iTunes Match and Mastered for iTunes: Which Tracks Do You Get When Matching CDs?

002.png

Apple touts its Mastered for iTunes tracks on the iTunes Store as “Music as the artist and sound engineer intended.” Mastered for iTunes tracks are therefore supposed to sound better than tracks you rip from CDs. The basic goal of Mastered for iTunes is to provide a direct downsampling of music from 24-bit, 96 kHz files to 256 kbps AAC files, rather than having a first downsampling to the CD format (16-bit, 44.1 kHz), then another conversion to AAC.

Whether or not the difference is audible is debatable; at a minimum, the conversion from higher quality masters can be seen as producing fewer “rounding errors,” though it’s a bit more complicated than that.[1] Also, the Mastered for iTunes process requires that record labels make some slight changes to their files, notably to avoid clipping,[2] but it allows them to create different masters for the iTunes Store than for CDs, if they so desire. (Yet as you’ll see below, I found much more clipping on certain Mastered for iTunes files than CD rips or older purchased files.)

I was wondering what link exists between Mastered for iTunes files and iTunes Match. If you rip a CD, and match it, will you get Mastered for iTunes files, which are currently sold on the iTunes Store (if available), or will you get equivalents of the CD’s ripped files? Since Apple says there’s a difference between the two, how do they handle this?

When you purchase a Mastered for iTunes file, there’s information in the file telling you this. You can see this by selecting a file, pressing Command-I (on Mac) or Control-I (on Windows, then viewing the Summary tab. The Mastered for iTunes badge below shows you that you have premium files.

006.png

If you rip a CD, match it, then download one of the files from the cloud, you don’t ever see the Mastered for iTunes badge. I have a handful of CDs for which only Mastered for iTunes files are available from the iTunes Store (these are new releases where labels only provide files for this format). I added them to my iTunes Match account, matched them, deleted my originals, then downloaded the matched files. I compared them with my original rips (using the methods described below), and saw that these files were not the same; I was clearly getting the Mastered for iTunes files from iTunes Match. But the files don’t display the Mastered for iTunes badge.

So for files that are matched, then downloaded, if there is only a Mastered for iTunes version on the iTunes store, the files you download will be in Mastered for iTunes format; yet you won’t know that. The Mastered for iTunes badge is displayed only when there is a bit of metadata in purchased files that tells iTunes to display it.

This begs the question: for releases where there were originally standard CD rips, does the iTunes Store keep older copies of files for matched tracks? Case in point: I bought all of Pink Floyd’s albums from the iTunes Store several years ago. Apple brought out a new set of these files in Mastered for iTunes format some time ago, but if I download the files from iTunes Match, do I get the original files, or has Apple replaced them with the newer files? In the former case, they may want me to buy the files again, because the newer versions presumably sound better. But if it’s the latter – they still only keep one set of files – my purchase would be a waste of money, since the files would be exactly the same.

It’s all very confusing. When I look at the Wish You Were Here album on the iTunes Store, I see that it is Mastered for iTunes, and it shows as Purchased for me.

iTunes001.png

So I wanted to know: when I download this from the cloud – from my Purchased list, or from iTunes Match, which shows the song as Purchased – which version of the track would I get?

To find out, I did an experiment. I took three files for the song Wish You Were Here. The first (1) was my purchased file. The second (2) a CD rip, at the iTunes Plus bit rate (which matches what Apple uses on the iTunes Store). And the third (3) was a purchased copy of the Mastered for iTunes version of the song.

I compared these files in Visual Differ; this app looks at the actual data in the files and shows any differences. With Visual Differ – or any other diff tool – you see each difference between two files, whether these are individual characters character (data shows up as characters), or whether the different section is very long. Each difference may be one character or 10,000.

When I compared files (1) and (2), there were only three minor differences, presumably differences in metadata. But when I compared files (1) with file (3) – the Mastered for iTunes file – there were 192 differences, many being very large sections of the files, showing that the actual music data was slightly different.

What is interesting, however, is that when looking at waveforms, the Mastered for iTunes file shows a bit of clipping; it’s clearly louder, and there are some peaks that shouldn’t be there. Below are the waveforms for files (1) and (3) in Fission.

Here’s the older file:

004.png

And here’s the Mastered for iTunes file:

005.png

As you can see in the second file, the overall volume is a bit louder, and there are some volume peaks in the second loud section. This isn’t audible, but it’s clearly at the limit of what should be done to a music file.[3]

But it gets more complicated. I compared several Bob Dylan tracks that I have in my library. I bought the Bob Dylan: The Collection back in 2009. I had updated the files to iTunes Plus versions, and very recently, new, remastered versions appeared on the iTunes Store. So I compared a few of these files. I took my 2009 upgrades and compared them to files that I downloaded from iTunes Match, which showed as Purchased.

In all cases, the files were different. The newer files are slightly larger (after removing album art), and there are many differences visible when comparing them in Visual Differ. When looking at the waveforms I can see differences in some of them, but they’re not as blatant as with Wish You Were Here above. Also, the newer files have much more clipping than the old ones; again, something that shouldn’t be the case with Mastered for iTunes files.[4]

To be fair, some files show a more judicious use of compression in the newer versions. Compare two versions of the Grateful Dead’s Ripple, from the album American Beauty. Here’s the waveform for my own CD rip, using CDs from the Grateful Dead’s remasters on the Golden Road box set of 2001:


Fission002.png

Now look at the file I downloaded from iTunes Match, which is certainly the newer Mastered for iTunes file:


Fission001.png

In the latter file, the audio looks less compressed, allowing more detail to come through. (I haven’t given a close listen to these files to be able to say if there’s a noticeable difference.)

Another example: Peter Frampton’s Something’s Happening. I compared my old file with a new download. There are many differences between the two. The old file has 78 clips; the new one has 12,042 clips. It’s noticeably distorted.

One more example is U2′s Zoo Station, from Achtung Baby. There are many differences, but at least U2′s engineer has mastered the clipping a bit better than Peter Frampton’s. There are no clips in the old version, and 11 in the new one.

Finally, The Clash’s London Calling. I have the recently released Clash Sound System box set, and compared my CD rip with the matched version; there are a number of differences, though just two clips in the ripped version and one in the matched version. Both are the most recent remasters, so if there are differences, the only explanation is that the matched files are Mastered for iTunes.

So, to sum up, here’s what you get.

  • If you rip a CD, and the only file on the iTunes Store is Mastered for iTunes, that’s what will get matched. When you download the file, it will be the Mastered for iTunes version. But the files won’t show the Mastered for iTunes badge. But it will never be the same file as your CD rip.
  • If the CD in question was available on the iTunes store before Mastered for iTunes, and there’s a newer version of the files in Mastered for iTunes format, you may or may not get the Mastered for iTunes file. I don’t see any logic to this, and it would require many more tests to really find out for how many artists you get Mastered for iTunes files, and for how many there exist older versions of files. It may be that the Pink Floyd is an anomaly.
  • If you match high-resolution files in Apple Lossless format, then download the matched files, you may see the Mastered for iTunes badge. I tested a handful of 24-bit 96 kHz files, for which the iTunes Store has Mastered for iTunes versions, and, of the ones that matched, some of them showed the badge, but others didn’t. I have no idea what causes this inconsistency.
  • Finally, it seems that for some or most of your purchased tracks, re-downloading them, or getting them from iTunes Match, will give you different tracks. You may or may not want this; given some of the examples of clipping I’ve seen, it looks as though Mastered for iTunes tracks may actually be worse than older tracks.[5]

What all this means is that it is very possible that the file that you download from iTunes Match is not the same as the CD rip that you matched.

Apple’s Mastered for iTunes is confusing. It’s similar to when Apple started selling its iTunes Plus format, which were music files at a higher bit rate (256 vs 128 kbps) without DRM. In the early days, you could pay to “upgrade” older files, but these upgrades went away when iTunes Match was introduced.

Apple can’t do the same thing with Mastered for iTunes. This format depends on the record labels providing higher resolution masters [6], and, while most new releases are recorded at these resolutions, this is not the case for older releases, or small labels, who may not have the know-how to work with these files.

So we’ll never see all iTunes Store files in Mastered for iTunes format. But I think Apple should not have Mastered for iTunes files as a premium format; they should clearly indicate which matched files are in this format. If anything, this would be a reason to subscribe to iTunes Match. I doubt anyone will buy the Mastered for iTunes versions if they already own older versions; the differences, if any, are very subtle, and nothing like the doubling in bit rate we saw with iTunes Plus.

Apple should probably make Mastered for iTunes more ubiquitous. This will enhance the iTunes brand, since other retailers, such as Amazon, don’t (yet) offer anything similar. Again, I don’t hear a difference, but I haven’t compared a lot of files. In some cases, labels and engineers may actually do a fair amount of post-production work on the files they send to Apple to use for Mastered for iTunes files, so differences may be apparent.

This said, many of the Mastered for iTunes files I looked at have a lot of clipping, something that Apple clearly warns engineers to be careful of. So perhaps Mastered for iTunes isn’t such a good thing after all…

Update 1: A Twitter follower asked about mono recordings, such as recently re-released sets by Bob Dylan, The Beatles, or Miles Davis. I checked some of mine – I have the Dylan and Beatles mono sets – and the waveforms don’t look any different. However, I rip mono CDs in true mono (as explained here); in other words, I tell iTunes to rip a single track, not two mono tracks in stereo. This means that my rips, which are usually 256 kbps, are only 128 kbps for mono albums.

When I matched some of the files – only the Dylan matched; the iTunes Store doesn’t sell the Beatles mono releases – the downloads I got were in 256 kbps, and twice as large. They have double mono tracks, instead of a true, single mono track. It’s not a big deal, but it’s worth bearing in mind.

Update 2: I started a thread on the Computer Audiophile forum to see if anyone else has done this kind of testing. What’s interesting is that one user sees the Mastered for iTunes badge on CD rips that are matched (when he downloads the files). I don’t see this on any CD rips.

However, I tried matching some high-resolution files, and a few of them show the Mastered for iTunes badge when I download the matched files.


  1. See Apple’s Mastered for iTunes PDF.  ↩

  2. Clipping is distortion that occurs when music is too loud.  ↩

  3. Apple’s pre-check tool, afclip, reports a total of 57 clips for this file; it should never have been accepted as a Mastered for iTunes file, since Apple’s guidelines are very clear about avoiding clipping. For that reason alone, the Mastered for iTunes file here is of lower quality than the original.  ↩

  4. Using afclip on Visions of Johanna, for example, shows that there are no clips in the 2009 version, but 60 clips in the new version.  ↩

  5. Apple performs the actual conversions to Mastered for iTunes files from the masters that record companies send them. Engineers have tools to be able to make test conversions, to hear how they’ll sound, but they don’t send final AAC files to Apple.  ↩

  6. Note that this isn’t the only difference with iTunes Match downloads. I’ve seen truncated tracks when downloading from iTunes Match, and I’ve gotten lots of emails from people who don’t like the fact that their “explicit” tracks – those with raunchy lyrics – are replaced by “clean” versions. Also, I found, in a couple of cases – including the Pink Floyd song – that the matched and downloaded tracks were not the exact same length or speed as the CD rips. This is puzzling; it suggests that Mastered for iTunes tracks may be tweaked in odd ways. Changing the speed of a track – which I saw with Wish You Were Here, and with a Jerry Garcia song – makes no sense.
     ↩

Follow @mcelhearn on Twitter.

Share this article:Tweet about this on Twitter0Share on Facebook0Share on Google+2Share on LinkedIn0





9 replies
  1. Johnny says:

    Perhaps worse are tracks that are simply different. The “Crazy Train” track that appears on an Ozzy Osbourne best of disc (that iTunes radio does use) includes different drum and bass parts over-dubbed during a dispute w/ the original session musicians. It sounds very different. The original albums were re-issued with the original drum and bass parts, yet I’ve yet to hear those versions on iTunes Radio since that best of album never got the update.

    When I did iTunes Match right when it came out, I was a bit horrified to find that the Pearl Jam song “Footsteps”, which is normally just vocals and 1 guitar, had harmonica parts added to it! My original rip of that song had no such thing. The imprecision, ambiguity, and inconsistency of these issues are why I listen to my own ripped music whenever possible, vs. some cloud thing.

    Reply
  2. Christopher says:

    I matched Nora Jones’ “Come Away With Me” album from a cd rip and re-downloaded the tracks. About half of the songs have the “Mastered for iTunes” logo. Thanks for the info of how to compare. It will be interesting to see if the other tracks that don’t have the “Mastered for iTunes” are indeed mastered for iTunes.

    Reply
  3. Ruediger Stang says:

    Hi Kirk,

    thanks for the interesting article, enjoyed reading it very much.
    iTunes Match is a great idea, and for most of my music, it was a boon (automagically getting improved quality, especially for older files). However, it does have its problems, especially with songs containing explicit lyrics. Very often, streamed or downloaded versions of songs are the clean version, although the adult one was originally matched. This for me is a lot more annoying than getting a file with a slightly different sound quality. I think this is an even more important issue Apple should fix soon.

    Regards,
    Ruediger

    Reply
  4. Cameron Saemann (@camerons) says:

    I will usually upload albums I buy from Amazon MP3 to my iTunes Match account, and if the album is a newer release as you state in your post, I will get the iTunes Match version with the badge. I have to delete the file and redownload, keeping the new file on my Mac before I view the metadata in iTunes, for the badge to show up.

    I’ve also noticed more clipping and “loudness” on Mastered for iTunes versions. I purchased the new version of Nevermind from Nirvana and the waveform was much larger and appeared to be clipped compared to my old version.

    Reply
  5. Jeff Hume (@jeffhume) says:

    My experience with the mono Beatles and some other mono releases was that it would sometimes match to stereo versions, which is very annoying. Some songs matched some uploaded in the same album. Do you have to do true mono to ensure that this doesn’t happen?

    It still annoys me that there’s no force upload feature. There’s some albums where I really care about the mastering, but there’s no way to ensure I get the right version with iTunes Match

    Reply
  6. Adrian says:

    I recently uploaded Dark Side of the Moon (remastered) that I’d got from Amazon to iTunes Match. Only one of the tracks came back with the ‘Remastered for iTunes’ logo and the difference in sound quality is noticeable. What is more, the second track was audibly clipped on the matched version – there was nearly half a second difference in length between the original Amazon mp3 and the version from iTunes Match. On a concept album, that kind of clipping is clearly a problem!

    I then purchased the iTunes version of the track, to see if that would give a smooth transition between the two tracks – it did. However, even though this version was ‘Mastered for iTunes’, when I ‘matched’ it, the re-downloaded version I got was the non-Mastered for iTunes (and clipped) version again. After much toing and froing between the Apple support desk and their engineers, it turns out that iTunes Match did not like the fact that I’d purchased my ‘Mastered for iTunes’ version from a different account to the one I have iTunes Match set up for…

    The long and short of it is that, at the moment, which tracks you get (Mastered for iTunes or otherwise) via iTunes Match appears to be bit of a lottery. I would say that Apple does keep more than one version of an album on its servers, regardless of whether or not they only sell a ‘Mastered for iTunes’ in the iTunes Store and that they still haven’t quite got their matching algorithms tuned optimally for the matching process…

    Reply
  7. mikey says:

    Nice article and comments. This is one reason I’ve held-off on bothering to use the service all this time.
    What is it about Apple these days, they keep releasing services that are half-baked, then either take an age (like YEARS!), or never bother at all, to do anything to fix them (thinking about iWork late 2013 update that’s completely incompatible with previous versions in order to get the iCloud versions to work).

    More importantly, isn’t it about time Apple upped their game and gave the option of buying/downloading/upgrading collection to ALAC (Apple Lossless) versions by now (many like myself rip to this, then transcode on moving to our iOS devices in iTunes settings, for smaller local device usage).
    Yes, many will not want them on their iOS devices yet (local storage isn’t there yet, even with 128GB storage devices which are welcomed), though they can at least transcode locally as per iTunes settings to do so that already exist. But a great many would LOVE to at least have them on their Macs/PCs – many of us have 100Mbps down broadband, so can easily handle these over our bandwidth, and MASSIVE storage too (I have 20TB’s available to me on a Pegasus R6, FFS!).

    Come on Apple, offer some market differentiation to make us want YOUR product, over competitors!

    Reply
  8. Adam Schoales says:

    It gets further complicated…

    I’m one of *those* guys who has vinyl rips in his iTunes. I know that may seem counter intuitive to most people but there *is* a reason.

    I downloaded my vinyl rip of Off The Wall to my iPhone the other day through iTunes Match and noticed something odd. I was getting a mix of both CD/iTunes versions and my original vinyl rips. How do I know? Well, it relates to why I have the vinyl rips in the first place.

    See, the CDs were made with alternate mixes for some of the songs than the original vinyls were. So when I downloaded the matched song and heard hand-claps on my “Rock With You” I knew that iTunes had delivered me a CD/iTunes version of the song. However when the track “Get On The Floor” played at a considerably quieter volume I realized I had the original vinyl version of the song (which also had a different mix on the CD). So for some reason iTunes match was delivering me a mix of CD/iTunes versions and my actual matched vinyl rips.

    This is probably a pretty niche case, but frustrating. I suppose the way to ensure my vinyl rips are saved to the cloud as vinyl rips is to tag each song title with “vinyl rip” so that iTunes doesn’t just assume it’s the same as what’s already in their library, but I honestly don’t know.

    Moreover, every year I make a Christmas Mix CD for friends and family. I spend time working on transitioning all the songs and adding little interludes here and there. Anyways this year I got a new Mac just before Christmas and downloaded all my previous mixes. I was shocked to discover that when I downloaded them many of the tracks came down as their original CD versions which means my minor changes didn’t come with them. So songs didn’t segue into each other in some cases, other songs were missing little enhancements I added. Again, a niche problem, but a frustration. In the end I simply copied over the files from my other iTunes library which resulted in iTunes becoming incredibly confused and showing them as “duplicates” along with the matched versions in the cloud.

    So while I love iTunes Match in theory, in practice it’s a different story.

    Reply

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply