Throughout my library I'm seeing cases where some songs from a given album that I've imported from CD are matched by iTunes Match, while others — on the same album — are not recognized and are instead uploaded (at a lower bitrate):

enter image description here

What accounts for this difference in treatment of tracks from the same album? Is there anything I can do to acquire the high bitrate tracks?

  • Note: This question is distinct from an earlier very narrow one (apple.stackexchange.com/questions/27571/…) about a similar situation regarding purchased tracks on a specific album and Match beta.
    – orome
    Nov 16, 2011 at 20:38
  • 1
    Note also that this is a matter or considerable concern: the point of paying for Match is get access to the bitrate upgrade, but for this (typical) case, that has happened for only 30% of my tracks.
    – orome
    Nov 16, 2011 at 20:46
  • Did you mean at a lower bit rate?
    – billygoat
    Nov 16, 2011 at 20:48
  • A question about how to fix this has also been asked.
    – orome
    Nov 17, 2011 at 21:53

2 Answers 2


It's been confirmed elsewhere that metadata alone will not give success in matching (if it were, someone would quickly find a way to just create thousands of dummy files with the right metadata). Some audio fingerprinting is going on, and you can test it by changing the metadata of one track to that of another one, and see if if matches the audio or the metadata. Still, metadata plays a role as well.

MacWorld have a note on the problem of one or a few songs not matching from an album: Basically, it seem to affect specific songs (rather than just randomly failing every n:th song). Two possible reasons for this:

  1. The audio or metadata specific to that track confuses the matching algorithm (as we know, audio fingerprinting is not an exact science)
  2. There is a rights issue with particular tracks that Apple haven't been able to find agreement for.

I think the problem stems from the fact that the album metadata with Itunes is different to the ones that are in your CD.

You can browse that particular album in Itunes store and change the metadata in your local library to match the ones in the store. This way the matching program will find all the songs.

Edit: The album's Metadata may not be the only dimensions used while performing the matching, but I am sure that, it is among the more dominant dimensions. Itunes match supposedly uses Gracenote's MusicID and it seems to use multiple identification techniques. More here.

  • Is (1) that the case even when only some tracks on an album are affected? If so, (2) how do I locate the corresponding album in the iTunes store; (3) how do I determine which metadata to change (I could wind up having to change the metadata on thousands of tracks)? And in any case, my understanding was that Match preserved user metadata; this would mean that that's not such a great feature.
    – orome
    Nov 16, 2011 at 21:17
  • That is the problem with varying metadata. I found out the hard way.. I had lost my Itunes purchases and Apple allows you to download your lost purchases. When I downloaded, some of the files were missing and I was told that the reason was that the data associated with the albums had changed ...
    – billygoat
    Nov 16, 2011 at 21:37
  • 2
    It was also my understanding that Match uses a matching algorithm based on the file's waveform, not (only) its metadata. @doc_180, can you substantiate your answer? Does your solution actually work, in your experience?
    – Dan J
    Nov 16, 2011 at 21:37
  • @djacobson I will do some more research. Can you please tell where you read about the matching algorithm?
    – billygoat
    Nov 16, 2011 at 21:38
  • @djacobson: Indeed, metadata would seem the wrong way to go about this. Certainly, based on the metadata in friends' and coworkers' the libraries I've seen, Match would have a challenging time matching if it relied heavily on metadata.
    – orome
    Nov 16, 2011 at 21:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .