How do you write a song?

[Mankind]

So. This is what I have at my disposal:

A trombone.
A crappy keyboard.
Myself.
A microphone.
And a sound card to plug it into.

But I really have no idea how to combine all of those elements to create an original song.

I'm willing to listen to any advice people have to offer.

[erik]

A crappy keyboard.

How crappy? Does it have MIDI? Does it have a line out?

What are the specs on your computer with regards to speed and memory?

[Mankind]

It has MIDI and a line out.

Computer performance is not something I have to worry about.

I'm more concerned about making music than operating hardware/software.

[erik]

It's possible to operate software/hardware in <i>order</i> to help you make music, and there are programs out there on all different levels of difficulty. So, is your question more "How do I write a song?" or "How do I end up with a recording of a song?"

[Mankind]

How do I write a song. Like, come up with musical ideas and so forth.

[erik]

How do I write a song. Like, come up with musical ideas and so forth.

What kind of song do you want to make? Rank these things in order of coolness, from coolest to least cool: Vocal Melody, Lyrics, Musical Riffage, Arrangement.

[Dan-O from Five-O]

Also, what's your musical background? Can you play any of these instruments?

[Mankind]

Arragement, Musical Riffage, Vocal Melody, Lyrics


As far as musical background goes, I've been playing trombone off and on since high school, so, something like 15 years, and I know how to play chords on the keyboard, but can't do anything that sounds very pianistic.

[roymond]

but can't do anything that sounds very pianistic.

Then definately don't go for the switched-on-Liszt style.

-Just put down some sounds that progress some how (either melodically, harmonically, rhythmically)
-And keep it simple at first
-Improvise some lyrics overtop that
-Mix and match
-Continue to stir and have fun
-Figure out what you've done after doing it a few times
-Do it some more

[Egg]

This is so much purer than the muck I started in the other help thread.

A wise man advised me to listen to songs I liked. Song Fight is a good place to start since you don't have any biases about these people and there's really close to no feeling that you should like a track or shouldn't. So listen to songs you like (and ones you dislike!) and figure out what you think you can infuse into your own creation.

As for the first few notes, I'm no expert, but I'd say figure out a goal like a mood that you're striving to convey. Then just do some of that riffage until something sticks. Then find a way to arrange that really well. Then you can go in and add and make it interesting and fine tune it until it feels like it's meeting your goals. And then you send it in and hope that you can bear the ridicule you've called down on yourself. heh, oh, me, oh, my.

And they're right. Make a simple choice and stick with it. Like, find just two or three chords that you really like to go between and find interesting ways to structure a song around them and you can't get all convoluted and messed up. Don't do that with 15 seconds of highpitched buzzing, no matter how much you like the highpitched buzzing, or nobody will like your song. That of course, is a road you might find yourself on despite the best warnings.

/oversimplify

[erik]

Arragement, Musical Riffage, Vocal Melody, Lyrics


As far as musical background goes, I've been playing trombone off and on since high school, so, something like 15 years, and I know how to play chords on the keyboard, but can't do anything that sounds very pianistic.

Start* by just playing the keyboard. You don't have to be all pianistic, you can play it like a pretty girl in a hip band from Brooklyn. One finger at a time is all you need to start. Find a series of notes that you like the sound of. Record that on some sort of multitracking software. Now, play that original section back and try and come up with either a harmony or another melody line which will compliment the original one you've already recorded. Your original playing could be a bassline, but it doesn't have to be. It could be a lead, or it could be more of a backing line, but it doesn't matter. Get down two or more good parts.

Play back all the parts and try and sing over it. Try not to sing the EXACT same thing that you hear one of the other instruments already singing. Don't worry about lyrics, just sing whatever comes off the top of your head. When you have something that you like, record it, even if the words are super dumb..

Play it all back. Does the music sound like it needs to change to different section at some point? If so, write another section, same process. Keep writing as many sections as you think the song needs. String them all together and you have a song. At this point, swap out any lyrics that you think are totally retarded.




*This is not the only way to write a song, but it might be a good one for someone such as yourself who has no idea how to start.

[Adam!]

Recently I've been cleaning out my harddrive and listening to some really old songs that I wrote as I find them. It seems that this is how I did it:

1) Write one hundred 3 minute pieces of music, with or without some mumbling over top of it
2) Throw them all away
3) Write a song

If you do that I think your one-hundred-and-first song will probably be pretty good.

[erik]

If you do that I think your one-hundred-and-first song will probably be pretty good.

Or you could have shitty songwriting down to a science. Everyone's early songs are bad, but they're not like baby teeth that will definitely be replaced by something better. You only get better if you examine your own work and identify things you'd like to change (and then try to change them) and if you have a idea in your head as to what you want your music to be (and then try to make that idea happen).

[WeaselSlayer]

I actually listened to some of my old songs recently and I think I had a much purer sound then. It was purely unadulterated experimentation. It was absolutely thrilling to listen to a time when I wasn't in a pretty comfortable groove. I actually think my early songs were my best work.

[Dan-O from Five-O]

I actually listened to some of my old songs recently and I think I had a much purer sound then. It was purely unadulterated experimentation. It was absolutely thrilling to listen to a time when I wasn't in a pretty comfortable groove. I actually think my early songs were my best work.

I know what you mean, it's one of the biggest hurdles I have to overcome. If you get too comfortable in that I,IV,V thing it's likely to become increasingly difficult to think outside that box.

I like Erik's advice, I just need to stop worrying that I'm wasting resources recording things that won't develop and go with the flow. If it does develop - great, if not move on. That in and of itself is something you need to be prepared for. You'll either end up with a bunch of incomplete stuff or recordings that years later you look back at and go "what was I doing?". Then again, you might just find a nugget or two.

Either way, you won't be alone.

[Adam!]

If you do that I think your one-hundred-and-first song will probably be pretty good.

No.

I didn't realize this was contentious. Pre-songfight I used to spend forever slaving away over songs, and therefore worked at a snail's pace. But I found that no matter how much polish I put on a song, about 50% of the time my next song came out better. I've come to the conclusion that, like nearly everything else I have ever tried in my life, the more often I do it the better I get at it. It's the Pauling Principal: "The best way to get a good idea is to get a lot of ideas".

I always assumed that this applies to other people as well. If anyone out there can write a hundred songs and end up a worse songwriter afterwards, well... that's just very sad then.

[erik]

The best way to get a good idea is to get a lot of ideas.

Well, sure. The best way to get a winning lottery ticket is to buy alot of lottery tickets. But that doesn't mean that the more tickets you buy, the better you will be at selecting winning numbers in the future. Increasing your odds doesn't mean you increase your ability. If you want to write a good song, then, yeah, write alot of songs. If you want to improve as a songwriter overall, then write alot of songs, think about how you could improve them, think about ways to challenge yourself as a songwriter, think about whether you are getting better or worse, think about whether you are repeating yourself. And I don't mean any sort of deep intellectual bullshit, I mean just like "I'm going to try using this technique next time because I think that would be cool" or "I'm going to try and write something in a minor key" or "I think this song sounds alot like another song I already have, so I'm going to keep working on it".

I'm probably real close, but I don't know if I've written my first 100 songs yet. If I tried, I could shit out 100 songs in a day just for the sake of shitting them out, and all of them would suck. It wouldn't make me a better songwriter.

[Mankind]

Ok, wow. This is all great advice. I feel really motivated now, if not inspired. But you can definitely expect an entry from me some time in the near future.

[Adam!]

The best way to get a good idea is to get a lot of ideas.

You've misused that quote.

Yeah, you're right. The original quote really just means a brute force approach. It still had a good ring to it.

If I tried, I could shit out 100 songs in a day just for the sake of shitting them out, and all of them would suck. It wouldn't make me a better songwriter.

Yes, it is possible to concoct a scenario where my "Writing a hundred songs will probably make you better" doesn't apply, and yes, that is why I used the word "probably" instead of "necessarily".

But, assuming Mankind isn't John Cage or Denyer (aka Mr. I-Recorded-An-AAD-In-30-Minutes), and actually puts some work and thought into those songs, then the experience will [most likely] improve his mixing ability, his grasp of progressions and melody, and his singing and playing, simply because people with a vested interest in their craft seldom make something and have no opinion about it. I think even the tiny act of forming an opinion on something you made is enough to help direct your artistic efforts in the future: if you make a song and you hate it, I can't see your next 99 songs being identical.

And yes, I'm sure there's some example, most likely involving Wesley Willis or parrots or other noise-producing members of the animal kingdom, where this doesn't apply. But I'm assuming Mankind has at least two brain cells to rub together.

Maybe part of this is wishful thinking: I, like you, haven't hit a hundred songs yet. But by the time I do, I hope I'll be awesome.

[erik]

But, assuming Mankind isn't John Cage or Denyer (aka Mr. I-Recorded-An-AAD-In-30-Minutes), and actually puts some work and thought into those songs, then the experience will [most likely] improve his mixing ability, his grasp of progressions and melody, and his singing and playing, simply because people with a vested interest in their craft seldom make something and have no opinion about it. I think even the tiny act of forming an opinion on something you made is enough to help direct your artistic efforts in the future: if you make a song and you hate it, I can't see your next 99 songs being identical.

I swear I'm not arguing a semantic point here. Ultimately we are saying (close to) the same thing: that people who put thought and work into their songs will improve over time. Where we differ in opinion is with regards to exactly what constitutes thought and work. I think it takes more than merely forming an opinion about one's music to really make you a better musician (although it does help you hammer out a firm sense of style).

I mean, I know I've drawn 100 pictures since high school. But I haven't gotten any better since then. I reached some sort of plateau, because the level that I'm at, I'm comfortable with. I took several art classes in high school, and the teacher would always come up with these excellent suggestions that would have improved the piece, but were either beyond my ability, or would have required alot of effort. I formed opinions about my work, sure. There were certain things that I could do decently, and I kind of stuck with that. My teachers were trying to make me a better artist, but ultimately, I couldn't be bothered. If I had followed their advice, I would have gotten better, but I never really did.

So I'm not really arguing for the existence of a theoretical person whose repeated artistic endeavors don't improve over time (but who can't be found in the real world), I'm relating this to myself and my experiences with how prolificacy does not have a direct correllation with improvement. And I've seen enough shitty bands stay together for years and make absolutely no improvement to know that I'm not the only one this applies to.

Which is not to say that I'm dissing on writing as many songs as you can. I think that writing songs is great, it's just the more work* you put into the process, the more rewards** you will reap in the end.


*defined however one would like
**also defined however one would like

[Adam!]

I'm not really arguing for the existence of a theoretical person whose repeated artistic endeavors don't improve over time (but who can't be found in the real world), I'm relating this to myself and my experiences with how prolificacy does not have a direct correllation with improvement.

OK, got it. But I'm talking about writing your first song (although I suppose it would technically be your 101st song), which is what I assumed Mankind was originally asking about. Were you any better at drawing by the time you had drawn 100 pictures than you where when you first laid pencil to sketchbook? I'd be surprised if you weren't. I think of any kind of creative effort, be it writing books, drawing, or making a song, as having some kind of "learning to ride a bike stage" where, no matter how much information you have, the only thing you can really do is get on it a fall over a hundred times. And the hundred-and-first time you get on that bike you may still fall over, but I'm sure you'll get farther than you did the first time.

My "start-by-writing-100-songs-and-throwing-them-away" statement was a paraphrased quote by some musician (I haven't attributed them because I can't remember who it was. For some reason I think it was Tricky, although I'm sure it's been said before him), and I've always treated it as common sense.

[erik]

Were you any better at drawing by the time you had drawn 100 pictures than you where when you first laid pencil to sketchbook?

Yes, but it's hard to compare songwriting to drawing because most people start drawing before they've mastered all their motor skills. And I'll leave it at that, because now I feel that I <i>would</i> be arguing semantics if I kept it up. Writing lots of songs is good advice for anyone who wants to be a songwriter. It's by no means the only advice I would give someone, but it is good advice nonetheless.

My "start-by-writing-100-songs-and-throwing-them-away" statement was a paraphrased quote by some musician (I haven't attributed them because I can't remember who it was. For some reason I think it was Tricky, although I'm sure it's been said before him), and I've always treated it as common sense.

I've heard the quote, too. I seem to remember (though I can't verify) it coming from the mouth of someone who is considered both very prolific and very good, like Paul Simon or Paul McCartney, or maybe Randy Newman, someone who had #1 hits within his first 100 songs. The way I remember it is "Everyone's first 100 songs suck". The way I interpreted it was to not rest on your laurels, and to always try and get better, and to never pat yourself on the back too much.