The Birth Of A Chord Melody

A couple of months ago, the regular newsletter from Jamie Andreas dropped into my mail-box. In it there was a short introduction to chord melody, in the form of Happy Birthday. As both my children (uh- children? 18 and 20?) had birthdays, how nice, I thought, to play them something instead of them being tortured with my “singing”. So, I sat down and started to learn how to play it. It so took my fancy that I decided that I had to follow this trail a little further, as and when a piece came along that would be a “suitable case for treatment”. I added “Chord Melody” to the (very long) list of “Things To-Do” and continued with my life.

Recently, whilst I was looking for a tab for something completely different, I sort of fell over the tab for Auld Lang Syne, which for some unknown reason sparked the words “Chord Melody” in my brain. Well, I thought, now that the idea is there, I really have to do something about it. The tune was simple, straightforward and not spread over several octaves, which in my simple mind made it an ideal candidate for experimentation. So without further intelligent thought (a not uncommon occurrence) I sat down at my PC and started into converting the song into my idea of chord melody. The thing is, I have not been playing a year yet (discounting the first attempt 40 years ago) and I have never tried to actively take part in the construction of a piece of music in any way, shape or form, other than changing the occasional note on a piece of tab. Perhaps I was biting off a bit more than I could chew, but I would never find out until I tried.

So I did.

As I had never tried anything like this and have read very little on the subject of chord melody, it was mostly a case of flying by the seat of my pants to start with. As I progressed, I found some sort of methodology, but the first moves were definitely “suck it and see”.

My main tools were Guitar Pro 4 and the Chord Guide.Com. Guitar pro was favoured for one or two features that I found in GP before I found them in Powertab (it started out in Powertab format, got transferred to GuitarPro and, eventually back into Powertab). The ChordGuide was an invaluable aid, especially in terms of time. I could have worked all the chords out myself, but it would have taken a lot of time and distracted me from the “main event” – the chord melody, itself.

The original score consists of 33 bars in two repeating patterns, which overlap at bar 17. That is, bars 1 to the first 3/4 of bar 17 are the same as the last quarter of bar 17 to bar 33. The reason for the quarters, is that the first and last bars share a full bar (there’s a name for this, but I can’t think what it is). The tune is in 4/4 and the first bar has one crotchet (quarter note), the last bar having the remaining three beats in the form of a dotted minim (half note + quarter note). The replication makes life easier, in that I only have to work with the first 17 bars, the rest is a repeat. These are the first 7 bars of the score in its original form:

You’ll see that it is in the key of F major (one flat, finishes with an F major chord). Rightly or wrongly, I discounted it being D minor, because of the last chord, which even allowing for a Tierce de Picardie, would have been D major.

Remembering the few lines that I had read on what chord melody is about and having taken a second look at Happy Birthday, the first thing that I had do was to get the basic tune up to the top 2 strings. This is so that the highest note of each chord played is the respective note from the melody. The original was in F, based on the 4th string 3rd fret, so the first step was to raise it an octave and everything was now in the right place (more or less, but more on that subject later), with only one note down on the G string.

These are the relevant bars after raising everything by one octave. As you can see, the second note of bar 17 is the same as the first on bar 1 (in standard notation on paper, the rests would not be included) – but, hang on, the tab comes out different. This is because both GuitarPro and Powertab try to make intelligent decisions based on the surrounding notes. This may or may not help in chord melody, as you can see clearly in bars 2 and 18.

My next decision was where to play the chords. Not having any background, I used what knowledge I do have of music theory to help me. 4/4 time consists of 3 emphases – strong (1st beat), middle (3rd beat) and weak (2nd & 4th beats). So I decided to put the chords into the tune on each of the stronger emphases. The first and third beats of each bar were, therefore, going to be juiced up with a chord, but what chord was that going to be? Again having no background to tell me otherwise, I decided the best way would be to simply merge the melody and rhythm lines. Just so you can follow along, let me give you what became my full transcription in F:

I went about it like this, I took the note that had to be played (let’s take the second bar as an example) and found a chord voicing that terminated with that note in the highest position. So, bar 2 starts with an F at the first fret of the first string, so I, therefore, need an F chord which uses string 1 fret 1 – easy – the normal F (4/3, 3/2, 2/1, 1/1) fits perfectly. The third note is also an F at the same position, so no problem there. Oh, man, am I on a roll here!

Bar 4. GP has transcribed the notes correctly, but the tab has been moved to the B string, presumably, to make the transition, in the next bar, to the 10th fret on the high E easier. Doing this, though, makes it more difficult to put in an F chord to terminate on the 6th fret of the B string, whereas we already have the F on the 1st fret of the high E in bar 2. This is one of those “more or less” cases I spoke about earlier – there is a certain amount of this tweaking to do. But no big deal, the roll picks up speed again!

Bar 8. Roll comes to a halt. After tweaking the tab, to bring the playing position onto the high E, the first chord was easy – a straightforward A-shaped Bb barre. However, I needed to look more closely at the 3rd and 4th beats.

I needed a C7 that ended on the 3rd fret of the 2nd string, which is not going to happen, because that’s a D and C7 has no D in it, but, as David Hodge kindly pointed out to me, a C9 does. The standard fingering of a C9 (032333) plays a G on the 1st string, but as this is the V of the chord, it can be left out in a 9 chord, which is what I did.

What started as:

became:

Having got over this hurdle, the remainder of the song was relatively straight forward. The major work being to ensure that the playing position of the notes was optimised for the chords that needed to be played and for the transitions to and from that chord.

However, playing the tune via MIDI made it clear that the song was imbalanced between the meaty chords, like G and the relatively thin open F, so you will find that a number of the “heavier” chords have had the note on the low E removed. Whilst this changes the voicing from a normal chord to a slash chord, where the lowest note is not the root, it does improve the overall balance of the piece.

Once I had a first version finished, I asked David Hodge if he would be kind enough to look it over and tell me if I actually had any idea of what I had done. To my amazement, he felt that I had done a good job, but wondered why I had not transposed into G, to make use of more open chords. Suddenly overwhelmed by enthusiasm, I immediately set about transposing up a whole tone.

So, back to the drawing board, but this time, with a little bit of background to help me through. Guitar Pro showed its value again, making the transposition of the original score (prior to chord-melodising, so to speak) really easy. I changed the key signature from F (1 flat) to G (1 sharp) and raised the entire track by 2 semitones. Done! Now for the chords!

Again, it would have taken much longer for me to have worked out the chords had I not been able to count on ChordGuide.com. As expected, the same problem arose in bar 8, as in the first time through, and I had to put in a D9 rather than the D7, which the original score used. However, one thing that did come out of the transition to G was that, without having to doctor any of the voicings, I ended up with a much better balanced song than had been the case in the key of F.

So, there we have it, my first attempt at chord melody, which will almost certainly be used to dampen raise everyone’s spirits on New Year’s Eve. I have certainly learnt from the exercise and will, no doubt, be trying other tunes as and when they come into my hands. I will, however, make the effort to learn more about chord melody before embarking upon another arrangement, if only to understand what it was that I did right and why (and, of course, anything that I could do better next time).

Final thought:

For anyone who has asked the question “Why learn music theory?”, the story this article tells is why. I could never have done it without a reasonable knowledge of theory. Tools like GuitarPro and ChordGuide.com are wonderful time-savers, but do not help you to think around a problem.

Here is the final result of my labours: Auld Lang Syne, a chord melody in G.