Audio overkill? Some question benefits of 'high-res' music

Audio overkill? Some question benefits of 'high-res' music
In this Jan. 7, 2015, file photo, Musician Neil Young speaks during a session at the International CES, in Las Vegas. Advocates like Neil Young and major record labels say the format that's the high end of what's known as "high-resolution" audio restores textures, nuances and tones that listeners sacrifice when opting for the convenience of music compressed into formats like MP3s or Apple's AAC. (AP Photo/John Locher, File)

Its backers say it does for music lovers what ultra high-definition television has done for couch potatoes.

It's a digital format that packs nearly seven times the data found on CDs, touted as producing crystal-clear sounds with a sharpness that'll blow consumers away. Advocates like Neil Young and major record labels say the format that's the high end of what's known as "high-resolution" audio restores textures, nuances and tones that listeners sacrifice when opting for the convenience of music compressed into formats like MP3s or Apple's AAC.

But some recording-technology experts say this super high-res format—known by its 192 kHz, 24-bit technical specs—is pricy digital overkill, an oversized "bit bucket" that contains sounds only dogs or dolphins can truly enjoy.

Some cynics say the push to high-res audio is just another attempt to get consumers to rebuy music they already own.

Marc de Oliveira did just that in February when he bought Bob Dylan's latest album, "Shadows In The Night" from the Young-backed PonoMusic store. Already having bought the CD from a physical record store, the Copenhagen-based 49-year-old IT consultant splurged on a 24-bit version, hoping to feel more present in the room where Dylan recorded.

Instead, he stumbled on a blog that analyzed the file and found no more than 16 of the 24 bits were used, the same as on the CD. After months of de Oliveira trying to get a refund, Pono's Vice President of Content Acquisition Bruce Botnick replied to his posts saying that Dylan himself liked sample CDs cut in the studio. Engineers mastered the album from those discs, forever locking this particular release at the lower specs.

Still, that hasn't changed what Pono is charging for the file, $17.99, versus the physical CD, which costs $9.70 on Amazon.

"They should have probably been more active about not accepting that as a real 24-bit file," de Oliveira said.

More than 90 percent of the PonoMusic store is represented essentially by digital copies, or rips, of CDs, Botnick acknowledged to The Associated Press in an interview at his Ojai, California-based studio. To be fair, they're labeled as such. And those files are still in a higher category than AAC files or MP3s, which eliminate some sounds in the compression process.

But of the other albums on PonoMusic labeled higher-than-CD quality, Botnick says about 70 to 75 percent "we know are real," meaning they've researched the recording history to verify the file has more information than just a CD rip or has some other quirk in the original recording justifying a mixed or lower resolution.

He said efforts are being made to further assure consumers of the "provenance," or origins of recordings, and how they got to be labeled high-resolution.

"It's a real fact-finding job" and "it's going to take some time" to handle the thousands of albums in question, he said. Until then, it's a case of "buyer beware," he said.

And while audiophiles may be aware of the rarified, often hard-to-detect benefits of the high-resolution files, average music lovers can easily over-value the claims made by backers, according to Mark Waldrep, a recording engineer, college professor and writer of the "Real HD-Audio" blog.

Studios are re-releasing older recordings in giant data containers that are sometimes barely merited, he says.

That conclusion was reinforced when he analyzed high-res Warner Music re-releases of Joni Mitchell's "All I Want" from the 1971 album "Blue" and "Ain't No Way" from Aretha Franklin's 1968 album "Lady Soul," which The Associated Press bought from the PonoMusic store.

"You're buying a container that's really 50-60 or even 70 percent zeroes. It's all empty information," he said. "The frequencies you're buying up here are either all zeroes, or hiss, which contributes nothing to the enjoyment of the music, unless you're into hiss."

And very few, if any new albums, are being made in the super-high resolution specs that Pono is touting.

Giles Martin, the Grammy-winning producer of the "Love" soundtrack for The Beatles-Cirque du Soleil show in Las Vegas, says the highest fidelity he records at is 96 kHz, 24 bits, after which there's no benefit in boosting the playback specs. "You can't upscale audio," he says. "There's a compromise in having huge high-res files that don't sound any different than other ones."

From the record labels' point of view, part of the re-mastering process is simply to preserve aging analog tapes at the highest practical digital format.

George Lydecker, a of engineering and archiving at Warner Music, says a CD-specification release of Franklin's "Lady Soul" wouldn't have been as accurate a reproduction partly because placing a necessary filter at the lower frequency required by CDs creates some distortion. Instead, the 192 kHz, 24-bit file that was released "is like standing in the studio live and hearing Aretha belt it out."

The album goes for $17.99 on the PonoMusic store. A CD can be had for $4.99 on Amazon.

While not all people will be able to hear a difference, some will.

"For the first time, you can get the file (that was) approved by the mastering engineer in the studio," says Jim Belcher, Universal Music's vice president of technology and production. "And for a lot of people that doesn't make sense. For a segment of the market that really cares about audio quality, they want that."

And that's the other thing. Even with a $400 PonoPlayer or some other high-end playback device like a Sony Hi-Res Walkman or Astell and Kern AK100II, or even the latest smartphones from Samsung and Apple, audiophiles who want to hear the true benefits of high-resolution audio should also have headphones or speakers capable of playing back those high frequencies that only few humans can hear. In some cases, that could require a headphone amplifier.

John Siau, director of engineering at high-end equipment maker Benchmark Audio, argues that consumers are fooling themselves if they believe they can appreciate high-res audio without the proper high-end equipment.

"There's no point in having high-resolution playback formats if your playback equipment can't even match CD quality," he says.


Explore further

Sounds of science: Japan leads push for high-res audio

© 2015 The Associated Press. All rights reserved.

Citation: Audio overkill? Some question benefits of 'high-res' music (2015, June 26) retrieved 20 May 2019 from https://phys.org/news/2015-06-audio-overkill-benefits-high-res-music.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
92 shares

Feedback to editors

User comments

Jun 26, 2015
This is a bit like buying "Its A Wonderful Life" on BluRay. You think you are going to get more detail? Probably not, because its only as good as the original film. Audio is even more difficult to "improve" or "upscale", because you cant make it better than it is originially.

I love higher quality mp3's, they are louder and distort less, but I will not be paying a premium price for it. Its not worth it. And your audio experience is only as good as your weakest link, from Recording -> Media -> Player -> Amplifier -> Speakers. This is psychoacoustics at work!

Jun 26, 2015
If you're into quality music, all you need is non-compressed lossless CD format, and good equipment. I have Sennheiser HD 800 Headphones with dedicated amp, and Paradigm Studio 100 speakers,.... but still wouldn't waste my time with hi-res music.

Jun 26, 2015
I find it ironic that most studios use 96kHz / 24 bit for recording yet PonoMusic sold Mr. de Oliveira an upsampled 192kHz / 24 bit copy of the 16 bit CD. That is false advertising and Mr. de Oliveira should get a refund.

Jun 26, 2015
Why anybody these days would use a lossy compression system given the cost of memory is beyond me.

Jun 26, 2015
While the higher frequencies are outside of human hearing the thing you miss is the 192kHz is a sample rate, which means a much lower frequency sound will have much much much better definition when recorded, and if you want to hear every nuance of that string or vocal line, the higher the sample rate the better.

I run all my audio systems at 192kHz anything less sounds like mud to me.

@Survtech
A 96Khz studio is regarded as a "B" studio,. A studios use 192khz or higher.


Jun 26, 2015
There is simply no critical listening study to date which passes muster with good experimental technique and with reproducibility that supports the ability of people to discern anything past 16 bit 44.1kHz reproduction with any statistical significance. Not even "golden ears."

It is all snake oil and Neil Young is either a con man or remarkably gullible. That and/or he is no longer able to hear with any acuity which, given his performance history, is extremely likely.

Jun 26, 2015
@DonGateley

Your statement is simply not true, like I said Sample rate is a lot more than frequency response.

A grade studios have been using 192kHz since the 50's for a damn good reason.

12 inch singles are made and released on radio for a damn good reason.

The reason is it sounds better, if you sample a bass line at 192kHz you get to see every little nuance of that string sound, you can literally hear the players fingers on the string.

It's obvious you have never heard a comparison between 192kHz and any lower sample rate.

Jun 26, 2015
This comment has been removed by a moderator.

Jun 26, 2015
@DonGateley

Your statement is simply not true, like I said Sample rate is a lot more than frequency response.


Citation, please, to a peer reviewed study that validates your personal opinion. It doesn't exist for good reason too.

One thing I will admit that the ear can hear without doubt is dollar signs.

Jun 27, 2015
Citations are ...

Neil Young a recording artist for the last 50+ years.
Every A grade studio engineer in history.
Me a recording artist for the last 20 years.
Any musician that has ever tried to do recording session work for public sale.

A grade musical equipment and recording facilities exist because we know the difference, and yes they cost more for a very good reason .... 192kHz capability.

Jun 27, 2015
Placebo effect--no evidence [i] done blindly that people can actually hear a difference, [ii] even where people can hear a difference (such as CD vs MP3) no evidence exists that adds any extra pleasure--Édith Piaf, Duke Ellington--I can hear the hiss and crackles--so what my brain can hear the music. It is not like vision where extra clarity usually creates a better experience--our ears are resilient in a way our eyes are not. PonoMusic is just selling placebo.

Jun 27, 2015
Placebo effect--no evidence [i] done blindly that people can actually hear a difference, [ii] even where people can hear a difference (such as CD vs MP3) no evidence exists that adds any extra pleasure--Édith Piaf, Duke Ellington--I can hear the hiss and crackles--so what my brain can hear the music. It is not like vision where extra clarity usually creates a better experience--our ears are resilient in a way our eyes are not. PonoMusic is just selling placebo.

Jun 27, 2015

A grade musical equipment and recording facilities exist because we know the difference, and yes they cost more for a very good reason .... 192kHz capability.


I said reproduction. There are reasons for higher sampling rate at recording, mixing, processing and mastering in that effects have non-linearities which can cause aliasing and the higher rates move the aliasing artifacts above the highest frequencies we can hear which is less than 22050 Hz. After all that is done, down sampling the master to 44100 with good brick wall filtering renders any differences completely inaudible to anyone. If you don't know this you are blowing smoke about all your experience. At least about the value of it.

As Squirrel says, you are being placeboed.

Jun 27, 2015
Sample rate is about how many samples per second are captured.
I.E.The definition of the sine wave captured is much more accurate, this renders a recording at much higher sample rates to be a much more realistic reproduction of the original sound being recorded..

Sample rate is not frequency response, and the fact you are arguing it is, means you have never heard an acoustic guitar at 192kHz, the entire recording industry knows this, and we have heard your arguments in every form since we started recording.

The problem is when talking about sound fidelity the argument is purely subjective and people like yourself think they know everything there is to know about sound because they have been listening to things since they were born and regard themselves as experts in this subject. This is a false perception of their own abilities, and the truth is they have never heard the difference because they have never heard a 192kHz recording.

Jun 27, 2015
Get a quality headphones and check frequencies your ears can hear.

https://www.youtu...bppCX6Rk

And then think again about 192kHz superiority.
Of course studios use 192kHz recordings with 24bit. They need to mix and add effects etc, and for that you need more precision to avoid artefacts. If I am not wrong, many softwares convert sound into floating point numbers for that. However, once it is finally downsampled to 44kHz 16bit I sincerely doubt that there are people able to discern this from the original 192kHz 24bit. To prove that you need to conduct proper blind studies. (not to mention state of the art equipment)

KBK
Jun 27, 2015
Ears are like intelligence, or like that of eyes.

Some people are intelligent, some people have exceptional eyesight.

Some people have exceptional hearing.

Attacks on high resolution audio come from people who don't have the mental function to hear it and probably don't have the ear function to hear it.

They are 'ear morons.' And, due to their missives on the subject....more than likely are also --brain morons.

A nice pair of ignorant and incapable bookends.

In stereoscopic hearing, we can easily hear to a clean 'jitter free' 100khz equivalent in ear arrival times. Just on the one function alone of a single 'ping' at 1khz, in a flat 2d plane of stereoscopic source placement/identification.

When we add in the response of the pinna (outer ear), and harmonics of complex function (instrument harmonics and multiple instruments), then even 192/24 is shown to be totally inadequate.

Essentially, don't let illiterate ear morons dictate reality.

Jun 27, 2015
I'm sorry for the people that can't tell the difference, because high res music is a beautiful thing. Sure a lot of past music will keep their original format but this is about the now and the future, not the past. Also let's not get confused between sample rate frequency and sound frequency, Sure 192KHz sample rate seems high if you're digitizing a 20Hz wave but it is actually very low when digitizing let's say a10KHz sound wave. That's only 19.2 samples per full wave. They should have probably raised the sample rate even higher and keep the resolution at 16bits.

Jun 27, 2015
I've upgraded all my collections to FLAC (usually specialized LP rips) and that's enough for me. The difference is more than obvious, but that's enough for me. No need to go any higher than that. The benefit is not that worth.

Jun 27, 2015
The audiophiles can throw as many specs as they want around this discussion. There simply is no apples to apples comparison when discussing HD visual to so called HD audio. You need to understand the limitation here is biological. The ear is just limited as a data collecting device. The human ear merely evolved to detect some atmospheric pressure differentials, all within a very narrow bandwidth, often quotes as 20Hz-20 KHz (signal amplitude is another interesting issue)...and this is for very young, undamaged ears, not us old folk.

Conversly, the human eye evolved to detect roughly 400nm-700nm ( a bandwidth of 360 THz..just no comparison). Some people can see a little deeper into the near UV and others as far out as 800nm, a common laser line. Additionally, with some practice, the human eye can actually see polarization in the sky above us. So sorry, but applying the moniker of 'HD' to a source with such a narrow bandwidth is simply absurd.

Jun 27, 2015
If you are into high res music just go to a frigging acoustic house concert for god's sake....

Great point! But then they wouldn't be able to brag about their perfect hearing and abilities to hear "obvious" difference between cd and hd sound. I remember I read once about a blind and then visual test of high end cd players. LOL. Very different results from same people if they knew or not which brand they were listening to. And did you know that expensive wine is always "better" tasting than the cheap one. Etc. I suggest to read something from phsycology of human behavior. Nothing can be trusted! We can't trust our memories, feelings, free will... But of course I will believe an older person who can't hear above say 12 kHz that he or she can "feel" the difference between 44 and 192.
If you are so certain then organize some testing!

Jun 27, 2015

If you are so certain then organize some testing!


Indeed. No one who actually has done that with scientific rigor has ever been able to show a perceivable difference when level matched and using hardware capable of reproducing the extra "HD" information (which, BTW, practically no real speaker or headphone can do according to any measurement.)

The AudioFreaks will never, ever accept that because they "know" better than science is able to demonstrate. What's weird is that these same people will bet on science in nearly any other circumstance. Their ears, however, uniquely stand outside of it.

Jun 27, 2015
Let's see what a difference there is at samplerates of CD (44kHz) and of HD (192kHz). Take apaper and draw a large blue circle onto it. Now mark red dots onto the circle line at 12 o'clock, then at 4 and at 8. Then connect those 3 dots with your red pen. The result is 12, 4, 8, 12 connected triangle. This is what one has as output when using CD sampling of 44.1kHz/3=14,7kHz sound. Can you see the difference there is between the circle and the triangle? Of course you can, and so does our ears. If the 14.7kHz circle is sampled ar 192 kHz, one is allowed to take 192kHz/14,7kHz=13 samples of the circle. It means green dot at every clock face number:1,2,3,4,5,6,7,8,9,10,11,12. Now connect all those green dots with green lines. Just increasing the sample rate 4x (from 3 to 12) we now have more hi-fidelity representation of the original signal.

Jun 27, 2015
Still don't belive me? Ask someone to play you from analog signal generator sine, triagle and then square wave at the same frequency, for example 15kHz. In theory all the harmonics in 15kHz square wave (30,45,60 kHz etc) are outside 'hearing range', but there is difference in what we hear. Therefore CD sample frequency of 44kHz is not good enough for hifi recording and playback of 10-20kHz sounds.

Jun 27, 2015
Not sure where to start here. I rip CD's using VBR (320 - 192), TBH, I can't tell the difference between that and FLAC, but my ears are not what they were :)
I have an analogue sound system* with a wooden cased woofer, which gives a nice sound at any volume. About harmonics, yeah, they are very important, without them, you lose so many overtones. I've listened to a lot of recent hardware (not high end stuff) and it sounds harsh.
Pfft, It's all subjective, as long as you enjoy it, it'll sound great :)

*Plugged into a computer, so actually a D2A.
/ends ramble

Jun 27, 2015
Using marks on paper to depict the accuracy of sound reproduction is ignorant and foolish. Unless,of course those marks are the sampling theorem and the simple proof thereof. PPihkala, your example is completely irrelevant to the reproduction of sound. There's more to this than meets the eye. Much more, and you should get some relevant education before making a fool of yourself. After that it's ok and quite common.

Jun 27, 2015
The value will likely emerge once the system is better established. In unpredictable ways.

With more data to work with there's more scope to enhance recordings. Same with live hearing in the brain. Raw audio data is processed to extract information about the location, movement, and nature of specific sound sources, plus disruptions and interference.

Live classical music concerts include rustling, restlessness, coughing and spluttering from the live audience. Plus the mobile phone alerts, security and fire alarms, and emergency vehicles in the vicinity.

The brain compensates for audio intrusions. But it's more effective if it has more source data to work with. Maybe software can assist. But music is mostly in the mind.

Beethoven didn't need high bandwidth audio. Just ink on paper.

Jun 28, 2015
Sure 192KHz sample rate seems high if you're digitizing a 20Hz wave but it is actually very low when digitizing let's say a10KHz sound wave. That's only 19.2 samples per full wave. They should have probably raised the sample rate even higher and keep the resolution at 16bits.


Sorry, this is just not true. You and other "audiophiles" saying similar things should learn about sampling theorem:

https://en.wikipe..._theorem

If you have more than two samples per wave, you have enough information to reconstruct the original signal PERFECTLY. That is a mathematical fact, no ifs and buts. It is a little counterintuitive but it is true nonetheless. Since human hearing only goes up to 20khz, traditional sampling of 44.1 khz is enough for perfect reproduction. Higher samplings may possibly be of use during mixing and such, which is one reason why it is used in studios. But it is of no use when listening to music.

Jun 28, 2015
Aside from your own hearing, two crucial factors come into play:

- The way the recording was made
- The type of equipment you're using to play the recording

You can't add data that wasn't there when the recording was made, and you don't profit from data your equipment can't play back. But to dismiss higher resolution recordings offhand is equally daft as to unconditionally embrace them.

Bitrate is FAR more important than sample rate, btw. Listening to a classical symphony (because of the large dynamic range) on very good speakers in regular and higher bitrates will reveal that.

Jun 28, 2015

Beethoven didn't need high bandwidth audio. Just ink on paper.


...the result being that no two performances of any of his works sound the same, because classical notation does not provide all aspects of a performance and is highly subjective, especially regarding indications on dynamics.

KBK
Jun 28, 2015
Sure 192KHz sample rate seems high if you're digitizing a 20Hz wave but it is actually very low when digitizing let's say a10KHz sound wave. That's only 19.2 samples per full wave. They should have probably raised the sample rate even higher and keep the resolution at 16bits.


Sorry, this is just not true. You and other "audiophiles" saying similar things should learn about sampling theorem:


The math and the methodology falls short of the human hearing mechanism. The reasons are complex. The reality is that some people can hear the difference and some cannot.

A simple if-then-so equation.

Beating it with dogma makes no sense, unless one is simply incapable of change and uses dogma as a projected reality.

The trick is: What concept created the math? Is the complex concept in error? Turns out it is, as some can hear the difference.

The engineering and math, as a set, are misapplied.

Understand: Engineering=dogma -- Reality=theory.

KBK
Jun 28, 2015
As an aside, to equal the inter-channel phase accuracy of a 33.33 12" stereo LP, a given sampling system would be required to operate in a perfected Jitter free 7 million samples per second.

That's 7Mhz, with zero jitter in the clock. Just to equal the native inter-channel phase accuracy of an LP.

This is part of why LP's are so good at stereo reproduction.

The math dogmatic folks use to say otherwise, is flawed as their understanding of the Question being asked is deeply flawed.

The human ear is a magnificent, complex, and highly phase sensitive device, and ~~~no two are built or wired the same~~. Just like athleticism, or intelligence.

Some ears/brains are basic..then... some are excellent, well wired, highly trained, and connected to an excellent discerning brain.

Virulent and projected ignorance of this can go straight to hell ....and I'll be right there kicking it on the curb it so nastily asks for.


Jun 28, 2015
This comment has been removed by a moderator.

Jun 28, 2015
Let's not forget the fact that harmonics and subharmonics, even those well outside the range of human hearing, can significantly affect the way we perceive those frequencies that do fall within our own hearing range, regardless of whether they originate in the original recording, post-processing, or due to errors in playback. Basically, when it comes to sound, pretty much everything has an effect, and the perception of sound varies strongly from person to person, for whatever reasons, whether psychological, neurological or through the physical characteristics of individual ears and auditory nerves. Which aspects of an audio signal are meaningful and which are not is for a large part the same across the board for all human beings, but there are also aspects about which we do not agree, we might say that they are a matter of taste. There is no such thing as the perfect recording/playback of a given piece of music for everyone - you have to make choices of a subjective nature.

Jun 28, 2015
This comment has been removed by a moderator.

Jun 28, 2015
I certainly don't mean to imply anything esoteric is going on with the interaction of harmonics and subharmonics (it's a lot of math, but not particularly unintelligible math), however, I do mean to point out that the subjective experience of sound can be affected differently for different people with the same objective change to the signal. The 'ideal' perception of sound and music is not the same as accurate signal reproduction, that's the essence of what I'm saying.

Jun 28, 2015
This comment has been removed by a moderator.

Jun 28, 2015
Perhaps I should try hi-res then. I had been suspicious given many forum posts from the audiophile world,.... for example reviews of amps and over-engineered cables with all kinds of descriptive auditory expressions, like 'they sound more bright' or 'less separation',...... my last post was "what brand Q-tips do you use?".

Jun 28, 2015
Yes, this is the point. The audiophiles actually don't expect faithful reproduction of signal - but as rich of phase/amplitude modulations, like the live presentation. For example the Stradivari violin are highly appreciated by their sound, because they're covered with thick layer of hard varnish, along which the sound resonates in longitudinal waves. This leads to formation of metallic tremolo (like banjo) and also supersonic carrier, which modulates the sound randomly, as the violin move with respect to listener. The audiophiles just want to keep and reproduce this effect - they don't care actually, if it's exactly the same, like at the live concert.


You can't simply say that the one view is better than the other, only that they are different. Accurate signal reproduction is sometimes the anwer, but not always. The problem in the 'fidelity wars' is one of communication: audiophiles say subjectively good sound = accurate sound, 'scientists' say accurate sound = good sound.

Jun 28, 2015
This comment has been removed by a moderator.

Jun 28, 2015
The ultimate test of a theory (if you don't accept mathematical proof) is experiment. Some non-believer please perform an unambiguous and scientifically rigorous experiment that shows sampling theory to be false based on perception and one that demonstrates the ability of humans to perceive any frequency above 22050 Hz (you can linearly mix it with anything else.) If you persuasively and repeatably succeed you will become famous and receive lots and lots of citations.

BTW, the argument for the non-linear mixing of higher harmonics to produce ones that can be heard only applies at the ear itself (any produced by transducers would be unnatural) and there is no evidence of that phenomenon to date. If the lower harmonics are produced by the system being recorded, those lower harmonics will be recorded and reproduced.

Jun 28, 2015
@DonGateley

It's not about the frequency it's about the accuracy of the sample that we can actually hear.
In any scientific endeavor the higher the sample rate the more accurate the wave form reproduction is.

It's obvious all the nay sayers hear have never heard a 192kHz recording.

The difference is obvious once it's been heard, and anyone that can hear will realize the difference if they experience it.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more