Need Help with Music Composition

[James--]

I was wondering if anyone here has any advice on how to compose music for adventure games, because I'm having trouble coming up with tunes.  If I look at the general pattern of another piece I can then go ahead and create something similar to that, but I want to create original pieces...

I've already got a great program for it, so that's not a problem... I'm just having trouble coming up with 'catchy' tunes.  I've done grade 5 piano and theory exams, so I understand a lot of the techniques, I just have trouble putting them into use really.

Another problem I have is with instrument selection.  I've got hundreds of instruments to choose from, but I have no idea how most of them sound.  Usually I write tunes as just piano first, and then try to choose instruments afterwards, but it's just a long process of trial and error.

So... how do you go about composing music for your games?  Do you come up with tunes in your head or do you have a technique for coming up with them?  And how do you choose which instruments to use for each track?

[thomasregin]

Hey James,

Hmm.. What you're asking here is not really related to adventure games as such, because what you're after are some general guidelines on how to compose music.

And those are _very_ difficult to give, because each composer has his own way of composing. Some like to meditate before writing, others take drugs. But I can tell you how I personally compose for movies/games/etc..:

First of all I have a talk with the director/producer about the project. It's important to know what he/she has in mind for the soundtrack so you don't start off with ultra-mega-death-techno when in fact the director was looking for religious church Hymns. After this, I need to have something to look at, whether it's 2 minutes of rough film-cuts or an intro to a computer game. Just to get an idea of what I'm dealing with. Having seen these rough cuts, I start composing in my head, while keeping in mind what the director wants. As I get more rough cuts or semi-completed parts, I start playing a long using a regular piano-sound until I find something that does what I want it to do.

Obviously I have a set of "tools" to choose from, seeing that I've been composing music for 15 years, but I always start from scratch and work my way through the soundtrack, one piece at a time.

The "tools" I'm talking about are "experience tools". For instance, I know that a solo violin is good to express sadness, and a crash cymbal does not do well as a lead instrument.

Apart from experience, there's one more important thing to remember (even though many tend to forget): Everything is team work - even if you sit alone in your basement composing, you need to be in regular contact with the director/producer. Send them loads of demos, so they can choose which style they'd like you to continue with. This way you'll show that you're really into the project and you'll keep the director happy, because he feels he's getting value for his money (which is very good for his stress-levels too).

For instance, I just finished writing the soundtrack for Dave Gilbert's Blackwell Unbound. During this process, I must have sent him at least 35 demos. And we ended up using around 18 for the game and that's, personally, how I like to do it! Keeps me calm too, to know that we have a wide selection of cues to choose from.

About instrumentation. Well.. Only one thing to say to that: Experience. Where do you start if you want to dissemble your car? I haven't got a clue, but my mechanic will surely know, because he's tried it so many times before. And that's the process you're in right now. You're gaining valuable experience by doing the trial and error thing. One day you'll know exactly what will work and what won't.

Well.. That was a bit of advice from me! I'm no expert and other composers may have a completely different working process, but this works for me! :-)

All the best and happy composing,

Tom

[m0ds]

Hi there. It's true we can't really help put catchy themes in your head because those sorts of inspiration need to come from yourself. Surely you find yourself whistling memorable tunes and stuff and sometimes even find yourself whistling or humming something you don't recognise. Well from there what I would do is just hum it again and think about other sounds and instruments to go around it. If you're able to, just try and imagine to things in your head. A melody and a bassline. From there you can create just about anything simple But try & let the bassline fall naturally as you hum or whistle a melody.

It's hard to describe, and it's not exactly a proven method - but it might help you a bit :p

Bit of an odd process I agree but you really have to let your imagination flow. There was a time when I was confident I could write anything down and it would be accepted as catchy. But that really isn't the case anymore. Nowadays it can take me months before I find a catchy tune I want to proceed with for a game. However, these days if I produce something catchy it's luck, because I don't set out with 'catchy' in mind, I set out with 'suitable'. From there you can learn & develop great things. Take Nelly Cootalot for example. I had some early ideas and I tried to think of ways to make them work but it just wasn't happening. It might just be me but there is a point when I know I've got something decent to work with and when I haven't. And for Nelly it was the latter for several months.

Tunes, ideas etc have always popped into my head but mostly I let them develop themselves. For these things to work for me they shouldn't be rushed which is why personally, with the right project - I now find it essential to get on board early & have a lot of time just for mental brainstorming. Case in point is Yahtzee's latest game which had me stumped for weeks and weeks yet again, as did 6DAS and Trilby's Notes. But through thinking some things out other things become clearer and that's all part of the fun. Through Yahtzee's new game I found the perfect melody for another game.

But going back to Nelly, sometimes that catchy tune you're after will just pop into your head. I think I was just walking home one day and had the title on my mind and just found myself humming something kinda cool. I hadn't even been thinking too hard. It just slipped in there. But once I'd whistled the melody and allowed the bassline to drop itself into my mind, I knew what I had was an everlasting melody for Nelly Cootalot. One of those amazing mysteries of life IT WAS LUCK REALLY. Well, for me it was luck, because I'd never felt prouder, and confident that this melody would hold up an entire game... Whether it did or not is other people's judgement, but for a theme tune I really like it. It's my favourite theme of all the ones I've ever composed because it seemed to say everything about her & the game I wanted it to say. But there is a lot of other key things to the motivation of that entire score which doesn't just fall back on me randomly humming a tune. But those are all personal elements that you'll discover in years to come "Warts & premature ejaculation, yeah, I heard"

(final version of the theme)
» http://www.screen7.co.uk/music/midi/scores/nelly_01maintheme.mid
(ending sequences)
» http://www.screen7.co.uk/music/midi/scores/nelly_21happyending.mid

You need a certain passion too. Try and tell a story through your music. Don't just create a tune because it sounds good, create it because it says something about what's going on. How you describe whats going on through music is up to you though, and it's something you've got to discover for yourself. That's why I find passion to be one of the most valuable aspects of music making for games. Choose a project you want to be involved with, one you know you're going to compliment because the story really intrigues you - and you'll probably end up thinking something catchy, moving or memorable or what not for it.

Thomas' advice is good, but it's unlikely you're going to find yourself working on a large commercial venture at this stage. I've been here 7 years and only done 1 so far. So with every other project the restrictions are usually fairly minimal, giving you creative freedom to design music you like. You may be given basic prompters, ie dark, suspense, fast paced, slow and creepy, all that kinda stuff. From there just start thinking about music to suit. I've been lucky with the fact that I've been given basic prompters through most of my works here so I've been able to develop my own style. But you may find that some developers have a really, really strong sense of what they want the music to sound like & something which can in some cases be quite tricky (well, it is for me).

There are lots of conventions and methods in music that even I don't understand yet but there are also some basic ones that just about anyone can follow, like ways to make the player feel they are moving, to make them feel in suspense etc. I was going to write these up with examples but it's not really what you were asking.

Anyway. You shouldn't even be asking this. You're Grade 5, so get thinking! :p

good luck

[Baron]

Me, I don't know anything about music theory.  The original music for my latest game -and granted there wasn't a lot of it -but that which was original was composed in the following way: 

1) I'd click and drag notes onto the musical staff at random (gasp!).  I'd use smaller notes if I needed a faster tune, larger ones if it had to be slower, but the pitch was all just random.  I'd also leave my program in the default piano mode unless I had a particular sound I was going for.

2) Then I'd listen to the "composition".  Ninety percent of the time it's crap, but since I hadn't invested a lot I'd just delete and try again.  It's usually not a long process before there's a bit of it that sounds kind of neat, and then I try to fill in the rest by ear (a tedious process for me, since I don't know off hand which notes have which pitch -usually I'd be thinking the next note must be "high-ish", etc.).

3) Once I was happy with the tune I would shop around for a suitable instrument, sometimes adding a baseline or a second instrument if it seemed appropriate.

Now I'm not saying this method produced super great tunes, but they sure were original!  There was also a strong tendency for them not to be anything like what I needed.  But I set myself the task of coming up with an original track every night, and after a month or so I had quite a bit to choose from.  Here are some ten second examples:

Discord in G*
Tribal Rhythm
Foxtrot Theme

*I don't know what the key of G is, the name just sounded cool.

Of the examples given I've only used the last one in a game, and it was dressed up quite a bit by another musician, but the others have potential, I think.  Anyway, just wanted to share an alternative method of music composition.

[James--]

Thanks a lot for the replies I really appreciate it

About instrumentation. Well.. Only one thing to say to that: Experience.

Yeah I suppose you're right - I've been looking through some midis from LucasArts adventure games to see what instruments they use, and it's already given me a much wider range of instruments I'm farmiliar with and can use myself now

(final version of the theme)
» http://www.screen7.co.uk/music/midi/scores/nelly_01maintheme.mid
(ending sequences)
» http://www.screen7.co.uk/music/midi/scores/nelly_21happyending.mid

Those are really good!  Definitely a catchy tune, it works great as a main theme

There are lots of conventions and methods in music that even I don't understand yet but there are also some basic ones that just about anyone can follow, like ways to make the player feel they are moving, to make them feel in suspense etc. I was going to write these up with examples but it's not really what you were asking.

That would actually be really useful actually - just the sort of thing I'm looking for.  If could write up a few of those that would be great

Anyway. You shouldn't even be asking this. You're Grade 5, so get thinking! :p

The extent of composition in grade 5 theory is completing an eight bar melody with a given two bar opening I think I'm going to re-read some of my theory books though, because I've forgotten a lot of the stuff and it's probably more useful than I thought

Discord in G*
Tribal Rhythm
Foxtrot Theme

Those are pretty good for someone with no theory.  The problem with using random notes though is that it will be hard to match the melody with backing chords from another instrument, as they have to sort of match the notes in the melody, which won't really work if they're random.  For relatively simple tunes with just one or two instruments that could work, but I'm trying to create more complex tunes


A method I have used a couple of times for composition is picking a key signature, then starting with the tonic triad (e.g. for C major the notes C, E and G), and then using just the notes in that chord for the other instruments (except for the melody which goes out of it occassionally), and then progressing the chord up or down one every few bars.  I think it's an ok method - pretty easy to do, but doesn't create a very interesting melody.


Thanks again
James