24/192 Downloads ... and why they make no sense?

[blueheronhollow]

Hi All,

 

Not sure if you've seen this article, and couldn't find anything when I searched the site, but it's sure to be controversial:

 

http://people.xiph.org/~xiphmont/demo/neil-young.html

 

What do you think?

 

Cheers,

Ian

 

[cclay]

I was just about to post the same link. It's certainly a countervailing opinion. I've bought plenty of high def audio and often find it better than redbook, but the author makes the strong point that it may be a matter of mastering rather than bit rates and depth alone.

 

The author's effort to correct misunderstandings of sampling theory challenge several of my conceptions or misconceptions about digital audio too, though some of the ultrasonic stuff was beyond my understanding. The analogy to eyesight has me piqued too. For sure, I expect confirmation bias does not get enough critical attention, in this hobby and elsewhere.

 

Among the more provocative reads for me lately!

 

[The Computer Audiophile]

Very interesting read. I really wish something like this could be debated by very knowledgable professionals on both sides.

 

[esldude]

I generally would agree with most of this article.

 

However, not with everything. For instance, for straight playback sure 16 bit is enough. However, when doing digital volume control and more so when doing things like room/speaker correction, it isn't much different than the needs of mastering on the other end of things. 24 bit can be helpful and certainly doesn't hurt. I do agree that I think there is little to be gained beyond 48 khz done well.

 

I guess my other point of contention would be the comments about headphones. Sometimes true, usually not. In my opinion headphones are more akin to looking at something with a magnifying glass. Those can have obvious distortions and color shifts from accurate reproduction. Yet you can still see more than without the magnification. Headphones aren't more accurate they are just aural magnifiers in ways even very loud playback over speakers aren't.

 

[Julf]

Agree about needing extra headroom for DSP operations, but that is a temporary need. You don't need 24-bit material in order to have the headroom, you just need the processing done in 24 or 32 bits.

 

[pacwin]

is not balanced by countervailing evidence/data. This paper was extensively discussed on the sa-cd.net forums:

 

http://www.sa-cd.net/showthread/42987

http://www.sa-cd.net/showthread/58757

 

Moran participated at various points. Essentially the chosen SACD'S were based on low resolution sources and were not a fair test.

 

This article is highly selective in its sources and makes no attempt to evaluate contradictory data. It reads like a fundamentalist manifesto.

 

[Julf]

"This article is highly selective in its sources and makes no attempt to evaluate contradictory data. It reads like a fundamentalist manifesto."

 

Not sure I would call it a fundamentalist manifesto - it is clearly a position paper / opinion piece, but well written and well argued. I do think it is a good basis for a discussion.

 

[ted_b]

high resolution, nor even a good representation of attempts to remaster to decent hirez.. Let's start there! And then they choose PCM-based SACD's which add insult to injury (i.e it's one thing to question average recordings remastered in PCM, but to then use as examples average recordings, mastered in PCM, then sampled/transferred to DSD...or even vice versa?) Seems skewed at the very least.

 

[audiozorro]

I knew from the conclusion in title of this article that I would disagree with much of what the writer has to say. So rather than follow his path of fallacies and opinions, I went straight to his recommendations.

 

Finally, the good news

 

What actually works to improve the quality of the digital audio to which we're listening?

 

Better headphones

 

The easiest fix isn't digital. The most dramatic possible fidelty improvement comes from a good pair of headphones. Over-ear, in ear, open or closed, it doesn't much matter. They don't even need to be expensive, though expensive headphones can be worth the money.

 

Keep in mind that some headphones are expensive because they're well made, durable and sound great. Others are expensive because they're $20 headphones under a several hundred dollar layer of styling, brand name, and marketing. I won't make specfic recommendations here, but I will say you're not likely to find good headphones in a big box store, even if it specializes in electronics or music. As in all other aspects of consumer hi-fi, do your research (and caveat emptor).

 

Really, do the majority of readers here feel that the number one thing to improve the quality of the digital audio to which we’re listening to is better headphones? The most dramatic possible fidelty improvement comes from a good pair of headphones. Over-ear, in ear, open or closed, it doesn't much matter.

 

Of course I prefer better headphones to bad sounding headphones for all music listening. But his first recommendation for folks to follow to improve the quality of the digital audio is proof enough for me that we are not on the same page.

 

So let me not waste any bandwidth and time on this false logic and nonsense. For me it’s not worth the time and effort to argue his positions, though the writer expects many to do so. I prefer to go back to listening to the critically acclaimed 24/192 Equinox album and let others wallow in audio nonsense.

 

 

[wgb113]

Any chance we can merge the two threads?

 

Bill

 

 

[exa]

24/192 Downloads ... and why they make sense? It is very easy to find out for yourself. Just download the same recording in various sampling rates and compare them. You don't need to have a golden year to notice the difference. 44.1kHz/16bit sounds plain and ordinary when played next to the 192kHz/24bit version of the same track. The difference comes from the incredible ability of the high-resolution recording to reproduce fine nuances and details.

 

Let's face it, the CD medium and format are outdated and high-resolution downloads are the way of the future.

 

 

[One and a half]

then let's read on:

 

http://www.physics.sc.edu/~kunchur/Acoustics-papers.htm

 

http://www.physics.sc.edu/~kunchur/papers/FAQs.pdf

 

http://www.physics.sc.edu/~kunchur/p...n--Kunchur.pdf

 

Kunchur makes a strong case for high resolution

formats since CD sampling rates can only resolve

timing differences down to 11?s, but clearly we need

much higher resolution playback at least 24/192kHz to

reach levels where time smearing will be

inaudible. This is especially the case when listening

to percussion where transients lasting less than 10?s

are common. The Red Book CD standard was never

good enough from this particular viewpoint.

 

http://www.physics.sc.edu/~kunchur/papers/HIFI-Critic-article-by-George-Foster.pdf Page 9

 

While he reads that, gonna listen to 192/24 2L awesome recordings with crystal clarity, obviously high sampled MP3's

 

[Julf]

OK, but don't you then also need a tweeter that goes to 96 kHz? Otherwise you just transfer the location of the time smearing from the DAc to the speaker...

 

[Akapod]

And if I'm wrong, I would love to be corrected.

 

A higher sampling rate allows two things. The first, an increase in the Nyquist number, doesn't really matter, as our ears just can't hear frequencies above 20 kHz (and that's on a good day, when we're young.) But a higher sampling rate also allows for greater precision in timing, and our ears actually do really, really well here, and that ability doesn't degrade as we age. (See this very cool New Yorker article for more on this.)

 

As an anti-aliasing/low-pass filtering would be used, there would be no sounds at 96 kHz. In fact, let's say say that the filter is set at 48 kHz. But those samples would still be recorded 96,000 times a second, which would help with the temporal resolution.

 

(And if the low-pass filter were set at 21 kHz, a scan of the file would reveal no frequencies above that number, even though the file was recorded at a higher sampling rate).

 

I'd love to hear someone with more expertise comment on this.

 

[chrisg3]

I wonder if author of that article actually listened to any high resolution music at all.

I trust my ears and feelings more than science or rather some scientific facts joint cleverly together in this case. I listened to enough 24/192 (and 24/96) downloads to know they sound different and better than equivalent 16/44 music. Then, I am positive, 99% people on this forum would agree with me. Sure, interesting reading at first but waste of time after all.

 

 

[Teresa]

exa said "The difference comes from the incredible ability of the high-resolution recording to reproduce fine nuances and details."

 

Exactly, however these so-called scientists measure and quantify, I doubt they ever actually listen to music for enjoyment.

 

The paper "24/192 Downloads ... and why they make no sense?" claims that a 16 bit word length is plenty for most music and that 24 bit wastes space. However they are only looking at "raw" dynamic range which is only one of the benefits of the higher resolution afforded by a greater word length. 24 bit actually has 256 times the resolution of 16 bit and I and many others feel it is even more important for realistic sound quality than sampling frequency.

 

In their 16/24 bit comparisons what they are focusing on is called "macro-dynamics" or "large scale" dynamics which equals the raw dynamic range.   The important part to me is "micro-dynamics" or "small scale" dynamics which a larger bit word with more values equals increased intricate sonic details, smoother and more comfortable sound.

 

In addition they are missing the reality that real musical instruments have frequencies up to 102.4 KHz http://www.cco.caltech.edu/~boyk/spectra/spectra.htm and these ultrasonic frequencies which are not in the normal audible range are important to the ultimate realism of the captured music. These ultrasonic overtones effect the frequencies downstream in the audible range and according to the latest theories can be felt on our skin, much like inaudible deep bass is not heard but felt in our bones. Take them away and the music is diminished. This they would know if they only bothered to actually listen to music instead of just measuring it!

 

Thus to me the paper "24/192 Downloads ... and why they make no sense?" makes no sense.

 

[KDinsmore]

Is this just a jealousy/envy thing because some don't own a good DAC? You'd have to have a good system to begin with plus have a great DAC to get to the 24/192 level. Then you're getting into USB?/SPDIF/Firewire debate, the quality of the recording.......blah..blah....blah..

 

I bet some perceive this as elitist: Someone has a DCS big rig and is enjoying audio nirvana and you don't. They have more money than you do, can listen to music the way you'd like to but can't so they must be wrong.

 

You can buy a Porsche with Ceramic Brakes at a large extra cost. When you push your foot to the brake pedal both the Ceramic/non-Ceramic cars will both stop! Those Ceramic brakes are even squeaky if you're just going to pick up some groceries. But if you get the car up to speed, stop again and again the Ceramic's are better. But you'd actually have to own the car to know that.

 

 

[Julf]

As someone wrote on another forum: "Science is for losers who aren't rich enough to buy insanely expensive gear".

 

[Julf]

"And if the low-pass filter were set at 21 kHz, a scan of the file would reveal no frequencies above that number, even though the file was recorded at a higher sampling rate"

 

But at the same time, that low-pass filter would also remove those "less than 10?s" percussion transients.

 

 

 

[Julf]

"however these so-called scientists measure and quantify, I doubt they ever actually listen to music for enjoyment."

 

And if they don't, does that make their observations and measurements less valid?

 

"These ultrasonic overtones effect the frequencies downstream in the audible range and according to the latest theories can be felt on our skin, much like inaudible deep bass is not heard but felt in our bones. Take them away and the music is diminished."

 

Any references?

 

[Akapod]

Not following you; please help me out.

 

A low-pass filter set at 21 kHz would eliminate any part of a signal whose frequency is at or above 21 kHz. A 10 microsecond transient with a frequency of 1 kHz would survive -- it would just be a very short sound.

 

 

 

 

 

[Downrange]

I'm sorry, but that reproducing chain they used barely qualifies as "mid-fi." (12 gauge stranded speaker cables, anyone? Kiss of death, right there.)

 

The thing I've noticed as I've made changes and improvements to the equipment in my systems is just how important little things can be. The final arbiter of a hi-res source's desirability is how it sounds on YOUR equipment. As an example, just in the past few weeks I've added the V-Link 192 to the chain in the office system. This made an immediate improvement in the sound of my CD rips (into FLAC). They still have "markers" for the 16 bit sound that I can hear in direct comparison with hi-res downloads of the same masters, but they're much, much closer to "good sound." This tells me that even the really very good V-Link II was not passing all the musical information (free of stuff that should NOT be there) that the new 192 model does. This is something of a revelation for me and helps me understand those who still defend CD as a viable medium. It can sound really decent when played back on superior equipment.

 

 

[Julf]

"A 10 microsecond transient with a frequency of 1 kHz would survive -- it would just be a very short sound."

 

Very short - considering 10 microseconds is 1/100th of even one wave of 1 kHz. A 10 microsecond transient has no components with a frequency lower than 50 kHz, so no, it would not get past a 21 kHz filter.

 

[Daphne]

The article in question is nothing more than a personal opinion. The author presented and accentuated carefully chosen facts that support his opinion, and “only his opinion.” This is not a research paper, because the author himself is the only source of reference on most of his footnotes. Who writes a paper on a technical subject and refers to himself, more often than not, as the only source of reference.

 

Is this paper a manifesto? Absolutely not, but one could easily describe it as a well organized rant.

 

However, it was posted on www.xiph.org which could easily be described as a modern day Internet Art Manifesto. http://en.wikipedia.org/wiki/Art_manifesto

 

So, I have always been a curious person, and it’s in my nature to question things that seem out of place. What relevance does a personal rant titled “Why 24bit/192KHz Music Downloads Make No Sense” serve posted on a .org website “dedicated to protecting the foundations of Internet multimedia from control by private interests?”

 

Well, gentlemen, I can string a thread of logic as well as most people. Since the article was posted on a .org seeking donations for a cause in a time of global economic strife, what better target audience to hit than those with disposal income.

 

Audiophiles in pursuit of top quality recorded music to satisfy their obsessive/compulsive disorder tend to have enough disposable income to acquire expensive audio systems. Instead of spending money on downloads which make no sense, consider making donations to an organization with a worthy cause such as free, open protocols and software that serve the public, developers, and business markets. This type of rant will surly draw attention and controversy, which in turn will lead those with computer technical backgrounds and interests with disposal income to the organization’s website. A most suitable and perfect target audience. Links to the article will be posted on the many audio related websites thus increasing traffic and creating quality back links improving the overall SEO of the website, and becoming an effective Internet marketing strategy.

 

So, what do you think, does my string on logic make more sense than music downloads which make no sense? Have we been led by the ears? Just curious. ;-)

 

 

[Akapod]

Can you walk me through what you just did? I'm not following.