June 8, 2009 What's your songwriting process?
If you’ve been writing songs for awhile, you’ve inevitably been asked, “Which do you start with - words or music?” It’s not always that simple!
I usually work from a title. When one hits me, I’ll rough out some stream-of-consciousness prose to make sense of it. It could end up meaning something completely different than what I thought it meant at the beginning. Next I’ll flesh out the song structure and melody. Then I’ll mold the useful bits of my garbled prose into a lyric. The production goodies come at the end - typically the hardest part for me. At that point, I just want it to be done. I can only spend so much time finessing automation envelopes.
How about you? Where do you draw your inspiration? What hits you first — a lyric, a melody, a groove, a bass riff? What’s your songwriting process?
Reader Comments (28)
Great question Brian..
It reminds of me of starting a company.. I always begin with the name.
Out of curiosity...how do you (or do you?) check to see if names, themes, or lyrics that have been over-used when it comes to songs?
-Bruce
Honestly, I don't have a single consistent songwriting process.
I suppose my usual process is as follows:
1. Come up with basic riff, melody, or chord progression. This may occur while randomly noodling on an instrument, or it may just pop into my head, frequently while I'm asleep or half-asleep. As I am primarily a guitar player, this will usually happen on guitar, but sometimes it'll be a keyboard / synth / orchestral thing instead.
2. Develop music for awhile.
3. Somehow an idea of what the song will "be about" comes to me, and from there, I start to develop lyrics. This isn't always how it works though; sometimes a basic phrase will accompany the original musical idea in step 1.
4. Continue to develop music and lyrics. Add other instruments.
On rare occasions I'll write music and put in some lyrics I already had written.
Also, when I'm writing songs to fit a specific concept/project (like I did with my album "Now You're Gone" that you're quite familiar with) I definitely have topics in mind as I'm writing, and certain topics or emotions I want to cover.
Though every once in a great while a song (or story) will spring fully formed from my brain like Proteus from the head of Zeus, most of the time I wait until I have at least two germ ideas that can be put together. For instance "It would be cool to write a salsa song", "I've always wanted to pastiche one of Pete Townshend's heavily arpeggiated synthesizer experiments", and "I wonder if Jeffrey Dahmer ever staged tea parties with the lobotomized corpses of his victims, and if so how did their pretend conversation go?" Everything else follows from the "interference pattern" created by trying to harmonize the various ideas - the more disparate the ideas, the more interesting the interference. Et voila - we get a salsa-inflected rock opera song about Jeffrey Dahmer's lobotomy tea party!
I typically don't worry about the uniqueness of my titles. I figure I'm okay as long as I avoid obvious clichés like "Stay" and "Miss You." I try to draw the listener in by piquing their curiosity. Examples from my last album are "I Stand By My Mistakes," "We're Not Getting Any Younger," "Zero Crossing," and "If It's My Time to Go."
James, I've done it your way too, but I find it trickier. Some of my older songs have three or four sets of lyrics, because I couldn't nail down what the song should be about. Obviously it works for you, because you've succeeded in creating a coherent concept album that tells a compelling story (apologies for the awful alliteration!)
That said, Emil is the king of the concept album, with ten concepts across ten genres. People probably think you're joking about the lobotomy tea party, but that sounds like a typical Mystechs song. ;)
Very good question...
I start messing around with some chords and lalala ing with some words thrown in and see what comes out. It's amazing the phrases you can come up with subconsiously. I then play it over and over until more words come...that way the scanning feels right. Finally I sort out the structure and, like yourself, I just want to to be done.
About 1 out of every 5 chord messing around sessions makes it into a finished song.
I've had a few.. Originally when I first started making music on guitar it was simply playing around with chord progressions and different chord voicings and finding something that didn't sound like other songs (which seemed like a chore for some reason back then).
When I started doing electronic music it was really about playing around with sounds and seemed very building blocks architecture-esque. The second TB record was very much a "concept"record where lyrical ideas existed first and played the largest part in the evolution of what a song would sound like.. The third TB record was a combination of everything.. Lots of moments played by a piano and on acoustic guitar and translated to electronic format (or a hybrid) with lyrical concepts. From there I wrote more organically like that...
Now I've taken along a Morrissey path which I found really interesting. He always writes the song titles first. So I'm working a project called Passing Thoughts. All the song titles come first from an idea (or passing thought, if you will) and then the songs really flow lyrically and musically from there.
My titles often come last. I'll have working titles like "The Aaaaaaay Song". I find what works worst for me is sitting down at the piano with the intention of "Okay, now I shall write a song!" I diddle at the keyboard for a little while, come up with soulless chord patterns, get discouraged, slump my shoulders, and head back to the safety of the couch to click hrefs.
What works best is to identify an emotional idea, some collision of two feelings that I find insightful, something that twists up unexpectedly. Like my buddy that told me how his daughter wanted him to tell her a story good night, and how the only thing he could think of was the story of Anakin Skywalker. His next bedtime story was about Superman. I started thinking of innocence and sweetness, combined with superheroes, and fatherhood insecurity. I came up a feeling-progression that eventually turned into musical swoopiness, that turned into harmonies, that turned into riffs and actual notes - somewhere in there the melody emerged, and then it was about jamming the actual words and lyrics on top. Tons of freewriting helped, but that was the hardest part. I think that process leads to the most organic song, but trying to write lyrics to jam into a mostly existing song is one of the hardest things to try. That song ended up as She Believes.
I've also freewritten lyrics without any music - Damn My Eyes is a good example. The title again came last - the concept of feeling like a loser in a bar was what came first. I was originally thinking it would be some harder-edged bar-band kind of tune with electric guitar, but I don't have much experience with that kind of music anyway. It sat on my hard drive for a year, and then I pulled it out, improvised some chord progressions (Sting's "It's Probably Me" was a very loose harmonic inspiration, but so was "One For My Baby"), and then my trio and I stumbled into a fun rendition. That music ended up much more approachable but also much less fresh. Most comments I get are that it's a great song, fun, good instrumentals, but not really anything "new". I've gotten more fans from that song than from She Believes, which I see as the better song.
I'm still feeling torn between those two general songwriting directions. (Any comments appreciated through my site.)
If I am trying to tell a specific story, or write for someone, the lyrics come first and then i write music that serves the song.
Usually I'm just fidding around on guitar or keyboard, and something musical will strike me. If I like it, I start recording what I'm doing immediately. (this way I won't forget anything, and if I try something in the moment I can go back and listen to what I did. I'm tired of kicking myself for forgetting I tried in passing). the addition of words/phrases and melody happen simultaneously most of the time, and I experiment until I land on something that feels right. from that point i try to nail down the form of the song, and then tie up the lyrics. I don't like to worry about additional instruments and sounds until the song has a solid skeleton.
Well, music then lyrics for me. but I first start with a rough production of ideas :
- for music, messing around with my guitar for half an hour everyday, trying different parts, trying things I don't usually do, recording everything. I do that for a week or two, so I get a lot of material to work on, generally a lot of good ideas, a lot of bad ones too. and i'm more trained to be inspired so other ideas come everyday...
- for lyrics : reading books, web articles, dictionaries and thesaurus, picking inspiring sentences and words... then trying to write what follows the sentences using the words picked.
With music ideas, I then pick the ones I feel something strong and build a simple structure, verse chorus bridge, verse chorus most of the time. well, i'd like my songs to be really really simple...
When structure is done, I write the lyrics on it, starting with or using some the sentences I already wrote, then writing new ones.... I rely a lot on the dynamic progression of the song, singing lalalas before jumping to real lyrics...
For the lyrics, I focus a lot on the first sentence... that's the beginning and it's got to be strong... The title comes at the end...
I know the song is ok when it works with those two elements : guitar + voice.
What follows is the recording step... the next step after writing a bunch of songs...
"Freewriting." I like that! It certainly sounds a lot nicer than "brainvomiting," as I've been calling it.
I too find myself singing nonsense syllables when I write at the piano. Occasionally, I'll stumble onto an interesting phrase. Other times I'll find myself repeating a silly or stupid phrase, which gets lodged in my brain and can't be shaken off.
In high school, a friend and I were in a band together called The Thought Chapter (which I adopted as the name of my latest Color Theory album). We put out fliers announcing our availability for parties and weddings, claiming we had a demo tape available, but no such tape existed. Of course, someone asked for the demo so we made one in a weekend, playing through every song we'd half-written. To this day, I still don't have any verses for one of those songs. I mumbled through them for both the tape and several performances. Hmm... actually, I think I mumbled the same words every time, so I guess that qualifies as a lyric!
my process most of the time:
melody most usually always comes first. i'll be walking, or riding on a train, or on a bus. i'm inspired by movement, it seems!
i then figure out which chords go behind it when at home with a guitar.
next is trimming it up, cutting uninteresting/ unnecessary bits, smashing similar melodies together, etc. this is where i start developing the form, and how all the parts will fit together in a smooth way.
last are lyrics. and unless i've been fixated on a certain phrase during the whole thing, it's usually hard to figure out what to write about. i've been known to dread the lyric-writing process, but i think i've gotten better at it over the years. i now focus on trying to tell a story.
the finished product always feels like a grand accomplishment, and makes me smile like this :D
A timeless question...
For me, there is no official process, no one approach that always delivers and that I always employ, but the process that seems to work best most often is to start with a title, write a verse or chorus, sometimes with a guitar in hand, sometimes not, then answer the question "What is this song about?". For me, answering that question creates boundaries, which I really, really need. The boundaries keep me focused on what I'm trying to say, and paradoxically, create freedom, in that I don't try to both think of what to say and how to say it. I just focus on how. I also make an effort to write without an instrument in my hand, to just imagine the song, then try to re-create what I've imagined.
Jeff
www.cerebellumblues.con
for our next EP, our band is playing around with the idea of naming songs based on the first or most obvious "thing" it reminds us of/sounds like/whatever. for example, we have a song that sounds like it should be the background music on an island. Boom, "Island Song." We also have a song that makes you feel like you are up in the clouds. You guessed it: "Cloud Song."
I wonder how our audience will respond!
_chris
---------------------------------------
Chris Bracco is an aspiring music producer/music biz entrepreneur. Chris currently attends Penn State University, working towards a major in Business Management and minor in Music Technology. He is also currently interning for Jive Records within their marketing/digital marketing department. He also plays guitar in & manages a funky rap/rock quintet named "A.S.B.P.K."
If you would like to learn a bit more about Chris, please visit his personal e-portfolio, his blog or his band's website:
Chris Bracco's E-Portfolio
Tight Mix -- The Future of Music & Audio Recording
A.S.B.P.K. Music
If you would like to contact Chris, please don’t hesitate to e-mail him at cob5020@psu.edu
That's a super question, and so difficult to answer.
I guess usually I start with a guitar or piano melody, record it, loop it, and wait for lyrical inspiration to strike.
As a solo artist, I use a looper to add layers, so once I've got the main instrumental melody and lyric, I'll add a couple of layers.
I almost always construct my chorus and my verses separately, which sometimes can lead to them being "patched" together in sound, but I haven't found a better way to do it yet.
Thanks for getting me to think about it though.
I usually find that a new riff or groove or chord change seems to suggest a bit of a lyric - maybe no more than a couple of words - and at least a hint of a tune.
Sometimes it all just follows easily from that, sometimes not. I just keep paying and humming until something makes sense to me.
Try to be open, especially in these early stages. If you chance upon a thread, follow it. Don't worry if it's taking you away from where you thought you wanted to go. You can always go back if it doesn't lead anywhere.
Sometimes I treat the tricky ones like a puzzle. Or like I'm a detective solving a crime. Examine the scene. Look for things out of place. Follow the clues. Every idea is a suspect until they can be eliminated. Where's the motive?
It's all about the higher power!
It just comes, not sure where from. If anything I spend the majority of time concentrating on the lyrics and delivery afterwards, but the structure and melody just flows )
I often find the hardest part is actually the title if it doesn't immediately jump out in the lyrics!
I'm 24 now, and for most of my songwriting "career" I've always started with the music first, then the melody and lyrics.
I started thinking about what was important in a song, what made it memorable, and why people liked songs so much and I realized it's all in the melody. No one walks around belting out the chords to a song. They sing the melody.
So that being said, I've started to try and write the melody first (lyrics or none) then the music. I figured I shouldn't let my melody be restricted by a chord progression. It's a bit of a challenge to write the music for a song after the melody is written, but it's been great writing melodies without any restrictions.
i tend to write in one of two ways:
the first being the way i originally started writing songs, which was born out of simply writing. writing to process. this mode of song writing tends, for me, to produce the more pensive and maybe melancholy "singer/songwriter" style music. more intent on the lyrical depth and a specific train of thought while the music mostly backs up the words. not necessarily provoking or flashy.
the other mode of writing that i've started developing the past 5 years or so is little more interesting and has helped me to write songs with "more room" in them. something that's complete but has a sense of space for a band or a producer to fill out.
i'll start with either an interesting riff, a solid chord progression or a seminal thought as the heart of the song. then, whether i started with the "thought" or the music, i focus on creating the other as i'm hearing or thinking about the first.
for example: if i started with a thought, as i'm working on the music i try to make it say the things i'm feeling about the thought i started with. and in the other i focus on the music i've started writing and i start to see in my mind what the music is saying. what story the music is telling and i try to put it into words- but not too specifically. i want to tell the sense and the feeling of an experience rather than the specific experience itself. i think leaving room for the listener to project their life and experience into your songs is important and this allows them to see themselves in your work more.
Simple. I just stare at the page until drops of blood fall from my forehead onto the page.
Fascinating stuff....I can relate to most of the methods that have been described. For me, it seems like the very best songs seem to come when I'm not actively trying to write anything. I can be doing just about anything; falling asleep, cooking, etc. and I'll get a song almost fully formed. If I'm away from my guitar, I'll use my cellphone to create a voice clip of me humming or singing...whatever I need to rememeber it for later. Otherwise i'll grab my guitar and demo it right away...that is the best, because the song sometimes just plays itself from start to finish, and I can tell it will be good. Usually the melody will be close to fully formed, with some key words and phrases locked in that seem to 'own' that song. At first, it can seem nonsensical, but after fleshing the lyrics out a bit more everything falls into place and know exactly what I'm writing about.
Other times, just picking up the guitar and singing while playing will yield some good ideas that take me in the direction of a good song. It can also just produce fragments that don't quite click, which i'll then put away for later.
When things are really going well with a song, it feels to me like i'm almost recieving it passively (like an antenna) rather than actively forcing it.
-Jay
www.jayscarmusic.com
I know what you mean about having best ideas when you're not really looking for them, like when you're in bed somewhere between wakefulness and sleep. I lose a lot of good stuff because I'm too lazy to get up. Y'know, I should probably download a sound recorder app for my iPhone. :)
If they have one, I'd definitely recommend it Brian! You'll rest easier knowing you've captured those ideas, even if you might never use them. I still record a few that don't seem nearly as cool the next day :)
Excellent post and comments, my songwriting process might be interesting to someone. I play bass in a post-hardcore/rock group but do most of the initial songwriting. I utilize some cheap gear (epiphone sg / ibanez bass / tascam us-122) along with FLstudio for recording ideas and drum sequencing.
I always start with drums, usually trying to find an intro beat or chorus beat i like. Once i find something i like, i build the song around that, jamming along on gtr, fleshing out ideas for the drums for the whole song before actually laying down the first guitar track. depending on what i've got so far, i usually end up getting most of the rhythm stuff down. I like to hit as many lead guitar parts as i can right after that, standing up, moving around and getting physically excited about what i'm creating. I know it's not good recording technique, but i like to have all my stuff eq'd and compressed from the get go, it amps me up and keeps me going when all the guitars always sound awesome. bass is always last for me, and then if anything is conflicting i'll make touchups and whatnot.
I tab out all the parts notepad style (sometimes noticing i should make some changes, or what i did was really sloppy) then i utilize box.net for my band, so i upload the mp3 to our shared box. luckily we all have iphones (box.net has an iphone app) so it doesnt take long for me to get comments back from the band. if everyone green lights it, then everyone learns the parts before coming to practice (usually). At rehersal, usually my lead gtr player and drummer have taken what i did and put their own spin on them. We make structuring changes if needed and after 3 or 4 practices, usually have everything down tight. as far as lyrics/vocals go, i always have ideas in my head but i dont touch em. We have 2 lead singers and i hear they like to have the tracks in the car immediately available whenever they want.
Over the years, i've played everything in bands (drums / lead vox / rhythm gtr / lead gtr / bass) and @ 25, i've figured out that i love to write songs by myself and be as free and energetic come stage time as i want. Handling the main songwriting and playing bass live seems to be a good fit. =)
It depends totally on the song's intention. If it's a commercial song for somebody who pretends to obtain a publishing contract or so, I tend to think about the theme (to go with the personality of the artist). It depends totally on the project.
As for MY projects, it also depends on the project, whether it's more musical-intended or if it's lyric/content minded... sometimes the best songs (and projects) come from simple doodling on an instrument (usually a guitar or a groove box) and sometimes you just file the idea for possible future use.
My process usually comes out of nowhere and slaps me on the back of the head when i least expect it, am nowhere near an instrument or studio, and am running late for something. Then, when i'm within range of some kind of writing tool, i attempt to extract that sucker like trying to pull out a splinter from somewhere where i can't reach...or it all happens purely serendipitously
There's some cool songwriting videos I saw recently that had some brilliant ideas and methods for songwriting. Check it out here
I do things similar to you, Brian. I come up with a title, or a series of ideas, and begin to build on what they mean to me. I write down tons of prose, usually at about 3 AM (A prime time for creative lyrics) on note cards I keep next to my bed. Sometimes I'll also draw as I write, just to kind of get the creativity rolling. Although I am a very musical person, I do not play the guitar or piano, bass, etc. So my melodic notes usually get forgotten. Hopefully my guitar player can write music himself! That's what Mick Jones did for Joe.
I feel very creative at 3 am too, but end up discarding most of the ideas I come up with in the sterile light of day.
Over the past couple weeks I've been posting little production snippets at www.colortheory.com. Eventually I'll expand some of them into songs. It'll be nice to have what I consider the hardest part out of the way first!