The first 12 notes of a song

If you are reading and come across a word that you do not understand, and you feel suitably motivated, it is straightforward to look it up in a dictionary. This is something that I do a lot, as part of my motivation for reading is to improve my own communication skills. Yet another reason why my Kindle is my friend. It is less easy if you just hear a word, as, in English, it can be challenging to find the right spelling, particularly when the speaker is lazy about pronunciation.

The problem gets more interesting when, instead of a word, it is a piece of music that you want to identify …

Nowadays, technology gives us some interesting ways to identify music. If the song you hear has words, then all you need to do is Google a suitably long sequence and you will most likely find yourself directed to one of the numerous song lyric sites. On one occasion, I mis-heard the words and Googled what I thought that I had heard. The lyrics site to which I was taken was so smartly designed that it cataloged mis-heard variants of songs as well as the accurate words.

If you have a smart phone, you just need the right app. I use Shazam, but I am sure there are others. All you do is let the app “listen” to the music for a few moments, then it accesses some remote database and comes back with the answer. I am constantly amazed at its precision. It seems to be able to identify quite obscure music, from only a short sample and seems impervious to quite high levels of background noise. I guess this is a high tech implementation of The Old Greys Whistle Test.

Many years ago, I heard about a much lower technology approach. A dictionary was published, which listed many thousands of pieces of music. The authors claimed that, if you could hum the first few [I think it may have been 12] notes of a song, you could look it up in this dictionary. All you needed to do was right down these notes in the following way:

  • The first note is written as “X”.
  • Each subsequent note is written thus:
    • “S” means that it is the same as the previous note.
    • “H” means that it is a higher pitch than the previous note.
    • “L” means that it is a lower pitch than the previous note.

So you end up with a string that looks like this: XSSLLSSLHSS…

Then you look through the dictionary where songs are listed as strings like this in alphabetical order.

I was very taken with the elegance of this idea and have often wondered whether that dictionary was successful and whether it still exists.

Leave a Reply

This article first appeared on the Siemens Digital Industries Software blog at https://blogs.sw.siemens.com/embedded-software/2012/02/16/the-first-12-notes-of-a-song/