High Definition iTunes Music Downloads May Be on the Horizon

[cdpage]

how about audiophile mastering?

If anyone can convince the music industry to wake the hell up it's apple.

If apple were to offer remastered (music that is dynamic not Flat where everything is loud and designed for the radio only music) I would bet you you would see a lot more peole coming to apple to download music for the headphones, 2 channel stereos and home theatre systems.

I want to listen to music again. and enjoy it.

 

[i4m]

Yep, it's placebo.

I can tell CD from 192k, but not at 256k. At 128k I can hear the difference, but 95% of the time, 256k is the sweet spot of files size and fidelity.

I don't know why iTunes would make a full rez file. A full rez CD is like 10x the size of the aac file, soooo..

Why not just make a new "mp3/acc" format that uses the 24-bit/192 files that to make a compressed 24-bit file? It'll be much smaller and I'm sure 99% of the population won't be able to tell the difference between a 24-bit lossless file and the 24-bit mp3/aac file, in the same way they can't tell the difference between the 16-bit CD and the 16-bit mp3.

If you can't tell the difference then consider the following possibilities:
1. you're not listening carefully
2. your hearing has been damaged
3. you're listening to crappy speakers.

 

[Poisednoise]

Can I play devil's advocate for a second?

First off, can I say I cannot refute all the evidence that seems to suggest that it is virtually (if not completely) impossible to distinguish between 44.1/16 bit and higher rates. I accept that, and I'm glad, because otherwise I'd be upset I had such a large CD collection. But, here's what I'd like someone to explain:

Take a clearly never-to-happen-in-real-life scenario: You have a sine wave at 22.05kHz. The sampling frequency is 44.1 - exactly double. The recording equipment takes samples, as luck would have it, precisely on the anti-nodes of the sine wave. It will then record that, surely, as no sound at all?

Like I say, this would clearly never happen in real life, and being only slightly off with your sine wave frequency at that high a rate would mean that the sound wave would be reproduced very close to accurately, but nevertheless one could extrapolate from this certain scenarios which could happen in which the wave would still not be perfectly reproduced. I've never really read up on Nyquist theory, so I'm sure this has been dealt with somewhere - it seems too easy a flaw for there not to be some reason why that's not how it works. But I'd love someone to explain it to me!

 

[csbo]

High Definition iTunes Music Downloads May Be on the Horizon

but nevertheless one could extrapolate from this certain scenarios which could happen in which the wave would still not be perfectly reproduced.

Why wouldn't it?

Also note that the filters involved remove virtually all information at 22.05

 

[Poisednoise]

Why wouldn't it?

Also note that the filters involved remove virtually all information at 22.05

Ok, well, I can't explain it very well - and as I say, I'm sure it's down to me fundamentally misunderstanding what's going on - but let's take a slower wave, and sample it more than twice a wave cycle - I've sketched here a black wave form, and then sampled it and sketched out in red the resultant information. I suppose it's then down to how the DAC actually converts that blocky form into something that's actually smooth and analogue, and I'm sure that's where the magic actually happens, but on first sight the red waveform certainly bares little resemblance to the original waveform:

 

[rhett7660]

I find that people who have spent their lives listening to music through crummy laptop speakers often say similar things.

Couldn't agree more.

I for one like this! I have the equipment to take advantage of this! I have a few high bite rate files already..... Would love to add to them.

 

[csbo]

Ok, well, I can't explain it very well - and as I say, I'm sure it's down to me fundamentally misunderstanding what's going on - but let's take a slower wave, and sample it more than twice a wave cycle - I've sketched here a black wave form, and then sampled it and sketched out in red the resultant information. I suppose it's then down to how the DAC actually converts that blocky form into something that's actually smooth and analogue, and I'm sure that's where the magic actually happens, but on first sight the red waveform certainly bares little resemblance to the original waveform:

The digital signal has a value at the sample point and then nothing until the next sample. So instead of that blocky graph where it samples and holds draw it as a lollipop graph with a vertical line to the waveform at each sample point

 

[NewishMacGuy]

Seriously foolish to insist that the human body has limitations, or that the way sampling works doesn't support the claims made about loss of audible content

Now if you'll excuse me, I have to fly somewhere by flapping my arms. Anyone who says that I can't fly that way just because they can't is a fool

----------



I have nothing against you spending money that way, it's your money. But the claims about it being more accurate are scientifically unsound, and they usually get accompanied by the sort of snobbery like above--sure, it's okay for those silly kids to like something terrible, but you are better than that

Insisting that the human body has limitations is not foolish. Insisting that you know exactly what they are (or that the scientific community does) is seriously foolish. Just because you've never seen a black swan doesn't mean that they don't exist. Just because no man had ever run a 4 minute mile didn't mean that it couldn't be done. Human knowledge is in a perpetual state of being imperfect at best and often flat out wrong, even human knowledge which has been "scientifically proven."

If you are of a certain age you will remember the "perfect sound forever" promise made by early proponents of the CD. The "perfect sound" part of which was also "scientifically proven" - except that the science was wrong.

We now understand better, even though everyone who said at the time that digital recordings sounded awful was pilloried by those who had "scientific proof" that the early 16/44.1 digital recording and playback technology was sufficiently advanced to be beyond the ability of humans to detect a qualitative difference between the analog master and the digital reproduction.

It was foolish to tell people that they couldn't actually hear the differences they claimed to be hearing because it was beyond the "perfectly established" limitations of their perceptive abilities in retrospect right, yet following the same line of argument which you have put forth above...

...a foolish one.

>

 

[Loge]

Take a clearly never-to-happen-in-real-life scenario: You have a sine wave at 22.05kHz. The sampling frequency is 44.1 - exactly double. The recording equipment takes samples, as luck would have it, precisely on the anti-nodes of the sine wave. It will then record that, surely, as no sound at all?

The theorem requires that the maximum frequency is strictly less than (not equal to) half the sampling frequency, though it only needs to be infinitesimally so. That way you will never get only the anti-nodes.

 

[chabig]

The theorem requires that the maximum frequency is strictly less than (not equal to) half the sampling frequency, though it only needs to be infinitesimally so. That way you will never get only the anti-nodes.

Exactly right. http://en.m.wikipedia.org/wiki/Nyquist_rate

 

[csbo]

Insisting that the human body has limitations is not foolish. Insisting that you know exactly what they are (or that the scientific community does) is seriously foolish.

not saying I know exactly what they are. Just that, given the normal range of human hearing and the effects of age along with the inability of people to hear the differences in controlled environments, all evidence points to the claimed advantages being outside the limits Just

If you are of a certain age you will remember the "perfect sound forever" promise made by early proponents of the CD. The "perfect sound" part of which was also "scientifically proven" - except that the science was wrong.

i think you're confusing a marketing pitch with science

It was foolish to tell people that they couldn't actually hear the differences they claimed to be hearing because it was beyond the "perfectly established" limitations of their perceptive abilities in retrospect right

except that the differences ended up being within the limitations

a foolish one.

im perfectly open to actual evidence to the contrary. Claims with nothing to back them up aren't evidence, though

 

[reel2reel]

Sorry, but 192k sampling rate is TOTAL overkill and simply a waste of space, even if you plug in your Mac/ future iPhone with 192k audio support into the best DAC/amplifier/speakers out there ...

Overkill for *you*

HD Audio is the future. It's just not for people who listen to music through their white earbuds.

I find that people who have spent their lives listening to music through crummy laptop speakers often say similar things.

Exactly. I know someone who's a HUGE Music Snob. And he listens to music through his MacBook Pro speakers. Just goes to prove that Music Snobs don't actually enjoy music, they just like talking about it.

 

[217833]

That is actually not true, as the parent explained frequencies above the Nyquist frequency would "disrupt" the correct measurement of frequencies at or under it unless you filter them out (in this case 2xNyquist sampling is enough) or sample at a very high rate (above the theoretical 2xNyquist). Today most digitizers internally do oversample to address this issue.

The point is, even if it matters for recording and/or production, it still doesn't matter for reproduction: in the context of distribution and end-user consumption 24/192 makes no sense and 16/44.1 or 16/48 are perfectly capable of providing the highest fidelity human hearing is able to perceive.

That's why I was saying than 16/44.1 is more than enough (over twice the human being capabilities).

And if we talk about the recording process, most of recording gear do filter out higher frequencies and use oversampling as you mentioned it. So it's not an issue and still not a reason to use 24/192. There's a great article from Dan Lavry that even explain how worst the quality gets over 88.2 or 96 kHz: http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf

 

[zhenya]

Overkill for *you*

HD Audio is the future. It's just not for people who listen to music through their white earbuds.

LOL.

You and your Golden Ears, right?

It's clear here who has actually done ABX testing and who hasn't.

----------

Exactly. I know someone who's a HUGE Music Snob. And he listens to music through his MacBook Pro speakers. Just goes to prove that Music Snobs don't actually enjoy music, they just like talking about it.

You're confusing people who love MUSIC with people who love EQUIPMENT.

Most of the best musicians I know are not particularly picky about what they listen to because the feeling of a given piece of music comes through pretty darn well on even the most basic of systems.

 

[217833]

Overkill for *you*

HD Audio is the future. It's just not for people who listen to music through their white earbuds.

Overkill for any human being on the planet.

Unless you got some bionic dog earring implants, you can't hear anything above 22kHz (at best), probably not much above 15-16kHz if you're an adult... and probably no much above 10-12kHz if you're over 60.

16bits / 44.1kHz for audio distribution is already HD Audio. If you can't understand this statement, you should probably not comment about it

For all "Golden Ears" out there, please go and pass the gold level on the Philips Golden Ears Challenge: https://www.goldenears.philips.com/en/introduction.html

 

[lars666]

What the guy above me said. reel2reel, you've just embarrassed yourself a lot, like some others here, too - please people, inform yourself about frequencies and the human ear before talking stupid nonsense. By the way, I use $1000 custom made in-ears for listening on the go and have a very good sound system for listening at home - I welcome every high definition sound possible, even if the improvement is just a subtle nuance. I'm one of those "audiophile" guys who makes common music listeners shake their heads, and even *I* say that 192k is just ******** overkill which doesn't improve anything. (And by the way, I didn't include the info that you won't hear a difference *even with the best DAC/amplifier/speakers* for nothing – so I don't really understand why you think my misconception of the greatness of "192k HD audio" lies in not listening to it through the correct equipment.)

Next step: cable sound voodoo and $500 plug sockets who deliver "cleaner" electricity for audiophile sensation.

 

[rhett7660]

Can someone explain to me why ANYONE would pay $2 per song for 256k when Spotify Premium offers playback at 320k on UNLIMITED songs for $10 a month?

What a joke..

Maybe because we want to own the song, and if spotify or another service goes belly up, you do not have access to THAT song.

I like to own my music and movies not rent them.

 

[bsolar]

Maybe because we want to own the song, and if spotify or another service goes belly up, you do not have access to THAT song.

I like to own my music and movies not rent them.

I understand what you mean but technically you don't own the songs you buy from iTunes either.

 

[Luba]

Not using a lowpass filter results in better sounding music?

Not using a lowpass filter results in better sounding music?

A recording studio that uses 192kHz sampling would know to turn-off the low pass filter since not needed? And the recording would be more accurate since the digital sine wave would closely match analog sine wave?

Most recording studios would then probably down sample to 24/96 from the 24/192 source?

If cost wasn't a consideration would you buy 24/96 or 24/192 music?

Thanks for the information! I'm very glad you participated in the thread.

This is not an inherent property of what the nyquist limit does. Your statement is only true if you've used a lowpass filtering in your mastering process so that frequencies and harmonics above the Nyquist limit are filtered out before they are recorded.

For example: A Nyquist limit of 22.05kHz does not mean that it will not sample 33kHz. ANY sampling frequency will intersect with that 33kHz wave in a way that produces a series of values that result in a different wave at a lower, and often audible frequency.

There are only two ways around this phenomenon:

1. Use a lowpass filter at the nyquist limit to completely prevent any frequencies above that limit from being captured at all. The nyquist limit is a conceptual limit, not an actual barrier to sampling frequencies above it.

2. Choose a Nyquist frequency so far above the A-weighted range that frequency roll off and frequency aliasing become completely immaterial, as they will not produce other artifacts inside the range of human hearing.

The other general rule is to work in multiples of 24. So if you're sampling at 192kHz during the recording stage, don't mix down to 44.1, but rather 48 or 96. This reduces downsampling error... though computer processing is so powerful these days that this is a far smaller concern than it used to be.

But again, I want to be clear that the Nyquist limit does not prohibit frequencies above it from being sampled. Even the sampling frequency doesn't prohibit that. If you sampled a 33kHz wave thirty-thousand times a second, it would produce an audible distortion... not the original wave, mind you, but the resulting ALIAS of the wave--a lower frequency wave--which ends up inside the A-weighted spectrum.

More detail here.

Excerpt:

"Although sampling at twice the Nyquist frequency will ensure that you measure the correct frequency of your signal, it will not be sufficient to capture the shape of the waveform. If the shape of the waveform is desired, you should sample at a rate approximately 10 times the Nyquist theory."

192 is closer to 10x nyquist than 22.

 

[chabig]

Not using a lowpass filter results in better sounding music?

There always has to be some kind of lowpass filter in the output.

A recording studio that uses 192kHz sampling would know to turn-off the low pass filter since not needed? And the recording would be more accurate since the digital sine wave would closely match analog sine wave?

No. Sampling at 192 kHz will not match the analog waveform any better than sampling at 44.1 kHz. Mathematically, you only need to sample a more than twice the frequency of the highest signal of interest to exactly match the analog waveform if you have a perfect lowpass filter. Since perfect lowpass filters don't exist in the real world, we always sample at more than twice the frequency. I know this isn't intuitive, but that's how it is.

 

[rhett7660]

I understand what you mean but technically you don't own the songs you buy from iTunes either.

You are correct, but I have them and copies of them. So if iTunes/Apple was to go belly up, I still have and can play those songs that I have purchased.

 

[Avatar74]

I don't know if you are able to listen to a 24 bit sound file on your computer now...

I have three sound cards with 24 channels of 24/96 input/output... I think I have this covered.

I would highly recommend you just take some time listening and avoid Nyquist theories that are just a bunch of math... <snip> ...Theories are only theories and they assume everything to be perfect for which they aren't in REALITY.

Thanks for the... errr, advice? I've done professional recording, mixing and mastering with an album credit or two to my name.... Can I make a recommendation for you? Read Ken Pohlmann's Principles of Digital Audio, for an introduction to the fundamentals of pulse code modulated audio. It was first published in 1985... and is widely regarded by most hardware engineers as the Bible on all things concerning digital encoding, a process which is due to its nature exactly as you say... "just a bunch of math."

 

[Avatar74]

Not using a lowpass filter results in better sounding music?

Well, what I was trying to illustrate is that the Nyquist limit doesn't actually prevent frequencies above it from being sampled. Even the sampling rate doesn't prevent frequencies above it from being sampled. Undersampling will produce a completely different wave because of the limited data points from which the DAC is trying to reconstruct the analogue waveform. Like this:

So there's two approaches to resolving it: Either by preserving an adequately high bit rate and sampling frequency, to make for as accurate reproduction as possible, or use the appropriate lowpass filter at every downsampling stage set to the nyquist frequency of the lower sampled format to ensure that nothing goes INTO that downmix that would result in aliased frequencies.

If cost wasn't a consideration would you buy 24/96 or 24/192 music?

Between 96 and 192kHz the sample rate isn't as important to me as long as the appropriate measures are taken as I mentioned above. Either way, 24 bit amplitude resolution is important because it alone affects the dynamic range of the sound recording.

Cost is not the limiting factor for me... What is a limiting factor is convenience and practicality. I don't carry my Nikon camera everywhere because it's large and my iPhone can do the job in a pinch.

Likewise, 24-bit stereo LPCM is pretty large, about 2307 Kbps uncompressed vs. 1411 Kbps for 16-bit CD-DA, or 256 Kbps for iTunes AAC.

Also, it's not practical unless MOST artists and producers were doing stuff that actually took advantage of the dynamic range overhead that 24-bit recordings provide... most don't, so the catalogue of songs I'm likely to even find that would make 24 a noticeable benefit is very, very small.

In addition to storage there are streaming considerations.... none of those uncompressed formats are easily accessible anywhere I want, any time I want. If I want to keep my storage requirements down when I'm on the go, AAC does the trick. That doesn't mean there can't be an AAC perceptual coding schema that doesn't shrink 24-bit LPCM.

Perceptual coding creates more efficient ways of reconstructing the same data... a 16- or 24-bit LPCM stream can be encoded to AAC and decoded and reconstructed pretty indiscernibly. It would just mean a larger AAC bitrate/file size for transcoding a 24-bit LPCM stream versus 16.

So it then just comes down to: How much storage space do I want this stuff to take up, and is the marginal, DISCERNIBLE difference worth the extra storage?

 

[Frankied22]

High Definition iTunes Music Downloads May Be on the Horizon

How many people are actually going to buy all their music again just so they can think it sounds better? Especially at $2 a song? No one apart from the crazy audiophiles. Apple just needs to accept the fact that music sales are going to keep declining because people can get music for free with Spotify, and if they don't want ads they pay the price of one album a month and have on demand streaming and offline access to millions of songs. It's essentially legalized piracy and now that people have had a taste they are not going to go back to buying every album individually for $10.

Apple should be focusing on branching iTunes Radio to include a $9.99 a month subscription service that gives you all access to the whole iTunes music catalog. Make it very elegant and user friendly and of course make it play very nicely with other Apple products.

 

[hayesk]

If you can't tell the difference then consider the following possibilities:
1. you're not listening carefully
2. your hearing has been damaged
3. you're listening to crappy speakers.

4. you understand that the human ear really can't tell the difference.

Maybe I missed it in the 19 pages, but everyone please read this:
http://people.xiph.org/~xiphmont/demo/neil-young.html

Feel free to prove this article wrong with actual evidence.

Besides, spend your money on a decent quality DAC instead of on high res music. It'll make an actual difference (up to a point).