So, I was curious what statistical behavior I have in regards to keys and scales.
What do I tend to write the most?
So, I grabbed a pile of my music (about 40 or so songs) and ran them all through an analyzer (because I’m lazy), and then counted up the amount of each scale that I produced from my music.
Firstly, what I find interesting is the entire lack of F from my toolbox. If I reach for F, I seem to tend to do it through Gbm rather than F. Huh. Interesting.
The other thing that I found rather surprising was that my choice in scales is almost a perfect mirror shape, though I favor Cm a bit more than Eb, so the mirror is slightly leaned to the left.
I was actually quite surprised by the diversity in my preferences.
I actually expected a lot more Am and Em leaning. If it would have come back something like Am at 60%, Em at 25%, and the the rest scattered a bout a bit, I would have thought, “Yeah, that’s about right.”
I hadn’t realized that I was leaning that much on Gbm, or Cm.
Curious about yourself?
Grab this piece of software (http://www.ibrahimshaath.co.uk/keyfinder/) (Windows uses the older version; guess they stopped updating windows in the most recent version), if you’re lazy like I am, and load a batch of songs and click Run Batch Analyze.
There’s no way to export the output, so if you want to count how many times, etc…, you have to manually add that information into excel yourself.
The way that I did this was that I just typed each of the returns into a separate cell, and then next to each cell I simply added a 1.
Like so…
|Abm|1|
|Abm|1|
|Am|1|
|Am|1|
|Am|1|
|Am|1|
|Am|1|
|Am|1|
|Bb|1|
|Bm|1|
|Bm|1|
|Bm|1|
|Bm|1|
|Bm|1|
|Cm|1|
|Cm|1|
|Cm|1|
etc…
This way you can sum the right column by the key index on the left.
This will give you…
|Abm|2|
|Am|6|
|Bb|1|
|Bm|5|
|Cm|3|
Then it’s just a matter of summing that right column and then dividing each key’s count (e.g. 2 for Abm) by that total to find your percent.
Cheers,
Jayson