How to tell the diamonds from the rough

Every time I come up with a melody that I like, whether it is while I'm plucking at keys, fooling around on my computer, humming, or simply thinking it up, I immediately believe I've come up with the best melody I've ever come up with. I think that's natural. If it doesn't seem great at first thought, why linger on it at all?

Sometimes it only takes a few minutes for me to realize that, well, maybe it isn't the greatest melody, or even really one worth keeping. Sometimes it takes a few days, or it takes coming up with a newer, better one that puts the first one in its place. Unfortunately, sometimes the melody makes it all the way into a full-blown song before I realize it kinda stinks. Such is the songwriter's struggle.

This last situation is one that every songwriter tries to avoid: writing a song that he or she ends up regretting. It feels like a waste of time and effort, and it can be kind of demoralizing. So are there ways to test the ideas right from the start, of separating the diamonds from the rough, of making sure that shiny stuff isn't fool's gold?

Well, there are some techniques that can be applied, though none are perfect and there are potential pitfalls to some of them. In this post I will descibe a few I have used.

The first is what I call the memory game. It was described to me by an old composer friend. In its raw form, this is a bit dangerous and may lose the melody for you completely, but it's a pretty good test to see how memorable the melody is. It is also incredibly simple. You do not record the melody or write it down or anything else that would serve to ingrain it more deeply in your mind than simply hearing it a few times would do. Just play it or sing it to yourself a bit and let it sit in your head. Then walk away. Listen to other music. Play video games. Make yourself put it totally out of your head. Do this for hours or even days. Then try to remember it. If you can remember it, well, that's a pretty darn memorable melody. If not, it wasn't particularly memorable, right?

The problem I've found with this method is obvious: you end up forgetting a lot of melodies. It is also not entirely clear that it weeds out all the bad ones or that all the ones it weeds out are bad. Having tried it myself, I can say that some non-stellar melodies have made it through, and I'm pretty sure I've lost some good ones.

A slightly different method is what I'll call the modified memory game, which adds the obvious step of recording the melody first. Yes, this may increase your chance of remembering a melody that isn't particularly good, but it's worth it, I think. The key is to record it simply and quickly. Sing it into a microphone or play it on a keyboard, record it, and stash it. An additional advantage here is you can see if the way you remember it is the way you recorded it. Sometimes, the mind plays games of its own and displaces a few notes over time.

An unrelated method for testing a melody is, of course, showing it to other people. I don't much like this, however, as you relly need to pick the right people. Unless you are intending it as some sort of a capella piece or solo line, you are effectively showing people an unfinished work. Only a few kinds of people can give good feedback on unfinished work. You don't want to show unfinished music to people who aren't used to hearing it. They will be judging your music against what they are used to hearing, which is completed, fully produced songs. You also don't want to look to loved ones who don't want to hurt your feelings. Find other songwriters, people who work on audio or produce albums, or anyone who is used to hearing things that are less than ideal. And people who won't be afraid to tell you that you stink. Never trust the opinion of a person who likes everything.

Now, I'm talking about melodies here, but melodies don't exist in a vacuum. I know that, for me, it is relatively rare that I just come up with a melody with no accompaniment ideas. So if your melody needs to have some context for it to make sense, make sure you include that context. If there is a counter-melody, include that. But don't go and produce a full song until you're sure it's built on a solid base of a good melody (assuming your song is meant to be a melodic one and not an experiment in anti-songwriting or something).

Probably the best way to tell if a melody is good is simply to give it time. If it still seems like a good, useful melody to you in a few months, it probably is. Of course, not everyone likes to wait a few months to see whether they still like an idea before using it. Certainly, if you're one of those people who write ten songs a day, this is a pretty useless method.

Sometimes maybe it is just best to not judge and just go with it. Write tons of music without pausing to see whether it's good or not. However, people who write a lot of music tend to write a lot of bad music. It is a rare genius who can crank out whole albums of good stuff in a week or two. Most of us have to rely, at least on some level, on craft to complement instinct. When you are writing quickly, it can be exhilerating, but you do sacrifice some amount of quality control. Everything becomes more hit-or-miss. Putting time in allows you to steer songs that might have gone off track into something better than they would otherwise have been. Of course, overthinking has its own dangers. But that's for another post.