High Definition iTunes Music Downloads May Be on the Horizon

[rboy505]

"High-definition 24-bit downloads are said to offer better detail, greater depth, and a deeper bass response compared to traditional 16-bit music downloads, but the file sizes are much larger."


What sad claptrap. 24 bit files as a final product have the potential for greater detail and depth, but it's in the lower regions of the audio, where sounds fall off and where ambience is either reproduced well or not. Commercial pop/rock/r & b/rap/country songs are all mastered for such loudness that they never fall below -10 (or thereabouts) ever. There's nothing about 24 bits that does anything magical at the top of the meter. If everything is completely remastered from the original analog mixes then sure, it'll sound better than the old releases, unless they get too shrill or squashed.


But the "deeper bass" thing is what really gets my goat. I've been in a million discussions about word length with audio files and no one has ever been dumb enough to say 24 bits gives you deeper bass than 16 with the same material, all else being equal, especially when transferring a finished mix. An acoustic bass in a very spare sounding recording may have better ambient detail, if another instrument isn't playing over the decays, but the only way a 24 bit version of a master has deeper bass is if the mastering engineer added it.

I guess saying "are said to have" makes it journalistically ok. 24 bits is the rule for input and the DSP gear is all higher bit internals, but let's just leave off the snake oil for anything as delivery medium.

 

[visualanté]

A 24bit sample is exactly 50% more than 16bit, how much that is per second depends on sample rate. Bit depth relates to dynamic range, sample rate to frequency response. The output is always a smooth curve as the DAC reconstruct an analog signal from the samples. Lower sample rates has the effect of a low pass filter according to the Nyquist theorem. Since it takes exatly two samples to represent a cycle, the highest frequency is half the sample rate, so 22050Hz for 44100 samples/second.

I'm trying to learn this so the DAC takes and interpolates the data to mathematically replace absent data I think right...so if there is more data there is less to interpolate which would mean a more accurate sound?

 

[csbo]

I'm trying to learn this so the DAC takes and interpolates the data to mathematically replace absent data I think right...so if there is more data there is less to interpolate which would mean a more accurate sound?

No. Additional bits give a lower noise floor, but no more accuracy.

 

[visualanté]

No. Additional bits give a lower noise floor, but no more accuracy.

noise floor is background noise right, so with 24bit little sounds won't be lost to static sound which would add to a fuller detailed recording?

 

[csbo]

noise floor is background noise right, so with 24bit little sounds won't be lost to static sound which would add to a fuller detailed recording?

To some extent. But we're talking incredibly quiet sounds when you are listening at high levels. And that's assuming an incredibly quiet environment as well, since mist rooms have too much ambient noise to hear such sounds clearly(don't breathe, it's too loud)

 

[Watersnaim]

Registered just for this thread

In the uk we have limited high res options.

This year HD tracks will be launching their HD service fully in the uk and hopefully iTunes will too and I for one encourage this.

I am not a really literate with respect to sampling rates frequency etc. However if I compare the few hires albums I did get from HD tracks e.g the eagles, the 24 bit version played via songbird blows the downloaded iTunes version or CD out of the water for clarity on my system which is not top end - naim 5si Quad classics and a Beresford DAC

Some of us are craving a flood of higher resolution files.

Maybe I am mad or maybe I have acute hearing

Do it apple please

 

[rntlee]

Link to the peer reviewed study that shows that. I am not aware of a single one.

No, a test conducted by Stereophile over an evening of cocktails does not count.

http://www.aes.org/e-lib/browse.cfm?elib=15398

Peer-reviewed papers indicating there may be audible differences for some do crop up occasionally. The differences are not "night and day" as some audiophiles might indicate. They are always described as very difficult to discern.

Personally, paying more for "high res" files is not something I'm likely to do. I can't differentiate above 256k mp3.

 

[tkermit]

In the uk we have limited high res options.

This year HD tracks will be launching their HD service fully in the uk and hopefully iTunes will too and I for one encourage this.

I think Qobuz is now available in the UK, too.

 

[MacFanatick]

Apple...

I'll take an iPhone 6 with a FPGA DAC please... Feel free to put them in all your devices while you're at it...

 

[Mago]

http://www.aes.org/e-lib/browse.cfm?elib=15398

Peer-reviewed papers indicating there may be audible differences for some do crop up occasionally. The differences are not "night and day" as some audiophiles might indicate. They are always described as very difficult to discern.

Personally, paying more for "high res" files is not something I'm likely to do. I can't differentiate above 256k mp3.

Actually it depends on which title are you hearing, try "lemon tree" the part the glass is broken on mp3 or aac you barely can discern what's broken, on flac 24bit@96kbps actually you can distinguish if where broken a glass or a cup or if it was made on glass or ceramic, no super human ears required.


But if you are earing hip hop, most rap and remixes, yes don't worth the bit rate there are nothing being capped by the codecs.

 

[Criptor]

HD Audio

Well I agree HD audio is nice. I am glad I didn't throw away my CDs like the old ipod commercials shown. However I am still concerned about first streaming taking over, second hard drive space for my large collection, the complete lack of any new features in the iTunes player with new updates in quite some time, and yes the fact that I have just about every song I could ever want in my life. I'll probably upgrade it from 128kbps (yes I downloaded from napster don't hate) but after it's 320 I really won't upgrade it for a long time, maybe 10 years or more. Most of the music is not in HD yet anyway 320 mp3 is still standard albeit FLAC is gaining popularity. And for crying out loud how many times can I re-download everything.

Now what about offering up all the international catalog to everyone around the world. You want to boost sales, stop putting country restrictions and you'll get quite a boost for a while. Allow anyone to download any song from anywhere. That might help, I've read comments from international users wishing they could get access to the US store, and US customers wishing to get access to the international catalog as well. Course not like The Pirate Bay hasn't been allowing this for years . But congrats on finally jumping on the HD bandwagon. Better late than never.

 

[jettredmont]

[Edit: Sorry, misread this as claiming the opposite of what it said, complete reading comprehension failure on my part. I still am surprised that HD Audio isn't all in 96 or 192kHz, though; it seems to be marketed towards the crowd that includes a huge portion of 96/192k true believers.]

Original post below only for posterity ...

A lot of DTS-HD Master Audio tracks are in 24bit 48kHz (96kHz isn't that common) format. Take a guess why exactly that is.

Because the size of the audio track on a Bluray doesn't matter, but specs on a sheet with higher numbers than the competition do matter in making sales? It doesn't cost any more to print that disk with extra bits than without, and the extra bits sell well to people convinced they hear something special in the ultrasound range?

DTS Master Audio is a brand of "HD" audio marketed at HD audio true believers. Of course most/all of it is in HD (I would be surprised if any were not, but looking at the DTS HD Audio description page it doesn't seem to be specified as a requirement). That isn't an endorsement of HD, it is capitalization of HD.

The "it sells, therefore is legit" argument is unfortunately a complete non-sequitor. Homeopathy sells, and it is just not-controlled-for-purity water diluting some active ingredient to the point that one atom of the active ingredient might be found in one of a trillion or more bottles but lots of the impurities of the source water are in every bottle. ******** sells, to be frank. That doesn't make it correct.

 

[jettredmont]

noise floor is background noise right, so with 24bit little sounds won't be lost to static sound which would add to a fuller detailed recording?

Noise that you are hearing in 16-bit audio is far more likely to have been from the original recording than an artificial noise floor from 16-bit audio. 16 bits is a huge dynamic range; 24 bits is 256 times as large. It isn't beyond scientific plausibility that more bits might benefit some people (there is far more variability to sensitivity to quiet sounds amongst the population than sensitivity to loud, near-ultrasound sounds), but chances are very high that you are not one of them (especially if you are over 20 years old, and especially if you are a male).

 

[jettredmont]

Actually it depends on which title are you hearing, try "lemon tree" the part the glass is broken on mp3 or aac you barely can discern what's broken, on flac 24bit@96kbps actually you can distinguish if where broken a glass or a cup or if it was made on glass or ceramic, no super human ears required.

Correct. There are many instances where lossy compression codecs remove or muddy sounds which are actually audible to the human ear. But, to be clear, there are three things which you are conflating:

1. bit depth (16-bit vs 24-bit) - this allows the DAC to distinguish loud versus soft sounds, roughly speaking. It "lowers the noise floor" with more bits.

2. sampling frequency (44.1kHz vs 96kHz in your example, 192kHz in the typical pro-HD-Audio post) - this allows the DAC to discern higher-frequency changes in the sound wave.

3. lossy vs non-lossy - lossy compression allows "most" of the waveform to be discarded while keeping "most" of the auditory effects.

(3) is what you are complaining about. Yes, that is absolutely correct. Lossy compression can remove important audio effects. In typical music it doesn't matter, but it can matter at specific spots in specific recordings. It all depends on what you enjoy listening to, but my impression of an audiophile is that they would rightly obsess over such minute changes in the audio. I wouldn't call even 256kbps AAC reference-quality audio. But, in 99% of the songs i listen to, it sounds identical to me, so for me, that's what I collect.

(2) is just scientifically unsound. The prototypical round-numbers human hearing range is listed as 20Hz-20kHz; a sampling rate of 40kHz can capture every waveform in this range accurately. See papers linked several times in this discussion for the theory as to why that is true. You are left with claims that people can hear beyond 20kHz (which is true for a small percentage of the population, greatly decreasing as you filter for adults), but then 44.1kHz sampling covers everything up to 22kHz waveforms. Simply put, the range of human hearing hits a wall between those two numbers. If you look at studies of near-ultrasound hearing in humans, the audible threshold line goes vertical pretty much right at 20kHz for the very young (as you get older the audible threshold line moves up, the vertical rightmost part of it moves left, and the pain threshold line above it moves down, all of which give you a much smaller area of audible but not painful sounds).

(1) may have some basis in fact, for some people. There isn't a wall here, and a significant variability; some people are sensitive to very quiet noises. So, it is at least plausible that 24 bits would help you personally. Not incredibly likely, but at least plausible.

So, that all having been said, if you are in fact comparing the same master, I'd expect that you might be able to hear the difference between a 256kbps AAC rendition of that glass breaking and the 24@96 FLAC recording. However, if you were to resample the FLAC recording to 16@44.1 using a good resampler, I suspect you would hear the same difference between the AAC and that FLAC recording, and have a very hard time hearing a difference between the two FLAC recordings.

 

[Mago]

So, that all having been said, if you are in fact comparing the same master, I'd expect that you might be able to hear the difference between a 256kbps AAC rendition of that glass breaking and the 24@96 FLAC recording. However, if you were to resample the FLAC recording to 16@44.1 using a good resampler, I suspect you would hear the same difference between the AAC and that FLAC recording, and have a very hard time hearing a difference between the two FLAC recordings.

You're rigth maybe not necessary HD audio as retina display since from the wole library of recordings only worth for few of them, but is welcome to have the audiophile choice at an source as iTunes, surely many people will start to download hd clips that don't worth the premium, but some with fine ears will have an opportunity to enjoy the best of the music on her pockets, previously only available at home.

 

[Yumbo]

iTunes needs tabs!

Safari has tabs, the Finder has tabs, ...

Now that be handy.

 

[Pman17]

That might be the theme of iTunes this year, HD. We'll get high quality tracks and support for H.265 will be added with 4K streaming on the new Apple TV. They'll say, "it's the first digital media store to offer 4K movies in H.265 and it looks gorgeous on the Retina MacBooks and 4K displays."

 

[Ozkiwitechgeek]

Audio Quality

I have had pretty much every format since the 80's for music and video so can share a little of what i have experienced in regards to quality of sound etc...

Super Audio CD was a brilliant ideas hampered by lack of content at the time and marketing

DSD is as close to Vinyl , real life as I have heard providing the mastering was recorded correctly

For the last two years due to moving cities / houses etc.. I tend to listen to AAC 256K format connected to a Cord Dock into Microlab Speakers, it provides a clean clear full range sound, BUT even I know this is lacking compared to even CD, DVD and especially Blu-Ray audio

Convenience is the fact of modern life, we all want it now and consume so much audio and video data

When I find a song or album from a new artist or past icon, I always buy the CD, DVD, Blu-Ray and sometimes Vinyl record version of their music

The only downside of these formats they are not portable

Pono player is a good concept however the design of the player is dead on arrival, it will not take off, it can't when everything on the planet in digital media is getting smaller, thinner, lighter

The humble iPod with an uncompressed ALAC, or PCM WAV file connected to a decent set of speakers or headphones can provide a full audio experience

Good in = good out, always been that way always will be, it applies to everything in life both organic and inorganic

All anyone wants is a good blend between convenience, price and best sound quality that can offer

Mass public need to be shown and told about how to get better sound, but the big electronic manufactures don't want to effect their bread and butter products that sell billions

I think in 3-5 years time SSD storage will be 500GB - 1TB and affordable so the future music players will be able to store larger more uncompressed audio files

It amazes me today you do not have to spend thousands let alone many hundreds of dollars to get really good quality audio

Interesting topic and could go on for ages but just adding my 2 cents worth

 

[EHFink]

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..

 

[milo]

I think the ABX test is a bit overkill for this. A simple blind test is enough. A v. B. Ask which is better? However, do not prompt or prime the person first.

ABX is better because it's a straightforward way to find out if the person can even tell the difference at all. With "hey listen to this...now this" it's just too easy for the person doing the test to give away which is which.

Free ABX software is available for mac, if people can really hear the difference they should be able to consistently pick out the difference on an ABX test.

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?

I assume because they want to have access to the music indefinitely and not lose it when a subscription ends. And to listen on devices that aren't capable of streaming. Spotify doesn't allow downloading all those song files to keep, does it?

 

[milo]

Try this: buy the DVD or Blu-Ray version of your favorite album. Compare it on a surround sound system to the CD version. Which one sounds better?

Bogus test since they are likely different mixes. And you're saying to compare bit and sample rates by listening to one format that's in surround? Is it really that hard to figure out why that defies common sense?

A much better test: get a 24/192 recording, use decent software to sample rate convert it to 44.1/16, then do an ABX comparison.

Sure there's no scientific reason to believe that 24/192 could ever improve audio quality in any way (nor are there any legitimate blind tests which show people successfully picking the 24/192 file over the 16/44.1).

I agree with you that 192 is a wild goose chase. But 24 bit, while subtle, is a genuine improvement in sound quality. It requires very good equipment and an extremely quiet listening environment so hardly anyone is going to hear the difference, but it does exist. But is it worth using as a delivery format? I don't think so, unless it's 24 bit masters used for lossless or lossy compression.

Oh, and ABX double blind tests often fail because all "switcher boxes" just murder sound quality. You should try manual double blind testing. It's more of a bear to co-ordinate but it's also a better test because you're not including the effect of crappy switching equipment in your testing results.

So have you done real double blind testing to verify that you can actually hear the difference you think you can hear? Claims of telling the difference are worthless without a listening test to back them up. And for ABX instead of using a switcher box just do it with software, no switcher box needed.

Me too,
The insistence of many on this forum that no one can or will appreciate better audio because they can't is childish.

That insistence is because of the results of many rigorous listening tests. I haven't read through the whole thread, has anyone here done an ABX test and consistently picked out the difference between 44.1 and 192?

 

[kemal]

Soapbox

Bitz - Schmitz.

My 12 year old $1500 CD player has a $45 CD mechanism that feeds its digital into some pricy D/A and a lot of well thought out analog circuitry. That's what makes it sound good. They took into account:

1) the quality of the digital filter: phase shift, ghosting, etc.
2) the quality of the D/A: Jitter, noise, etc.
3) the quality of the analog preamp and headphone amp: single-ended/balanced, capacitor coupling, feedback, etc.

Few IT devices such as iPods have anything going for audio quality. Pono may just get this right.

 

[milo]

But if you've ever heard hi-res on a quality DAC with a good stereo system, you wouldn't say it's a placebo effect.

You know what else? If you ever took a sugar pill and it made you feel better, you wouldn't say it's a placebo effect.

That's the whole point of the placebo effect, it's a psychological issue where people imagine an improvement that doesn't actually exist. You listened with all the hype in place (hi-res! quality DAC! good stereo!) instead of doing a blind comparison. Sounds like you didn't even compare the same masterings of that album.

check
http://www.tested.com/tech/1905-the-real-differences-between-16-bit-and-24-bit-audio/

Frankly that writer hasn't the foggiest understanding how digital audio works. Completely worthless. It's never a good sign when an article about higher resolution doesn't even mention dynamic range.

 

[oneMadRssn]

Free ABX software is available for mac, if people can really hear the difference they should be able to consistently pick out the difference on an ABX test.

I'll bite. Do you have any recommendation of what ABX software I should try?

I have a pretty stellar Neil Young recording in FLAC that I just converted to MP3.

 

[milo]

I've used ABXer but it looks like these days ABXTester may be better supported.