Proven: Good Old Redbook CD Sounds the Same as the Hi-Rez Formats

Everything Else
21
Elk said: The high resolution signal went around the converter. This *favors* SACD and would increase the likelihood a difference would be detected between SACD and 44/16.

Exactly. Imagine, you add a new component to the audio chain. Something like a preamp. Will you notice a difference? I bet yes. So will I. But they did not.

Was there anything that "masked" the most obvious difference, let alone subtle ones?
22
Alekz said: Was there anything that "masked" the most obvious difference, let alone subtle ones?

Nope. Nothing in the equipment or signal chain to filter out SACD fairy dust. Unless Nordost cables do not pass DSD playback.

The purpose of the study was to dispassionately test SACD's proponents' claims that SACD is clearly superior. This includes the "most obvious difference," as well as subtle ones. Even with the added A/D/A hardware and cables in place on the 44/16 signal chain, no one could hear the "obvious difference," let alone subtle ones.

Four different excellent high end systems (plus headphones), skilled listeners, plenty of time to go back and forth between sources at will, opportunity to bring in your favored SACDs - yet no discernible difference could be heard.

I understand wanting to believe that DSD is clearly superior. Yet, every time this belief is objectively put to the test no one can identify this "superior source." Without AX testing, the transition between auditory perception and imagination becomes quite narrow. :)

(The only time listeners could reliably tell the difference was at silly high playback levels when the higher noise floor of 44/16 could be heard. As we all know, this is incredibly loud and not at a level anyone would actually listen, unless will to risk hearing damage.)
23
jamesie said: If they had taken an actual analog source (such as real to real master tape) and converted it to 44.1kHz, 96kHz and also 192kHz I would think they may have found/heard differently.

Perhaps. Many of us like to think so. :)

The only similar study of which I am aware is where they did exactly this (precisely the same analog to digital) with 176.4/24 and DSD. This was a comparison of these two encoding schemes. No one could reliably discern one from the other. http://old.hfm-detmold.de/eti/projekte/diplomarbeiten/dsdvspcm/aes_paper_6086.pdf

jamesie said: 96Khz being the best in terms of headroom and background noise.

Yes, 24-bit provides a greater noise floor and allows one to record with peak levels at -20dbFS which provides a great deal of headroom, making setting levels easy.

24
Elk said: I understand wanting to believe that DSD is clearly superior.

No, I believe that an extra A/D/A converter (or converters: ADC and DAC ) with added cables must be audibly noticed in the chain.
25
Elk said: I understand wanting to believe that DSD is clearly superior.

No, I believe that an extra A/D/A converter (or converters: ADC and DAC ) with added cables must be audibly noticed in the chain.

The whole thing seems puzzling. On the one hand, the process seems to have been very thorough as these things go. On the other hand, as Alekz suggests, you would have thought there would have been an audible difference even if they had started with red book CDs just because of the additional equipment.

I have bought more PCM high res material than I probably should have. Some of it clearly beats my CD copies, some doesn't. For those high res downloads that are clearly better, I've never been sure whether it was the higher sample or bit rates, or just differneces in the mastering choices (e.g., how much dynamic compression) or equipment used. If red book CD really can sound just as good as high res, it says a lot about the mastering of most CDs, and none of it good.
26
stevem2 said: Some of it clearly beats my CD copies, some doesn't.

If in doubt I usually check the spectrogram. Some of the so called hi-rez recordings have a clear cut-off at 22kHz :D
27
Alekz said: I believe that an extra A/D/A converter (or converters: ADC and DAC ) with added cables must be audibly noticed in the chain.

Audiophile belief and faith is why such tests are enlightening. Audiophiles "believe" many things, including the efficacy of plastic bags of shiny pebbles. Thus rigorous tests are appropriate. Faith often fails upon scrutiny.

I also appreciate wanting to believe an added A/D/C chain would be audible. But it was not. Good pro equipment will do this. :) Remember also, once again, the tests were conducted with experienced listeners employing far better equipment than most of us own.

stevem2 said: I have bought more PCM high res material than I probably should have. Some of it clearly beats my CD copies, some doesn't.

The researchers allowed subjects to bring in their favorite SACDs, those the listeners were convinced were clearly superior. They still could not detect a difference.

(There certainly are plenty of SACDs which are mere transfers of 44/16. We hope that at least the Chesky and Kimber discs they used as part of the software tested were legitimate SACDs. :) )

Keep in mind the audibility of ultra-high frequencies has also been investigated. There is no evidence such frequencies are audible and/or make any difference in sound perception. Kal Rubinson has written on this (he has both the audiophile and medical chops to do so - plus he loves SACDs. He finds high frequency audibility theories absolute bunk.)

stevem2 said: . . .you would have thought there would have been an audible difference even if they had started with red book CDs just because of the additional equipment.

This is the danger of assumption and why we test hypothesis.

We convince ourselves tweaks, frequencies outside of human hearing, freezing CDs, green markers, Armor All - all sorts of things improve sound. Yet, when studied with a proper double blind test they are no longer audible.

If these things markedly improve the sound, why do they so readily evaporate? DVD to Blu-Ray is easily detectable on a double-blind. Why does high resolution sound routinely fail?

Why is audio reproduction so fragile that Revelatory! cannot be heard except under exceedingly limited circumstances?

Do we actually hear what we think we hear, is there a psychological aspect favoring high resolution so powerful it affects are perception when we know we are listening to it, is there something so unique about listening to music that the normal method of studying psycho-acoustics suddenly and completely fails?

Is there truly no discernible difference and we are fooling ourselves?

Fascinating to ponder.



28
Elk said: I also appreciate wanting to believe an added A/D/C chain would be audible. But it was not. Good pro equipment will do this.

I bet it was that good pro ABX switch that made A/D/A converter disappear :D
29

Go on, keep poking the bear (or deer in this case), I want to learn more. :smiley: I find Elk’s posts in this thread fascinating.



J.P.

30
wingsounds13 said: keep poking the bear (or deer in this case),

That was never my intention. I express my humble opinion, based on my personal experience with my personal audio system. I'm yet to hear a component (let alone pro-components connected with pro-cables) that I'm not able to notice in my system. Actually, my wife is much better at spotting differences, and I trust her opinion more than my own, because I know what I changed (if I did) in my systems and she does not. Yet she can pinpoint the subtle nuances. E.g. Cat6 vs Cat5 diff was pretty much obvious for her and she never confused their sound.
31

@Elk- Regarding frequencies beyond 20kHz or so, how much energy is produced in this range by instruments outside of cymbals and a few other percussion instruments? The violin seems like an instrument that might be capable, but I have no idea if that information is actually present in the music that we listen to outside of a few select instruments that are usually of intermittent use (at least in orchestral, chamber and choral music).

32
Alekz said: I bet it was that good pro ABX switch that made A/D/A converter disappear

Nice try grasping at yet more straws, but it was only a switchbox comprising two RCA connectors, a total of 4 inches of hookup wire, and a reed relay with 0.2-ohm dc resistance.

Additionally, none of the subjects, including the owner of the audiophile home system in a custom room, felt the addition of these components changed the sound of the SACD audio in any way.

Again, do you really believe the claimed revelatory, "immediately obvious" improvements of high frequency playback are so fragile that only laboratory grade equipment can possible reproduce it? If this was the case, very few people would even be in a position to properly audition SACD playback.

wglenn said: Regarding frequencies beyond 20kHz or so, how much energy is produced in this range by instruments outside of cymbals and a few other percussion instruments?

Most have little energy above 20kHz, but quite a few do produce harmonics or other frequencies above 20kHz. As examples,

The harmonics of a muted trumpet extend to 80 kHz. (2% of the energy)
Violin and oboe to above 40 kHz. (violin, 0.04%; oboe, 0.01%)
An orchestral cymbal extends to 100 kHz. (40% !)

33
Alekz said: I bet it was that good pro ABX switch that made A/D/A converter disappear
:D
I think the emphasis was on the " :D " my dear Wapiti. :)>-
34

That would be great, as it would make more sense. :slight_smile: OTOH, the transparency of the switch is important and was directly addressed by the researchers.



I remember Alekz claiming however that the reason DSD was indistinguishable from 176/24 PCM in a double-blind was the cables used on playback somehow did not pass the DSD analog signal and masked the difference.

35

You can’t measure without:



o- knowledge what to measure

o- proper measuring tools



Human eyes/ears/brain (especially trained ones) are very complicated and sensitive instruments, that are very good at detecting differences, registering events and guessing but not good at measuring or knowing (feeling? hearing? seeing? dunno what word to use here) the reasons.



Here’s an example: I have no idea why network cables sound differently. I have no idea in what area I hear the differences (phase response, amplitude response, frequency response, noise, group time delay, etc.). I have no idea what to measure or if it’s measurable at all with the current tools. What I hear is The Difference and I know my preference.



Again, I hear any new/changed component in my system. And what I find very strange and suspicious is that they did not hear the added A/D/A box. Period. The rest can be only (most likely wrongly) guessed, because none of us discussing the testing method were there.

36

I wonder why they did not simply downsample the tracks on a computer and played on a high-end network player, or record them on a CD-R and play on the hi-rez transport they used in the test using the same audio chain. No switches, no weird A/D/A boxes, simplified design. BTW, everybody can do that at home with the PWD/PWT and let us all know what they hear.



Personally, I prefer upsampled sound. I have no idea how to measure the result, but even when there is no extra information, the upsampled tracks sound better. Also, you can find a corresponding article in Stereophile.



This is yet another reason why I do not trust the research. Sorry, but they wasted time, money and efforts for nothing…

37

And here are the links:

http://www.stereophile.com/how-is-ted-coding-the-fpgaent/upsampling-or-oversampling-letters

http://www.stereophile.com/hirezplayers/814/index.html

http://www.stereophile.com/digitalprocessors/454/index.html
38

No measurements were being made. Alekz, this is a red herring - completely irrelevant.



The single question was whether whether an A/D/A chain inserted int he play back system was audible, even by people that want to hear a difference and believe there is one.



They were unable to discern a difference. This study is a valuable collection of data into human perception.



Alekz said: Human eyes/ears/brain (especially trained ones) are very complicated and sensitive instruments . . .


Complicated, yes. Sensitive, not very. We are easily beat out by many animals, let alone instruments. But again, red herring. Sensitive or not, there was no discernable difference.



Alekz said: . . . none of us discussing the testing method were there.


True. However, I am accurately repeating what the authors describe as their test. I have not speculated on any point. The test protocol and results were published in a peer reviewed journal. All of the information I have cited is readily available.



Alekz said: Personally, I prefer upsampled sound.


Yes, and many do not and prefer non-upsampled playback. Most can also provide detailed theory as to why their preference is valid. :slight_smile:



Similarly, the test subjects all claimed they could hear a difference when playing an SACD. This is what was under test. They were wrong.



The problem your anecdotal report is you, as listener, know what you are listening to. You hear what you want to believe - it must be better and you therefore hear it as better. We all do this.



The hard test data however suggests any stated preference is not based in the sound, but rather on pre-conclusion; we hear what we want to hear. Humans are very susceptible to suggestions, emotional preferences, intellectual conclusions. In fact, nothing is more unreliable than eye witness testimony.



This is why double-blind is routinely employed in psycho-acoustic research. Emotional preferences and belief are taken out of the equation. You do not know what you are hearing and therefore must rely on your ears! Is this not the mantra of the subjectivist?



All studio engineers have had the experience of carefully adjusted EQ on a particular track until it is perfect, hearing every change of the knob. And then they discover the track was not active and no changes were made. Been there. The brain is a powerful trickster.



Alekz said: I find very strange and suspicious is that they did not hear the added A/D/A box.


Yes. You have faith that there must be a difference, based solely on anecdotal experience. When faith is questioned, believers often become “suspicious.” It cannot possibly be true! I believe otherwise! Therefore, the science is wrong! :slight_smile:



Sorry, but the earth really, truly is revolving around the sun. This, in spite of the fact our personal observations and anecdotal evidence tells us exactly the opposite.



Alekz said: . . . they wasted time, money and efforts for nothing…


Only because you do not like the result. >:)



The experiment has held up to scrutiny: the protocol is clean and well designed, the equipment more than adequate, the sample size meaningful, the subjects well-versed in what they were listening for.



(By the way, blind tests are typically conducted where the subject knows what to listen for. The test is not to fool the listener. The test is to determine whether something is detectible. We want them to know what they are listening for.)



The only flaw is you do not like the conclusion. :slight_smile:

39
Elk said: The only flaw is you do not like the conclusion.

You did not read what I wrote here, did you:

Alekz said: I wonder why they did not simply downsample the tracks on a computer and played on a high-end network player, or record them on a CD-R and play on the hi-rez transport they used in the test using the same audio chain. No switches, no weird A/D/A boxes, simplified design. BTW, everybody can do that at home with the PWD/PWT and let us all know what they hear.


40

Because this is not a live test of a single analog stream. They were testing the audibility of an inserted A/D/A loop.



Edit:



I realize this statement implies a couple of things I find self-evident, but may not be apparent to others.



First, real time switching allows random direct comparisons between SACD and 44/16 at the listener’s discretion: switch back and forth, rewind, play again, etc.



Second, the characteristic under test is whether DSD playback is inherently somehow different sounding than Redbook CD and whether Redbook can record and playback this analog stream. That is, does it recreate something different in analog playback that Redbook cannot. One cannot make this determination by transcoding the file prior to playback.



Moreover, even if there are alternative methods this does not invalidate the method used. This is another red herring. Come on, Alekz.



Alekz said: You did not read what I wrote here, did you:


You need even more shot down?