Video: iterating towards the final design of my 'These go to five' banner logo

I recently launched a new blog, These go to five, which is attempting to (very slowly) chronicle all of the five-star songs that make up my iTunes music library. These songs represent only about 1% of my ~13,000 song music collection, so I thought I would use the blog to explain why they are so special to me.

Once I had settled on the name — a bit of an oblique reference to these go to eleven — I then started work on my design for a banner logo. The following video shows how I iterated over several designs before arriving at the final version. The backing music is an instrumental song that I wrote as part of the RPM music challenge.

The never-ending quest to rate and organize my iTunes library - part 2 of 2

In part1 of this blog post, I indicated that I had been rating the music in my iTunes library for about 8 years. Having thought about this some more, I think its probably been at least a decade. But after all of that time, I am finally finished! For the last four years I've kept track of how many songs remained in my 'To Rate' playlist, and so I can graphically show you how I finished this very long marathon.

Here is an explanation for the various milestones along the way to having no unrated songs remaining in my library.

a) January 2009 At this point I was only rating music that I had purchased (CDs, iTunes, and Amazon) and had just over 1,500 songs left.

b) August 2009 I tend to buy CDs in splurges rather than regular purchases every week or so. This jump probably represented about 10 new CDs. Most of the small spikes on this graph represent new music purchases.

c) February 2010 Finished! All of my purchased music had been rated...but what about all of the other music in my library which I had legally acquired for free. How much could there be?

d) February 2010 Well quite a lot as it turns out. Since 2005, the South by Southwest festival (SXSW) has provided free torrents of music by people performing at SXSW (one song per artist). If you didn't know, a lot of artists play at SXSW and these torrents are big. You can still get hold of all torrents from this site. The current running total is about 9,000 songs (45 GB!) and every year the torrents seem to get bigger.

These torrents are usually provided as two files, with the second one being smaller (and released a little bit after the first file). So this big spike represents me starting to rate 5 and a half years of SXSW music. Sidenote: a shout out to Danny Novo who had made some cool SXSW iTunes album artwork every year that the torrents have been available.

e) March 2010 I think this smaller spike represented the second of the two 2010 SXSW torrent files, with the majority of the SXSW music included in the previous huge spike.

f) and g) February & March 2011 The two SXSW torrents for 2011. You may note that initially I seem to rate the music very quickly and then it slows. This is because I tend to sort the yearly SXSW torrent by song length. Songs that are extremely short (< 1 minute) or extremely long (>7 minutes) tend to be music that I don't like and I can quickly process these. I have also found that songs that feature 'death' in the song title, or 'black' in the artist name are also not to my liking. Finally, I am not a fan of rap music so artists with 'MC' or 'DJ' in their name can also be quickly filtered.

h) March 2012 SXSW 2012.

i) August 2012 A bunch of new CD purchases.

j) January 2013 I think I had finished all of the SXSW music at this point, but then discovered a whole bunch of previously purchased music that had slipped through the cracks (pretty much of all of my classical music), plus I also had another bunch of new CDs to rate.

k) March 2013 SXSW 2013.

l) April 2013 Each year I get quicker at doing the initial pass of that year's SXSW torrent, and I think you can see how quickly I go through the long/short/black/death/rap songs (see steps f) and g) above).

m) August 2013 Finished. Done. The End. No more rating. right?

What now?

The SXSW torrents represent a real mixed bag of music. If you drove coast-to-coast and kept on randomly flipping through radio stations — preferably college radio stations, there is not much mainstream pop music at SXSW — it would get close to the experience of listening to a SXSW torrent. So unlike the music that I buy, where I have pre-selected albums based on one or more songs that I have heard and that I like, the SXSW music provides no such guarantees.

What tends to happen is that I discover a bunch of new artists through songs that I really love. However, a large swathe of SXSW music is either not to my taste, or is just a bit 'meh'. In fact, I realized that I had ended up with about 1,500 2-star rated songs from my SXSW torrents. While having some 2-star songs in the middle of a CD that own is one thing, these 2-star SXSW songs represent a potential waste of disk space.

As I am a iTunes Match user, I deleted all of these songs from my iTunes library (after ensuring that they had first been matched by iTunes). This frees up some disk space but means I can still play these songs and download them again if I really want them.

However, it still irks me a little to know that these 2-star songs are part of my library, even if they are not taking up any space. So because I am a masochist, and because I can't abide to delete them all in case there are some gems buried amongst them, I decided to re-rate all of these songs. If I decide that I really do like a song, I upgrade it to a 3-star rating. Otherwise I permanently delete it (removing it from iCloud). So far, I've processed about 200 songs and it will probably take me several more months before I finish this.

And then I am done.

Really done.

Until the next SXSW!

The never-ending quest to rate and organize my iTunes library - part 1 of 2

This is a blog post that I've been intending to write for many, many years. The final impetus for getting on with it was seeing David Sparks write about how he rates his music in iTunes. Before I reveal the minutiae of how I rate my own music — a system which is very similar to David's — I should give some background to my music collection and my long attempt at organizing it.

My music

I consider that I own a fair amount of music. I think that the new standard to be used when deciding if you own a lot of music, is whether you find iTunes Match's 25,000 song limit restrictive. Currently, I have 13,275 items in my iTunes library. These include:

  • 9,715 songs ripped from CDs I own
  • 2,209 songs from free SXSW downloads (more of this in part 2)
  • 840 songs purchased from iTunes
  • 121 songs purchased from Amazon MP3 store
  • 70 songs that were available as free downloads (predominantly from artists websites)
  • 35 self-recorded songs or musical sketches
  • 23 audiobooks
  • 22 podcasts (I mostly use Downcast to manage podcasts)
  • 2 E-books (I mostly buy Kindle books)

I can be so precise about this breakdown because I use a set of custom 'tags' for nearly all items in my iTunes library. These tags simply consist of adding text to the 'Comments' field of tracks; I use tags such as [CD], [ITUNES], [AMAZON] etc. The addition of square brackets makes it very unlikely that this text will exist anywhere else in a comments field. It also makes it easy to include as part of a Smart playlist. I'm curious whether the tagging feature of OS X Mavericks will allow tagging of songs in iTunes, which might mean I can replace these comments.

An alternative way of categorizing my music is to break it down by decade of recording. I try to change the dates on tracks to the original recording dates rather than the date that the CD/album was issued. This is mostly a problem for older music that only exists on iTunes from compilation albums or re-issues.

I think this represents a good spread across the decades (maybe Jazz fans would beg to differ). Another way of looking at this is that 11% of the music that I own was recorded before I was born (1971).

Rating songs

I've been using iTunes ever since I purchased my first Mac in 2001 (the era of rip, mix, and burn). I started ripping CDs from around this time, slowly working my way through my CD collection. At some point (maybe around 2005) I started to systematically rate my songs as follows:

1 Star - 2.8% of my library

Things I never want to hear. This sometimes include songs that I really don't enjoy on albums that I otherwise like (typical culprits being low-quality live tracks, or extremely long and unfocused instrumental jams). However, it mostly includes things like silent tracks that are used as interludes on some albums, or spoken tracks (think of all the classic radio DJ interludes on the Reservoir Dogs soundtrack). Giving all of these tracks 1 star means I can create smart playlists that will never randomly shuffle between music and spoken word (or silence).

2 Stars - 44.2% of my library

Songs that I am largely indifferent to. On a good day I might be in the mood for a 2 Star song, but on a bad day I might be offended by one. This category represents a large slice of my iTunes library, reflecting my opinion that there are many great songs which exist on largely mediocre albums (and I like to buy albums rather than individual songs). Perhaps my tendency to down-rate rather than up-rate songs is reflected by the fact that out of about 750 full-length albums, I have only 33 with an average rating greater than 3.5 stars (I have a Perl script that generates average ratings).

3 Stars - 44.2% of my library

Currently, I have almost an identical number of 2 star and 3 star songs. A 3 star song must have at least one great quality about it. This could be the lyrics, the melody, or even the performance of a single instrument. Often it is just a good hook in part of the song, or just the energy that the song imparts. I pretty much always enjoy listening to a 3 star song, but I wouldn't want to listen to the same song over and over again.

4 Stars - 7.9% of my library

There is a big drop in the number of songs that make it to this level. A 4 star song must be amazing. It must have at least two or three great qualities about it and I will almost never tire of hearing a 4 star song. The things that stop them becoming a 5 star song could be something minor, but something that might seep through into my consciousness when I listen to it (a song that outstays its welcome, a bad note in one part that can be jarring, or a single line with a flawed lyric).

5 Stars - 1.2% of my library

These songs represent songs that I find exceptional in almost every way. Usually I find them to have a stunning combination of lyrics and music, with fantastic vocal and instrumental performances. Furthermore I find these songs to have superb production values, with great arrangements and pristine sound (letting you hear details of every instrument on the record). In nearly all cases, I find these songs to be flawless. There is nothing about them that I could imagine being improved. It perhaps reflects my overall preferences for what I might broadly define as 'classic pop/rock' that 90% of these songs are less than 5 minutes long. My iTunes library does comprise genres such as Jazz, Classical, and Prog Rock, but they reflect a small proportion of the music that I own.

Finally, I should add that I will often listen to a song 10 times or more before making my mind up as to what the most suitable star rating should be. This, coupled with the tendency to keep on buying new music, is the major reason that it has taken me almost a decade to finish rating my entire iTunes library. My progress in getting there will be explored in part 2, along with a few comments on the challenges provided by the free SXSW downloads, and the question that now looms large on my horizon:

Once you have finished rating the entirety of your 13,275 song iTunes library, what do you do next?

The Littlest Birds Sing the Prettiest Songs — or do they?

According to this video testimony from musical performers The Be Good Tanyas, ‘the littlest birds sing the prettiest songs’. If we take their claim at face value, one might presuppose that ‘size of bird’ and ‘prettiness of song’ might be connected in some form of linear relationship where smaller birds sing prettier songs:

Littlest1However, as they fail to clarify the nature of the relationship between ‘size of bird’ and ‘prettiness of song’ one could postulate that the following examples also satisfy their description:

Littlest2


Littlest3


Thus we feel that — in the absence of supporting data — their observation that smaller sized avians produce more pleasing vocalizations, should not be regarded as a relationship which necessarily suggests that larger-sized birds sing notably less pretty songs.

Regardless of whether there is a relationship between these two metrics, we might still expect to find supporting evidence for the notion that diminutive avians produce trillings, warblings, and other chirruping-like sounds that are of a fetching nature.

One might therefore assume that the songs of Hummingbirds should be among the prettiest of all bird songs. However, according to the Hummingbird-themed site World of Hummingbirds:


“the sounds of hummingbirds are not exactly considered music…it would be more appropriate to call them chirps then it would be to call it music.”

This testimony, along with the audio evidence provided on their site, suggests that the claims of The Be Good Tanyas should not be taken at face value. Further undermining their claims is this list of the top ten song birds from ‘Bird On! News’. While some of the entries on this list also feature in whatbird.com’s list of small-sized birds (those birds 5–9 inches in length), the top 10 list also includes the mute swan which can grow up to 1.5 meters in length, a size described by Wikipedia as ‘impressive’.

Without published data in a peer-reviewed journal we find the evidence to support any claim that ‘the littlest birds sing the prettiest songs’ to be lacking and we humbly request that The Be Good Tanyas ‘show us the data’.