I’ve been listening to quite a few new tracks from iTunes under Apple’s new “Mastered for iTunes” program. This is something Apple voluntarily started to help improve the audio quality of the music we download from iTunes. From my perspective, it’s great, but Apple has received some self-indulgent criticism in some circles.
For instance, a Web site in the UK wrote an article refuting claims that Mastered for iTunes sound quality was comparable to CDs. The Web site even went so far as to make a video using null testing to prove its point.
Here’s the problem. Apple never said that Mastered for iTunes was CD-quality at all. The Web site later updated its posted admitting that one of its commenters, not Apple, made the comment about CD quality.
From what I can see, Apple only wants to provide the highest quality audio it can. What can possibly be wrong with that? The simple answer is nothing.
Here’s what Apple’s mission is in the Mastered for iTunes program as I understand it.
1. Provide better quality audio to its customers
2. Give mastering engineers tips and tools to submit the best quality audio possible to Apple
3. Make sure that audio professionals are optimizing content for the digital medium.
Apple is now accepting high-quality masters from engineers because its new encoders can preserve more dynamic range, resulting in better quality music for the consumer to download.
Mastering engineer Bob Ludwig (Coldplay, Bruce Springsteen, Nirvana, Soundgarden) told me that Apple spent 18 months investigating how to make the AAC encoder better. The results, he says, are “a great thing.”
Ludwig explained that in the past, AAC songs were ripped from 16-bit CDs, but now Apple can handle higher resolution files. In fact, Apple says an ideal master will have 24-bit 96kHz resolution, but sample rates of 48kHz, 88.2kHz, 96kHz, and 192kHz can also benefit from the encoding process.
The creative part of the mastering process hasn’t changed one bit with the Mastered for iTunes program. Ludwig said he masters his music exactly the same way. The change comes after the creative process is finished — that’s where Mastered for iTunes comes in.
Instead of sending it to CD and then having the AAC files ripped from that, mastering engineers can now use Apple’s tools to create custom AAC files from the high-resoultion master.
Apple even documents the trend towards louder mastered music, commonly referred to as “the loudness wars” and offers tips and advice. With music being so loud these days, running a song through any encoder will often cause clipping — it’s the nature of encoders. However, one of the tools Apple provides to engineers allows them to check the AAC files for clipping and then make adjustments before the songs are sent to iTunes.
Some people think that mastering for a digital format is silly, but it’s downright stupid to ignore iTunes. Apple is the largest distributor of music and if you don’t think that improving the quality of your music for millions of people, then you don’t really care about your fans.
As a musician and a music lover, I’m in favor of any process that captures more dynamic range of the music I’m going to buy. It’s a good thing for everyone involved, but most especially for the consumers that purchase and listen to the music.
Engineers and musicians should embrace better-quality digital music, not fight about frequencies and philosophical differences in the process. Your customers will thank you for it.
Great info from Bob Ludwig, thanks.
Hope they will upgrade our already purchased albums and not want to rebuy them all again.
My main concern is that AAC (and MP3) is in fact indistinguishable from uncompressed audio (perceptually lossless) at fairly low bitrates — lower than the great quality iTunes already provides. (You can go to Hydrogenaudio to look at the listening tests.) So what exactly are these new masterings doing to the original recordings? Apple is making it sound like these guidelies will help remove AAC artifacts, but double-blind tests have shown that there aren’t really any audible artifacts in the vast majority of encodings. This means that the new masterings will be compensating for something that’s not actually there, thus reducing the audio quality.
From what I understand, there should really be no difference between encoding an AAC from a CD and encoding an AAC directly from the high-fidelity master. “Optimizing content for the digital medium” doesn’t make much sense to me because audio engineers have been doing this ever since CDs first came out. If we get more dynamic range from these new tracks, then that’s wonderful, but it has nothing to do with the fact that they were encoded with AAC in the end.
In short, I guess I just don’t understand where the “for iTunes” part comes in. “Remastered for Apple” would make much more sense to me, but as it stands, it all sounds a bit too snake-oil-ish.
“My main concern is that AAC (and MP3) is in fact indistinguishable from
uncompressed audio (perceptually lossless) at fairly low bitrates —
lower than the great quality iTunes already provides.”
?????
You’re deaf. And ignorant.
I’d bet money that you’ve never been anywhere close to a real (not bedroom) recording studio because it’s obvious that you don’t have a clue – not even an inkling of a clue – when it comes to audio quality.
“From what I understand, there should really be no difference between
encoding an AAC from a CD and encoding an AAC directly from the
high-fidelity master.”
That because you don’t UNDERSTAND anything about audio. I’ve been involved in professional audio since the early 1970s. Anyone who tells you there’s no difference is either hearing-impaired or has never had the opportunity to listen to well recorded material over a high quality system.
Why do you think “there should really be no difference between encoding an AAC from a CD and encoding an AAC directly from the high-fidelity master”?
A master has a higher sample rate and a higher bit rate, resulting in cleaner highs, better harmonics, and more dynamic range than a CD.
With a CD, a lot of the information, including dynamic range, has already been thrown out. So yes, starting with a master can lead to a better AAC, because AAC allows for more dynamic range than a CD.
There is no need to get confrontational. It’s true that I’m not an audio engineer, but the folks at Hydrogenaudio (and elsewhere) have been doing double-blind audio tests to test the veracity of such claims for years, and I trust their scientific methods and results. They have a number of exceptional audio engineers in their community as well.
The main issue is that the human brain is very suggestible, especially when it comes to audio. A listener may very well believe they hear improvement when they know that they’re comparing, say, 192kbps MP3 and WAV, but if they do the test blind, they will not be able to tell the difference except for specially-constructed problem samples. This has been shown again and again, and it’s the main reasons why companies like Monster Cable continue to make big bucks off of so-called audiophiles.
Do you have any double-blind tests to support your claims? By all means, please prove me wrong — I’m always happy to improve my knowledge.
Well, since AAC has the same resolution and frequency as CD (16bit/44.1khz, last I checked), what difference does it make at what point the high-fidelity information is thrown out?
I admit there may be some technical issues that I’m overlooking. I’m not too familiar with the way the AAC encoder works under the hood, or with the way downsampling is usually done in the recording industry. Why is it better to let the encoder do the downsampling?
From the very article we are commenting on: “Apple says an ideal master will have 24-bit 96kHz resolution, but sample rates of 48kHz, 88.2kHz, 96kHz, and 192kHz can also benefit from the encoding process.”
But to fully explain, well, how versed are you in signal processing. I don’t mean to sound arrogant, just that to explain requires both you (and I to be able to explain it well) need to have a lot more knowledge in this field.
A science fiction writer I highly admire likes to complain that Apple “ruined” the quality of music. I suspect he doesn’t keep up with new developments, as Apple attempts to genuinely improve the commodity that has become the way most people now buy and listen to their music.
You’re probably right — I don’t know too much about the signal processing aspect of it. In any case, it’s definitely great that they have their hands on these high-fidelity masters.
By the way, for some non-self-indulgent discussion of the issue, I recommend this thread: http://www.hydrogenaudio.org/forums/index.php?showtopic=93628
I should stress that I’m in no way affiliated with HA. It’s a great forum with bullshit-averse members that taught me many interesting (and often counter-intuitive) things about audio. I also fully admit that I might be wrong on many counts in regards to this issue. I just care a lot about the quality of my audio, so I’m naturally skeptical with so much FUD about lossy encoding circulating in the industry — even among professionals.
the tech:
http://www.apple.com/itunes/mastered-for-itunes/
the music:
http://www.itunes.com/masteredforitunes
Since my original tests were “self-indulgent”, I don’t see that it can do any harm to answer your points !
http://productionadvice.co.uk/mfi-was-i-wrong/
One other thing about your post specifically – you’re right that it was “a commenter” on my blog who made the claim that MFi sounds “closer to CD” – but remember that commenter was a representative of the mastering firm responsible for “I’m With You” and also apparently the whole “tweaked masters for AAC/MFi” publicity in the first place.
I think Mastered for iTunes is a good thing. But if the music producers still use hypercompression (i.e. make their music “loud”) the sound quality still sucks.
E.g. Red Hot Chili Peppers’ “I’m with you” has horrible sound quality (DR3, no headroom, horrible clipping).
Aft 48khz sample rate, who gives a shit?