper positive ABX test results published showing that a high-res file resampled to 16/44 is audibly discernible but ask Waki what he means by "evidence with provenance" for your get-out clause which questions the honesty of the tester.

 

Furthermore, the AES paper by Robert Stuart further shows that there is a discernible difference between high-res & RB due to "The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System

 

You're constantly bitching at me about accepting things without questioning.

 

No-one should really accept this paper at face value without being familiar with some of the reservations expressed here.

 

The Audibility of Typical Digital Audio Filters in a High-Fidelity Pla - Hydrogenaudio Forums

[mmerrill99]

You're constantly bitching at me about accepting things without questioning.

Am I?

 

No-one should really accept this paper at face value without being familiar with some of the reservations expressed here.

 

The Audibility of Typical Digital Audio Filters in a High-Fidelity Pla - Hydrogenaudio Forums

 

You mean the objections by Arnie Kreuger who can't resample files correctly as was shown in his Jangling Keys fiasco & referred to by Don Hills "There have been several of these format comparison file sets produced recently, by people who should know how to do it right, which have proved to be not as well converted as they could have been."

 

You mean the objections by Arnie Kreuger who cheats an ABX test & posts the null results for all to see?

 

You mean the Arnie Kruger who posts his objections to the Stuart paper here & is answered by Stuart?

[wakibaki]

Well, of course if you want to judge it on the basis of personalities, that's only one of the behaviours I'm objecting to.

 

I try to detach myself from issues of personality where technical argument is concerned.

 

You do realise that if there is a breakthrough in this area, hydrogenaudio will be the place on the internet that it will be confirmed?

[mmerrill99]

Well, of course if you want to judge it on the basis of personalities, that's only one of the behaviours I'm objecting to.

 

I try to detach myself from issues of personality where technical argument is concerned.

I'm not talking about Kreuger's personality - I'm talking about his technical abilities & credibility when it come to making technical criticisms in areas he has shown himself to be incompetent & dishonest

 

You do realise that if there is a breakthrough in this area, hydrogenaudio will be the place on the internet that it will be confirmed?

What you talking about, now, Waki - that forum is the church to renew your beliefs in the so-called objectivist religion. They will be the last to accept any breakthrough - claiming, like you do, dishonesty in those who present evidence that is unpalatable to their religion

[wakibaki]

Anybody who thinks they've found credible evidence posts it there first, because there's no kudos in getting a rave review at WBF.

[Jud]

Unfortunately, it is not simple to re-sample between formats where the sampling rate differs by a non-integer ratio. This process is imperfect and there are different algorithms used, which is what Chris was referring to.

 

For what it's worth, I am in the camp defending 16/44. I truly wish that HD audio added some magic - any audiophile would - but the reality is that it makes no difference and there are other much weaker parts of the reproductive chain which CAN be improved and this is where people should be spending their money if they want better sound.

 

And I enjoy LP sound too but never thought of it as 'accurate' as 16/44.

 

16/44.1 hasn't been supported by manufacturers since the 1980s, so depending on your age you may never have heard it. Between introduction of the first CD player in 1982 and the late 1980s, the industry standardized on 8x oversampling. Except for a tiny, tiny handful of DACs, every DAC made since then has a chip in it that raises the sample rate to 352.8 or 384kHz. Since the 1990s, these chips have used sigma-delta modulation, meaning after the 8x oversampling the bitstream is converted to a DSD-like format at mHz rates, which is then converted to analog with a relatively simple filter.

 

Looking at what actually occurs in the DAC, and correspondingly during ADC, it becomes evident that most people in the high-res "debate" are asking the wrong questions, and therefore unsurprisingly coming up with the wrong answers.

 

ADC first: Other than a couple of Pacific Microsonics decks in obscure places, the first step in analog to digital conversion at the studio is sigma-delta modulation. Only a few companies use this sigma-delta modulated (SDM) bitstream from the ADC. Most configure the ADC to output a PCM bitstream (my impression is that 24/96 is a common resolution), so in nearly all the music you listen to, unless you patronize the few small companies that take the SDM output from the ADC, that's the first format change right in the ADC. When you buy a CD or a 16/44.1 download, you're buying something that's been downsampled, or "decimated," from that first conversion to 24/96. So that's two conversions.

 

Then DAC: Nearly all DAC chips will do the 8x oversampling in 2x steps: 44.1 -> 88.2 -> 176.4 -> 352.8. So that's three more conversions. After that, as noted above, there's another format change to a DSD-like bitstream at mHz rates before the actual digital to analog conversion. Thus when you're listening to something you purchased in 16/44.1 format, between the A/D conversion and the D/A conversion there are six sample rate/format changes.

 

It's not only not simple to re-sample between formats where the rate differs by a non-integer ratio; mathematically, it's not simple (in fact, it's impossible to do without introducing distortions) even where the rate differs by an integer ratio. So conceptually, it would be a good thing to eliminate some of these conversion steps. If you purchase a recording in a 24/96 format, where 24/96 was the format the ADC was configured to output, you'll eliminate two steps: (1) decimation from 24/96 to 16/44.1; and (2) you'll have two rather than three doubling ("interpolation") steps in the DAC: 96 -> 192 -> 384.

 

When you're asking whether a difference can be heard between 16/44.1 and, for example, 24/96, what you're really asking is whether the elimination of those two steps is audible to you. This depends on the audibility of the decimation filtering at the recording end; the interaction between the ADC filtering/conversion and your DAC's filtering/conversion; on the particular interpolation filtering used by your DAC; and on the audibility of the sigma-delta modulation step in your DAC. (Or better, it depends at least on those things.) The interaction of all these factors is complex, and thus the answer will not be simple or uniform across recordings and DACs. When trying to test something this complex with blind tests that depend at least in part on echoic memory that starts decaying within a couple of seconds, it's not at all surprising to get results no better than chance in most instances. But realize that what you're comparing isn't sample rates or formats, it's filtering and conversions. Most people have no idea what differences in filtering and conversions sound like, and therefore what to listen for, so there's no way to nail such a difference with a sample lasting just a couple of seconds. But if you listen to longer samples, echoic memory goes. Catch-22.

 

Edit: Here's the thing, though - if, like Tony, Barry, JonP, etc., you've played around with filtering, you've heard differences in sound. These differences are measurable. If you've done enough of it, you can correlate the measurements and the audible differences. Or to shortcut things, you can try out hardware or software that offers a choice of filters. So there are audible/measurable differences. But you're not going to be in a position to hear them unless they override whatever other filtering/conversion is going on at the ADC and/or DAC.

[wakibaki]

Very good Jud, but none of this speaks to the superiority or otherwise of RB as a distribution system. It all hinges on whether the differences are audible, not whether there's a possibility that skipping 2 steps might make a difference.

 

I wish you'd confine yourself to less prolix posts, but I suppose that's all part of trying to hide the concept that you're trying to palm.

 

Maybe that's a bit unfair, but it's a trait of lawyers to become entranced by a good argument, to the point of considering the truth irrelevant, but I guess in a crowd of audiophiles you're not going to stick out.

[lamode]

Speak for yourself. The dynamic range of 16 bit audio is adequate for mid-fi reproduction. It is not adequate for high fidelity reproduction because when music is played at live concert levels the peak SPLs are at least at 105 dB and human hearing is sensitive down to 0 dB at some frequencies. To get sufficient quiet I do have to turn off all my appliances. I live in a rural area, with no road traffic closer than 300 feet.

 

There are 2 problems with what you wrote.

 

1) Yes, the human ear can hear noises at 0dB at certain frequencies, but how many recordings capture music at this level? None. Even a classical recording will typically not use more than 30dB of DR, which means that the quietest parts are still around 66dB above the noise floor. The noise is rendered insignificant.

 

2) Even if you live in an anechoic chamber in the middle of a desert, you can't escape noise in the recording itself. You will never achieve a S/N ratio better than or even equal to the microphone used in the recording. Let's take a typical studio microphone like the Neumann KM184:

 

Signal-to-noise ratio, CCIR¹⁾ (rel. 94 dB SPL): 72 dB

Signal-to-noise ratio, A-weighted¹⁾ (rel. 94 dB SPL): 81 dB

(https://www.neumann.com/?lang=en&id=current_microphones&cid=km180_data)

 

Who needs 144dB of DR when you are using microphones with a S/N ratio of 70-80dB?

 

(Btw, 300 feet from a road is nothing. As you know, traffic noise can travel a mile or more in rural areas. I've experienced this at my Dad's farm).

[Jud]

Very good Jud, but none of this speaks to the superiority or otherwise of RB as a distribution system. It all hinges on whether the differences are audible, not whether there's a possibility that skipping 2 steps might make a difference.

 

I wish you'd confine yourself to less prolix posts, but I suppose that's all part of trying to hide the concept that you're trying to palm.

 

Maybe that's a bit unfair, but it's a trait of lawyers to become entranced by a good argument, to the point of considering the truth irrelevant, but I guess in a crowd of audiophiles you're not going to stick out.

 

I was actually trying to speak to the topic that holds some interest, rather than trying to "hide" the original issue. The answer to whether there's an audible difference between 16/44.1 and high res is trivial and was answered more than 25 years ago. The reason 8x oversampling became a standard is that going direct to analog from 16/44.1, and to a lesser extent from 88.2 (2x oversampling was tried for a short time before 8x oversampling took over), creates audible distortion. So an entire industry - virtually all audio engineers working in the field at the time - agreed 16/44.1 was simply unworkable, 88.2 was no picnic either, and 352.8/384 was where we should be.

 

Back then storage and transmission wouldn't allow distributing 352.8/384, so we got the present-day kludge of downsampling at the recording end, then upsampling at the consumer end. That's been pretty well set in stone until now just from sheer economics: due to the fact everyone's doing it so you have economies of scale, it's a hell of a lot cheaper to stay with the same distribution model. But if you were starting with a clean sheet of paper, these days you'd have the music companies sending us DSD or DXD (352.8/384) instead of fiddling with all these conversions.