Your digital library statistics

edited March 2013 in Diversions
Just a one-off fun-off: List your digital music library statistics here (you can add to mine if you find some other data interesting):

Tracks: 45k
Size: 300GB
Median track length: 4:14
Median year: 2000
Median track plays: 4
% of library with no plays: 0.86%

Those are the easy data to acquire from iTunes. Anyone else?

(Median is acquired by sorting and finding the value halfway through the list, in case you weren't familiar.)

(I thought we'd done this before, but I can't find the thread. Anyone else recall the specific one?)


  • Tracks = 5,686
    Size = 47.93gb

    Not sure how you determine where the halfway point is to determine the median.

    That No Plays stat isn't accurate on my iTunes. I'm seeing albums that show zero under no plays, yet I've reviewed the album, so there's no way that can be right.

    I keep my iTunes library pretty clean. I delete regularly. Some stuff I'll burn to a CD if I think there's a good chance I'll listen to it later. Otherwise, it's gone if I can't even remember what the hell it sounds like and it's been sitting there for more than a year. And with my review schedule what it has been the last year, if I haven't reviewed it after a year (and don't feel a sense of disappointment at realizing it), then I'll never get to it and it's just clutter.

    Fun thread idea, though. I'll be interested to see others.
  • Tracks: 52,337
    Size: 916.38 GB
    Days: 142.1

    Of the total number of tracks, 29,853 are Apple Lossless (ALAC) and the rest are MP3/AAC. Interestingly, the lossless tracks account for 57% of my library but take up 81% (742.96 GB) of the space.

    Play counts are unreliable because I've transferred my library multiple times and re-ripped all my CDs to lossless a couple of years ago. According to iTunes stats, 36,033 of the tracks in my library are unplayed. That ain't right.
  • edited March 2013
    Not sure how you determine where the halfway point is to determine the median.
    As a decent estimate, you simply sort the full library on the measured value (time, year, plays, etc), and drag the scrollbar to the halfway point or so (or if you have a numbered playlist, you can choose the exact halfway point). Median values are generally better representatives of "average individuals" in a distribution (they ignore the weights of outliers).

    E.g. for the following list: 4 is the median value
  • Tracks: 21,302 (In iTunes. The actual number is much higher, as you might expect from a member of this board.)
    Size: 144.64 GB
    Median Track Length: 3:58
    Median Track Plays: 4
    % of library with no plays: 1.08%

  • @Karg

    Can't believe how burned out I really must be to not have intuitively thought to use the scroll bar to measure the half way point.

    Median song length = 5.08
    Median listens = 2

    I can really see the difference between the albums in my iTunes for reviewing vs. the ones that I just enjoy. The number jumps from single digits up into the twenties quite suddenly.

  • jonah:
    Median song length = 5.08
    Heh, that's quite high - it's obvious how much jazz probably dominates your library. :)
  • I'm hoping for some more median-year stats - that strikes me as a potentially interesting axis...
  • I guess one more interesting stat - my iTunes library is intact from Nov 2003 (my oldest last-played is 11/18/2003), working in a 10 yr anniversary. I'm obsessive about listening history/playcounts (many of my playlists are based on selecting stuff I haven't heard in a while), so I've tried to be careful.
  • I'll work out my stats later, but one small point. The median year will be skewed by the way that emusic dates albums. For example, all the latest ECM drop in Europe have gone down as 2013, even though some are actually 20+ years old
  • Well, now, Greg, I think your median year will be skewed by your not correcting your tags! :)
  • I know that!! I admit I do not often retag, other than category - Reggae for West African music, eg, annoys me

    Median Year 2004

    Plays 1!! The range is 33 to 0, with about one third unplayed, since the counter was reset when I changed computer. Unlike Jonnah I never delete music.

    25541 Items taking 79.9 days to play

    I haven't found Size yet

    Items is perhaps out by a couple of thousand or so if I ignore duplicates. I am in the process of adding all my CDs as lossless files, but retaining the much smaller files as my ipod only holds 30GB. But to compensate I have a number of older CDs still to put on itumes.
  • Another: % library heard in the last N years (rounding) -
    1 y: 25%
    2 y: 50%
    3 y: 70%
    4 y: 80%
    5 y: 90%
  • 27,541 tracks, 98.9 days, 225.92 GB. That's a little skewed, as some percentage of those tracks are "unchecked" and not likely to be listened to much, if at all...a lot of old Amie stuff.

    Median year is 1979, reflecting all that old jazz. (Track #13,766 is "Passenger" by the Grateful Dead - a show from the Live Music Archive, and good example of an album I don't listen to much.)

    Median play count = 5, median time = 4:10. 434 items unplayed...that's high, reflecting lots of recent pop additions from Guvera. Also just "checked" items. Because of smart playlists, I don't typically have a lot of unplayed songs. The song "Late This Morning" by the Freddy Jones Band has been unplayed since I first added it on November 1, 2005!

    Now my other library has about 3,600 items, another 36 GB! That's the office computer, mostly classical...probably also a good bit of duplication with the main music computer. This exercise makes clear that I'm not very organized in terms of deleting stuff I don't listen to.
  • The current iTunes library contains:

    19,974 tracks, 96 days, 307.2 GB.

    Median year: 2001 - the media track being "Spectral Mornings" by Steve Hackett (recorded 1979, released 2001).
    Median play count: 1 - but library was rebuilt last year following a disc failure...
    Median track length: 4:54 (some 82 tracks are over an hour)
  • edited March 2013
    Median year is 1979, reflecting all that old jazz.
    Wow, that is a lot of old jazz. Now I'm tempted to ask you what your numbers are by decade. I guess I'll see what mine are:
    20s: 89 (0.2%)
    30s: 238 (0.6%)
    40s: 482 (1.1%)
    50s: 1264 (3.0%)
    60s: 2313 (5.5%)
    70s: 2687 (6.4%)
    80s: 4205 (10%)
    90s: 9560 (23%)
    00s: 17189 (41%)
    10s: 4101

    So far, monotonic! The 2000s are probably boosted by lack-of-date-correction, but I don't think too much.
  • I'm about 100 CDs away from getting my full classical collection onto a hard drive, and the stats so far are:

    Tracks: 54,813
    Size: 303 GB
    Total time: 4,229 hours

    Median track length: 3:36

    Median year of tracks that have a year: 2000

    Median plays: 0. Not surprising given that most of the music has been added in the past year.
  • Now I'm tempted to ask you what your numbers are by decade.

    20s - 0.5%
    30s - 2.0%
    40s - 3.0%
    50s - 7.9%
    60s - 16.4%
    70s - 20.6%
    80s - 9.2%
    90s - 9.1%
    00s - 18.1%
    10s - 12.6%

    So there you go - Thanks largely to Guvera, I have built a pretty extensive library of mid-50s to early-70s post-bop. The upswing in the 00s of course reflects first iPod/iTunes in 2005...
  • edited April 2013
    Finally back at my computer.

    Tracks: 16811 - 68.2 days
    Albums: 2414
    Size: 164GB
    Median track length: 4:59
    Median year: 2006
    Median track plays: 2
    % of library with no plays: 3% (but see below)

    20s - 2 (0.01%)
    30s - 2 (0.01%)
    40s - 4 (0.02%)
    50s - 285 (1.7%)
    60s - 233 (1.4%)
    70s - 319 (1.9%)
    80s - 566 (3.3%)
    90s - 1483 (8.8%)
    00s - 8526 (50.7%)
    10s - 4897 (29.1%)

    Genres over 10%:
    Ambient 3159 (18.8%) - 15.4 days (22.6% by time)
    Electronic 4737 (28.2%) - 18.3 days (26.8% by time)
    Jazz 2182 (13%) - 8.7 days (12.7% by time)

    Most listened to albums:
    Brokeback - Field Recordings from the Cook County Water Table
    The Gentleman Losers - Rural Route No. 5
    Tetsu Inoue - Inland
    Nils Frahm - Felt

    - The decades distribution reflects the price of music in tandem with my shifting tastes. In the 00s and 10s my growing interest in experimental electronic and ambient music is symbiotic with with music becoming cheap (emusic, Amie) or free (netlabels, Guvera). Before that I had a small but beloved record collection, much of it in other genres. Guvera is denting the distribution a little with older jazz, but even there I've downloaded a lot of recent ECM.
    - I owned almost no jazz before the beginning of 2012. Guvera is largely responsible for the surge.
    - Those three top genres do also reflect my listening, though ambient would be a higher percentage of the listening in terms of time (a lot of longer tracks, as evident in the percentages by time under genres).
    - I included # of albums because I was curious what the average tracks per album would be. It's 7. Again that's showing the effect of a lot of one- or two-track ambient albums.
    - The % with no plays is really lower than that - the tracks found in the library that have no plays include stuff bought for other family members that I might never listen to. Should clean that up some day. I do have a playlist that uses some criteria to filter out those unplayed tracks that need my attention - it currently has 189 tracks, or 1.1% (I've almost caught up with Guvera!)
    - I actually have over a thousand classical tracks but don't really listen to them much at all except for some contemporary things and Arvo P
  • The recent thread "Pruning your hard drive" reminded me of this old thread.

    A couple factoid additions (not bothering updating above stats): 

    How fast do you "cycle through" your library? One fast measure is the median last-played date for all your played tracks (ignore unplayed). Of my current 63k tracks, my median last-played is about 2 years, which seems pretty good to me, in terms of bubbling older stuff up in a reliable way.

    The number of tracks I haven't heard in >9 years is 47. (Choose some illuminating time of your choice - I have that as a smart criterion to chip away at those "ancient" listens).
  • Maybe the answers are buried somewhere in this thread (and thanks for dredging it back up...I expect there are updates to be had in the last seven years), but I'm curious about the shortcuts that yield these delightful statistics short of doing a ton of gruntwork w/ smart playlists in iTunes/Music.  Median #s especially I'm curious about methodology:  are you just eyeballing the middle of the list?
  • omnifoo, I just noticed I never responded to you. Median I get from copying lib contents (if applicable) into a playlist, which is always numbered. Then just sort and go to mid-way by track count.
  • Btw, I still like this idea, and this thread is old - will be interesting to see what's changed when I have a chance.
  • Good to know there's always a system w/ labor saving methods.  I didn't use Apple Music enough for my 2021 "Replay" to be interesting.  I still prefer using the library of local files most of the time, but still have played nothing more than 16 times in the past three years.  I'd use a median-finding method like yours if my stats were more interesting.
  • edited February 2022
    9y later update, no surprises really:

    Tracks: 66k
    Size: 600GB
    Median track length: 4:25
    Median year: 2002
    Median track plays: 5
    % of library with no plays: 0.6%

    Listened to 55% of the lib in the last 2 years, keeping steady pace there.
  • My 22 year old monitor went blank three weeks ago,
    so I can’t access my stats, but I’m at somewhere around
    1.3 million tracks. The other info will come next week once
    my new monitor arrives.
  • That's literally 20x the size of my or likely anyone else's lib here. :-P Have you managed to actually tag it all correctly??
  • Well, I’m not sure of what your definition of “correctly” is,
    but if track, artist, album, genre, year, etc. being filled in
    is what that refers to, then yes, that’s what every one of
    those have. It’s really not difficult because it’s mostly
    done for me thru the CDDB. I may make an adjustment
    here and there to suit a desire (composer instead of
    performer listed, for example). I also have software that
    adjusts the volume of each track with that automatically
    listing date and the time it was adjusted within each track.
    About once a quarter, I’ll run software that finds duplicates
    and removes them, but pretty much for over twenty years,
    iTunes has kept track of all that stuff concerning number of
    times and when played without any intervention from me.
  • edited February 2022
    March 2013:
    Tracks: 52,337
    Size: 916.38 GB
    Days: 142.1

    February 2022:
    Tracks: 74,113
    Size: 1.50 TB
    Days: 200.4

    All of my music (except vinyl) is on an external NAS drive that I can stream through my home stereo or on any Bluetooth speaker attached to my home network, which is awesome. Play counts on iTunes are useless because streaming plays don't register. According to iTunes, 2/3 of my library tracks have never been played. I'm sure the actual number is pretty big, from the days when I was downloading albums by the dozens from eMusic (roughly 8,500 tracks), Amie Street (5,800), Guvera (2,700), etc. But not that big.

  • edited February 2022

    February 2022 (compared to April 2013)

    Tracks: 53111 – 206.6 days (up 315%).
    Albums: 5637 (Up 234%) [compared to the stat above, this suggests I have bought more albums with higher track counts].
    Size: 568.35GB (up 346%) [but size relates to file formats..].
    Median length: 4:42 (down 17 seconds) [shrug. Larger file count makes the epics have less effect]
    Median Year: 2012 (6 years later) [shrug].
    Median play count: 2 [see below].
    Highest play count: 328, for Hear the Sun by Pete Namlook & Tetsu Inoue. (Notable outlier: 56 plays of Remain by Stephan Mathieu, an hour-long track)
    Played in last 2 years: 16765 (32%) (no 2013 data).
    % with no plays: 0.001% (51 tracks queued up in my new acquisitions folder) (down 3%).

    Genres over 10% of total:
    Ambient 12042 (23%) (up 4.2%).
    Jazz 11541 (22%) (Up 9%!!)
    Classical: 10031 (19%)  (did not appear). [up a lot but this number feels fake, accurate as it is - see below]
    Electronic 6249 (12%) (down 16%!!)
    (Honorable mention: Prog 4040 (8%) (up quite a bit I suspect).
    (Doing this by total time instead of track count boosts ambient by 5%, and prog by 1%)

    Distorting factors:
    - cheap/free classical buckets-o-stuff, like the 99c Bach Guild boxes - inflates the classical number (I certainly don't listen to classical 19% of the time - maybe 5%) and deflates the median play count.
    - cheap entire-label all-you-can-eat deals at Bandcamp where I have a few times bought, say, 140 albums at once and listened to most of them once sifting for the gems - also deflates the median play count.
    - CDs. The music I like most I buy on CD and listen to on the good speakers - many of my most listened to albums don't make a dent in the MP3 play counts.
    - genre economy effects - there are many more ambient labels giving stuff away for free than prog labels.

    I have bought two thirds of my music collection since 2013.

  • As of Feb. 2022

    Tracks 73,112

    Size 521.71GB (small likely due to lots of old, smaller mp3s.  Clearly I'm no audiophile.)

    Days 224

    Artists 4742

    Albums 6945

    I’m still behind the curve on next-gen stats but am sure I haven’t played most of it in the three years since starting from scratch on playcounts in 2019 when I switched to a Macbook.  I’d like to think the “Unplayed” smart playlists since the 2010s would have played the majority in the past decade.  Max plays = 16 (as w/ Apple Music Replay, in preparation for podcast episodes on particular albums).

    My genre & year tags are all out of sorts, unhelpful.

    Your stats are fascinating for teasing out our different listening habits & tastes.  Thanks for sharing!

Sign In or Register to comment.