Skip to main content

Amazon music storage

Topics/tags: Rants, music, the cloud

People who know me know that I accumulate things. Since my early days of College [1], I’ve collected popular music. I used to collect vinyl. When CDs came out, I started accumulating CDs. And, when digital music came out, well, I kept accumulating CDs, but started ripping them [2]. One of my first big personal projects was a huge database of my albums. I could tell you where I bought any album and how much I’d paid, which albums any particular song appeared on, and other trivia both useless and useful. Unfortunately, I wrote the damn thing in HyperCard, and so it is now long gone [3].

In any case, I have a lot of music in both physical and electronic form. Hence, when Amazon started offering a cloud storage service of 250,000 songs for $29.99/year, I jumped at it. 250,000 songs is somewhere between 10,000 and 25,000 albums. I’m not even sure that I have that many. In any case, I figured that it would be a way for me to get everything together and in order.

But, well, ripping CDs takes time. Tagging them appropriately and adding covers adds more time. So I never quite got around to uploading everything I ripped, let alone everything I had. It feeels like every time I dig out an old external drive, I find a folder labeled Upload that contains a few discs I’d ripped at some point that I had intended to upload. Sometimes I did, sometimes I didn’t.

Early this year, Amazon announced that they were discontinuing Amazon music. Folks who timed their accounts correctly could continue to upload until the end of this year. Other folks might have a less generous deadline. Mine was September 10 [4]. I had four goals for that date: Arrange the music I seem to have put on too many external drives, tag and add images to the music that needed it, upload everything, and download what was already on Amazon music [5].

But, well, the summer was way too busy. So while I made some progress, I didn’t make enough. The week of September 3 I spent downloading what I already had on Amazon, just in case. Of course, I encountered some problems with the downloader, which meant that I had to restart a lot. What problems? It seems to give up if it encounters a song with no length. It also doesn’t like to download more than 1800 songs at a time before freezing up. So I’d start it going, do some other work, and glance over from time to time to see if it was making progress. Of course, it doesn’t tell you when it’s stopped, so I wrote a small program that counted how many songs were downloaded and printed it out every minute; when that number stopped changing, I knew that it was time to quit and start again [6].

How many songs had I attempted to upload in the years that I had Amazon music? Slightly under 60,000. Clearly, I had not spend as much time uploading as I planned.

But what about the last-minute rush to get more songs onto Amazon Music Storage? I didn’t get to uploading until the weekend; September 8, or so. I kept telling myself First you should get the music in order. But once I hit the 8th, I gave up and just uploaded what I could find. Unfortunately, the Amazon uploader is even worse than the Amazon downloader. It would tell me things like 2,000 songs uploaded, 300 songs matched, 200 songs could not be uploaded because of a format error, 100 songs could not be uploaded for an unknown error, 50 songs could not be uploaded because of a server error. Would it tell me what songs could not be uploaded? No; that would make too much sense. So, whenever a batch upload would fail, I’d have to redo the same batch upload and wait for Amazon to realize that most of the songs had been uploaded already. A batch of 2000 or so songs would usually take three or four tries before Amazon finally accepted everything. Hence, even though I had a bunch that I expected to be able to upload, I could only upload some of them. I guess I’ll not achieve my goal of having an organized music collection, at least not on Amazon.

That may be for the best. I’ve been looking at what I got from Amazon and, well, Amazon does not do a particularly good of either naming the downloads or, in the cases it adds an album cover, of choosing the right cover. Take, for example, Ani DiFranco’s Living in Clip. I’m not sure why I looked at it; I seem to recall that I’d chosen it because it’s a multi-disc album.

Problem the first: Some of the songs are named things like 01-02- Wherever.mp3, the first part of which represents Disc 01, Song 02. But others are named things like -03- Gravel.mp3, which don’t include the disc number. Why not? They’re on the same album. They had better names when I uploaded them. So why didn’t Amazon use either the names I had named them or a consistent naming scheme? It looks like the names of songs that Amazon matched don’t include the disc number and the ones that actually uploaded do.

Equally surprising is that Amazon has chosen at least three different covers for the album. Most of the songs that it auto-matched have the correct cover. Most of the ones that I uploaded have ended up with the cover of Monkey Business by the Black Eyed Peas. Since I have no albums by the Black Eyed Peas, I have no idea how that happened. And one song, Letter to a John, has the cover of DiFranco’s Out of Range. Did the auto-match fail and give me the wrong song, or at least the wrong version of the song? While it’s not a song I know well, a little detective work suggests that although I uploaded the one from Living in Clip, the one Amazon decided to store does, in fact, come from Out of Range. No, they are not the same version of the song.

What else? Oh, for compilations, instead of setting up a separate Various Artists folder, Amazon puts each song on the compilation into a folder associated with the artist. For example, the songs on Badlands: A Tribute to Bruce Springsteen’s Nebraska appear in 13 different folders. Getting the compilations back together will be an adventure [7].

Badly designed software for uploading music. Badly designed software for downloading music. Bad file names for songs. Incorrect song matches. Maybe it’s good that Amazon discontinued the service. Perhaps I should not have counted on it in the first place.


[1] Officially The College at The University of Chicago. I’m pretty sure that some people pronounce it so that you can hear the capital T’s.

[2] That’s not quite true. I did join eMusic in the early all you can download days. Somewhere I have a bunch of CDs with those wonderful 128K mp3’s. If I recall correctly, eMusic had a bunch of Miles Davis as well as a lot of Rykodisc titles. And it was a good source for artists not on major labels.

[3] There’s a chance that I have an 800K floppy somewhere with the database on it. But it’s unlikely that a floppy that old still works. At some point, I could try with one of the Macs in the CS museum.

[4] My notes said September 11. But upload privs seemed to stop at 12:01 a.m. on September 11, rather than midnight. That strikes me as ending on September 10. Yet another thing to dislike about Amazon.

[5] Amazon claims that it will continue to store everything. It’s still safer to have a local copy.

[6] No, I did not look over every minute. I just glanced over every fifteen minutes or so. I suppose if I were smarter, I would have written a program that made a noise when the downloads stopped. I didn’t think of it at the time.

[7] Something in the back of my brain tells me that iTunes will help. I’ll have to see.


Version 1.0 of 2018-09-16.