Fitting chords to a bass line...suggestions?

[Gruuve]

Hi comrades:

 

I've attached the standard notation for a bass line that kind of came out of my hands today. When I write any new music, it usually starts with a bass line, then I fit chords to it. I've got some experimenting to do, but I was hoping maybe some of the folks with good theory knowledge might suggest some interesting chord progressions to try.

 

My first thought would be perhaps Amaj bar 1 count 1, Dmaj bar 2 count 1, Gmin bar 2 count 3, Dmaj bar 2 count 4, and repeat.

 

What I normally do is take something like this, and plug the midi file into Band-In-A-Box and start throwing some chords at it until I get something that has a sound I like. I (mostly) know the chord quality for each scale degree, but there's still a bit of trial-n-error (at least that's really easy to do in BiaB).

 

Bass Riff PDF file

 

Here's the midi file with just the bass riff if you'd rather hear it. For whatever strange reason (probably just my laptop), the midi file plays correctly the first time through, but starts skipping notes on the 3 repeats...don't know if you'll get the same behavior...let me know.

 

Bass riff midi file

 

TIA,

Dave

[alexclaber]

I know this isn't the theoretical answer but I'd be inclined to loop your bass riff and then start improvising over it either high up the bass neck or on guitar, just with simple 2, 3 or 4 note chords. I wouldn't think about the theory or what the bass was playing, I'd just mess around and find something that sounds good.

 

That's the approach I take either way, whether starting with bass or guitar - and if I get stuck, I go back to my theory to see if it can get me over the hurdle.

 

Alex

[Gruuve]

Hey Alex:

 

Actually, that's kind of what I usually do (although I generally just use BiaB, one of my goals for improving on guitar is so I can do just that...bang out simple chords to see how they sound). I guess I start with a little bit of theory (ie. knowing which chord quality to initially try based on the scale degree, etc.)...but from there it's just a matter of using my ears. I kind of think of it in terms of "which notes in the bass line do I want to accent?", then I put chords over those notes. Sometimes what seems to need the accent lands on a downbeat, sometimes not. With a bass line like this one that touches just about every note in the scale, there's plenty of possibilities!

 

So, I guess I'm looking for suggestions on how to use theory to maybe narrow down the possible chord choices some more, or something along those lines.

 

Regards,

Dave

[jeremy c]

That's some weird chords, Dave.

 

I'm trying to figure out why you would want a Gm chord there.

 

|A7 |DMaj7 | is unusual enough already.

 

I'd prefer:

|A7 / / /|A7 / D7 /|

 

If you are writing a song and not just a groove, I really really would recommend writing the chords first.

[Steve Force]

Like jeremy c posted, doesn't it make more sense to chord the tune, and then work the bassline around them?

 

This way, you can work the established musical theory/best practices into the line, rather than the other way around.

 

My 02 cents.

[Gruuve]

Hey guys:

 

Yes, I could define the chords first, then work a bass line around the chords. That's the way it's usually done. That's also exactly why I'm doing it differently.

 

I like songs that have a strong groove, and I believe that much of the groove comes from the bass. So, if I start with a grooving bass line, then work out one of the chord progressions that's implied by the bass line, I end up with a grooving song.

 

I realize this is not the conventional approach...but to me that's all the more reason to get better at doing composition this way. Nothing wrong with thinking outside the box, right?

 

Dave

[Eric VB]

Dave, this is a great songwriting question!

 

With my own stuff, I probably tend towards Alex' seasoned answer: experimentation over analysis. That's because I often "hear" other things in my head.

 

Jeremy has a good idea, too, in keeping it simple. Sometimes throwing in too many chords spoils the "soup", as it were.

 

Writing a song from a bassline on up can be quite challenging.

 

I've listened to John Paul Jones' solo recordings, and to me it seems like some of his compositions (from bass) have a great groove but not always a lot more. One song, "Bass 'n' Drums", he must have felt was better off leaving as just bass and drums (duh), without further arrangement. (I consider this recording to be his only real "bass solo", so maybe that's what he was going for.) You can usually tell what he's written from piano or guitar (or mandolin, for that matter). On his latest CD, "Thunderthief", a lot of these POVs converge (and he shows off some pretty good guitar playing as well) for better overall compositions, IMO, than from his earlier "Scream for Help" movie sound track. (I see "Zooma" as being transitional between the two.)

 

I found it interesting -- not just because I was a big Zep fan -- to hear how a bass player could write songs because I was struggling to do so myself. I found JPJ more inspirational because he always seemed to want to showcase the bass more prominantly than other bassist/songwriters like Sir Paul or Sting. Maybe they are less commercially successful compositions that way, but I think a lot of bass players can appreciate what JPJ has done.

 

(I have more to say, but the wife is calling.)

[I I mjrn]

Originally posted by Dave Sisk:

Hey guys:

 

Yes, I could define the chords first, then work a bass line around the chords. That's the way it's usually done. That's also exactly why I'm doing it differently.

 

I like songs that have a strong groove, and I believe that much of the groove comes from the bass. So, if I start with a grooving bass line, then work out one of the chord progressions that's implied by the bass line, I end up with a grooving song.

 

I realize this is not the conventional approach...but to me that's all the more reason to get better at doing composition this way. Nothing wrong with thinking outside the box, right?

 

Dave

Not at all anything wrong with thinking outside the box...but what often happens is that U may just put Urself N another type of box.

The approach of writing from a either a groove or a chord progression isn't all that "outside"; it's how many beginning (or even veteran) writers work....but it's also the way bebop improvisers took established tunes N2 new sonic landscapes, so there's nothing necessarily "beginner" about it.

It can lead 2 a certain formulaic outcome, though, that may not B as distinctive as it seems 2 the composer.

 

What defines a "strong groove" 2 U ? The particular rhythmic phrasing or the bass lines/riffs ?

It's been fairly conclusively demonstrated that rhythm is a predominant element of musical recognition. People will recognize rhythmic phrases with different notes as particular songs much more readily than they'll recognize precise note series with altered rhythms...so it could well B that U can alter the exact notes of a bass line & keep the power of Ur groove.

 

U might want 2 get a basic idea & then, before U elaborate 2 much on that, develop an idea of a tune/melody that will B the basis 4 developing the accompaniment. That's hardly required; we can hear commercially (& musically) successful records that R based completely around some groove but how many of those become memorable songs ?

 

I'm not suggesting that U R on the wrong track overall, just that U view the different components (melody, harmony, rhythm) as more interactive & look 4 ways 2 alter them so that they fit together N whatever way U find most pleasing rather than develop any single 1 too far & then try 2 hammer them together around that single focal point...which is actually just an elaboration on what some others have said, I guess.

 

James Brown was famous 4 the tight focus of his creations (a favorite quote: "If U have an idea, thats' a song; if U have another idea, save it 4 the next song") but remember (Bcause I suspect this isn't something new 2 U), that as U develop an idea U should B flexible about adapting it Bcause new elements can make it stronger.

[Gruuve]

I&I...definitely understood...I often find that a bass line may change as I work on the other compositional elements.

 

Eric, I'll have to check out some of the JPJ stuff you've mentioned.

 

Dave

[Dave Brown]

I looked at this riff a bit, but I didn't pull out and play it. I'm trying to replace the AC belt in my daughter's car.

 

At first blush, measure 1 seems to be based around an A7 chord, with that weird D note...making it a sus. Then measure 2 seems to be an F#m7 chord, again with the D, but then those G's make it kinda wierd.

 

So when I get a chance to play it, my ear may inform what my eyes can't see.

 

It does look interesting, though.

[Eric VB]

[Good luck on that AC belt, DB!]

 

Dave, the other thing I was going to bring up was "instrument interface" (again). Sometimes it's just tough to get what's in our head (musically) to come out of a given instrument. But you have command of an instrument with a very intuitive interface: your voice! And you can multi-track record, so you can build chords a note at a time that way. (Or Alex' suggestion of playing higher up on bass, or even just playing monophonically -- one note at a time -- on guitar.)

 

Now, from a theory standpoint, if you have two notes -- say F# and A -- what chord is it? Well, it could be F#min (missing the 5th), or it could be something else (like a Dmaj missing the root). Sometimes it's easier to just let it be F# and A and not worry so much about the rest of it.

 

Now, you could go the Zep route and write riff-based songs. A lot of times the trick is to play something plain vanilla during the singing and have the vocals and riffs take turns. For example, for your riff, you could play it a couple times and then (because everybody agrees it's pretty much in A) hold a some-flavor-of-A chord to sing over, sing something, then go back to the riff. (But it sounds like this isn't where you want to go.)

 

You could also try looping that riff and come up with a melody/lyric over it without worrying about the chords. Once you have the melody defined, it may be easier to hear the chords.

[Eric VB]

Here's a visual art analogy, if it helps. Paint some big splashes of color on a blank page. When the paint dries, draw line drawings on top. This is the "backwards" way of doing things, but can yield some interesting results. (The "correct" way to do it, of course, is to draw first and then paint within the lines.)

[Gruuve]

RBG:

 

Good suggestions (as I'd expect from you ). This is just one riff...I'd expect the end product to end up being more than just this of course. I often use the idea you mentioned (holding out the implied chord to sing over, etc.)...that could work well for this. I could possibly see doing the same or similar rhythm holding out the A, sing something, then do the riff a time or two, then a straightforward chorus...something like that perhaps. Or the riff might get moved around to a different root, etc. (Approach #1 is probably what would work best...kind of what I was originally thinking.)

 

Unfortunately (or fortunately, depending on how you look at it), Band In A Box can play chords on a variety of instruments much better than I can. Notating out this bass line actually had two purposes. First, to force me to continue to improve my reading skills (transcribing certainly does that). Second, once I've notated it, I can then export a midi file from the notation and import that into BiaB (or Jammer, or whatever) and try a whole bunch different chords with it by merely typing in chord symbols. Once I get something I like, I can then render each instrument as a WAV file, import it into my recording gear as reference tracks, and away I go. (This is actually the process I used for the last collaboration song, Primitive Existence.)

 

I of course could just hook the laptop up to the mixer and play along with BiaB, but I like forcing myself to notate bass lines. The more you do it the easier it gets.

 

This bass line was originally in my head...I'm actually pretty pleased that I got it to come out of my hands as easily as it did (and almost completely unchanged from what was in my head). I think I'm slowly getting better at that part as well. Having that direct connection between "think it" and "play it" is where I want to get, but I've got quite a ways to go yet. Things like this certainly help though.

 

Dave

[Eric VB]

Dave, that's certainly a working construction formula, as "Primitive Existance" shows.

 

I think the part that's tripping up the rest of us is this part (italics emphasis added):

 

Originally posted by Dave Sisk:

Second, once I've notated it, I can then export a midi file from the notation and import that into BiaB (or Jammer, or whatever) and try a whole bunch different chords with it by merely typing in chord symbols.

Certainly this approach has merits. I think a lot of people would be inclined, for example, to put a Cmaj chord along with any monophonic passage that featured some permutation of the notes C E G. And if you want to get "fancy" and analyze it further (and taking from a recent thread on stacked triads), in a jazz idiom you could try Fmaj9. (Or play the Fmaj triad and imply Fmaj9.)

 

And, I admit, sometimes I use a similar approach when writing, playing "the usual suspect" chords on guitar until I find the one that fits.

 

But after spending more time with a keyboard instrument, and laboring over reading piano arrangements, now I lean towards building harmonic accompaniment a note at a time. For the C E G passage above, maybe for some reason I want to add an A (to continue a pitch I was previously playing, for example). It could still be Fmaj9, or it could be Am7. It doesn't really matter at that point. I could even use an F as a passing tone in the melody and that wouldn't necessarily preclude Am7 (or something else, like C6 for example).

 

[i used to think piano was all about strictly playing chords/bass with your left hand and melody with the right. I would play and hold a Cmaj triad with my left and then noodle a melody on top with my right. That's rarely how piano is played based on the sheet music I've come across, despite what that infomercial guy on PBS says.]

 

On guitar, the same kind of approach leads me away from, for example, always playing Amaj in the standard open chord voicing: A E A C# E. Maybe I want an A(no 3rd) or A(no 5th). Although I could try every possible chord name available, I think I can be more effecient by simply finding the pitches I want, and then analyzing later to figure out what chord that may be.

 

Does that make any sense?

 

I wouldn't say there's a right or wrong way to do this. I'm just trying to give you an alternative.

[Gruuve]

RBG:

 

Yes, good ideas here. Your one note at a time chord construction method is interesting...isn't that called "contrapuntal"? Thanks for the ideas...

 

Yeah, I'm sure the statement about just typing in some chord symbols is a bit confusing. When you type in a list of chord symbols and click "Play", BiaB will actually play them as accompaniment. Go to www.pgmusic.com and download a trial copy of Band-In-A-Box...tinker with it for a mere 5 mins and then you'll understand what I'm referring to. (Jammer will do the same sort of thing, and do it better...unfortunately, Jammer won't render WAV files while BiaB will...and that's a capability that I use...Jammer is a much better and more solid seuqencer than BiaB.)

 

Dave

[Eric VB]

I just wrote a quick comp. using all programmed MIDI for something else ( mp3 here ). I started with a melody/lyrics, then added the cheesy "drums", threw on a bass line (via keyboard, as everything else), and the last thing was the arpeggiated chords. (The arrangement builds in a different order.)

 

[Lest anyone get the wrong idea, I did a lot of note editing; I don't play keys that cleanly. I left the bass a little "sloppy" but quantized everything else.]

 

I couldn't tell you what all the chords are without a full analysis. I tried not to over analyze as I was writing, and tried not to let the different parts influence each other too much.

 

At the beginning of the phrase, the accompaniment is playing an arpeggiated Ebmaj7 voiced as root-5th-maj7th. So really it's Ebmaj7(no 3rd). In fact, nobody plays a 3rd there. But I got there by experimenting instead of analyzing.

 

Keys really force me to experiment differently than if I were messing around on guitar/bass. I'm sure if I tracked bass instead, that part would be different.

[Phil W]

Looking at the music, I was thinking either

|A7///|F#m7/G7/| repeated or |A7///|D7/G7|

Hearing it now I would plump for the latter as I play along (badly) on guitar.

As has been said before, the first bar fits A7 fairly well except for the D natural which is a little discordant but only lasts for a sixteenth beat. C# is a perfectly acceptable note to play on D7 (listen to the bass I Feel Good by James Brown, or any bebop solo). The Ds and Es the passing G# also work fine over G7.

There are myriad possibilities. I sometimes find myself writing chords to fit a bassline too, I don't think it's too weird a concept. In an improvisng situation, chordal players constantly have to adapt their choice of chords and voicings to fit the bassline as well as the soloist.

Regarding Ric's points about chord voicings, people mind find this thread from the keyboard forum eye-opening as they debate long and hard about voicings for a 13 chord.

[Phil W]

Reading back carefully over your initial post:

Regarding interesting chord progressions, I think that there's some interesting 'weirdness' going on in the bassline which creates interesting tensions against simple chords as I've gone for. The weird thing is that looking at the line - all the notes are in the key of D major, except for the last passing note. Despite this, the choice of notes really doesn't suggest D major 7 in the second bar.

As to my choices above, they're all 7th chords you could try a bunch of extensions. The 9th works well on all, . Some kind of Asus7 would be good in the first bar(though I haven't tried it)as C# is a sweet note against an Asus. The D7 has lots of dischord witha C# and that G in the bass which doesn't fit well on a D7; although you should probably see the G natural as anticpating a G7 chord a 1/16 early.

[I I mjrn]

Not being a great sight reader & having no way 2 use the midi file, I took a moment 2 play w/this.

 

Looking at the offbeat accents I thought there might B a funky faktor 2 this but, at least 2 me, it sounded more "poppy" than anything else.

 

Here's my offering:

(marks Ndicate 1/16ths)

 

 

[ A//// //D// G//// D//// | F#m//// //G// D//// E(7)////]

 

A variation might B 2 play the same basic series but sub a Gadd9 & a Dadd9 4 the final chords, respectively, N each measure.

 

Simple, mayB even simplistic.

I'd leave most extensions, etc., 'til after U had a more developed idea of a melody...but that's just me.

[Phil W]

I'd leave most extensions, etc., 'til after U had a more developed idea of a melody...but that's just me.

Good point.

Looking at the chord sequence I&I mjrn, it seems to work well - I haven't heard it but it would definitely work.

[jeremy c]

I&I's chords work, but that's way too many chords for a two measure riff.

 

I still say, write the chords first.

If you consider yourself a grooving bassist, you will be able to come up with a funky part to anything.

 

Another suggestion is to have no chords at all. Why do you need them? Make up a melody that goes with your part. Make up a single note kind of guitar part. Make up a single note keyboard riff or a few choice horn riffs. No one needs to be playing a chord.

[Gruuve]

Hmmm...contrapuntal...isn't that the correct word, JC? I may give that approach a try...

 

Dave

[alexclaber]

Just been messing about with this. Try strumming a 16th note funk guitar part, based around an G octave-fifth double-stop high up on the D and G string. Experiment with moving to other double-stops when the bassline is either stable or has a suitable space for an interjection. Funky.

 

Alex

[Gruuve]

Oops...hit the wrong button...

 

I still haven't had a good opportunity to actually sit down and work on this some more...maybe this week. I'll give the 16th based guitar a try. I'm thinking something that bounces back and forth between 7th chords and sus4 7th chords might work well.

 

Dave

[alexclaber]

Originally posted by C.Alexander Claber:

Try strumming a 16th note funk guitar part, based around an G octave-fifth double-stop high up on the D and G string.

I should probably clarify, this is on the bass: Funk faux guitar! One nice thing about playing such parts on the bass is that the fatter strings allow you to use really stripped back chord voicings without it sounding empty or boring. My guitarist has some real fun trying to make similar faux guitar parts work on real guitar...

 

Alex

[I I mjrn]

Originally posted by jeremy c:

I&I's chords work, but that's way too many chords for a two measure riff.

MayB...but this is a fairly slow tempo (90 bpm). Anyway it was just one suggestion, not the ultimate best.

 

I&I like the suggestion from JC that another option would B a more free, harmonically indefinite setting, variable by soloists choice.

[Eric VB]

Originally posted by Dave Sisk:

Hmmm...contrapuntal...isn't that the correct word [...]

Well, reading the Wikipedia take on "contrapuntal" -- or "counterpoint" -- it certainly seems to fit in the more general sense. I'm just not used to hearing that word, even though I use "countermelody" all the time (which I guess is a kind of counterpoint).

[Gruuve]

OK, so here's a chord progression I like that seems to fit well, and convey the emotion that I was after:

 

A / Asus4 / D2 / D Dsus4 A / Asus4 / G / D Dsus4

 

Here's a small MP3 file (800K) rendered with this chord progression and the bass line in question (all midi-generated at this point):

 

http://www.ipass.net/davesisk/music/other/SongIdea03.mp3

 

Not positive that I'm happy with that one G chord...what y'all think?

 

Dave

[alexclaber]

They work. Quite a lot of melodic possibilities over them though there is a whiff of cheese...

 

Isn't the bass line playing an octave too high? If you've simply scored it out on bass clef remember that bass is a transposing instrument and the system needs to play everything an octave lower than written.

 

If you want to minimise the slight darkness that the G chord brings, try juggling the voicing so it sounds higher. Maybe pivot around the 4th in the previous chord to the G's 5th and keep that as the lowest note and play the tonic in the octave above. Of course, if you want to mix things up, try removing the root notes from some of the chords and see how it sounds - I'm a huge fan of such open voicings.

 

Alex

[groovyjazzyfunky]

Nice groove, I really dig the bassline, and I totally understand where you're coming from in writing the bassline first and then fitting chords to it. I do it that way with roughly 1/2 of the songs I write. I think it all depends on what part of the song you envision to be "the hook." That is a rather "hooky" bassline you've got there, it seems obvious that it's what inspired you to begin fleshing out this particular song in the first place, and I don't think there's anything wrong with leaving it as a sort of "centerpiece" of the song and writing everything else to fit around it. Sure, you could do what Jeremy was saying and write the chords first then come up with a cool bass part to fit them, but why should you have to? IMHO, bass doesn't always have to be relegated to "the supporting role."

 

That said, some of your chords there are a bit "iffy." Especially the last bar, starting with G. Your bassline lands on an F# there, so G is going to clash pretty badly with that (at least in this instance, to my ear). And then your bassline goes to a G, so the D chord (D *F#* A) you have there doesn't fit real well with that. You sort of fix it by quickly going to Dsus4 (D G A), but it's still not sounding quite right to me. This is also the case the first time around (bass landing on a G, chord being a D), so maybe forget the regular D altogether and go straight for the Dsus4. Also, I think a DMaj7 might sound better than the D2 at the beginning of your second measure (unless by "D2" you meant DMaj9, which would automatically include the Major 7th, which is a C#, which is a prominent note there in your bass line).

 

Try this:

 

A / Asus4 | DMaj7 / Dsus4 | A / Asus4 | F#m7 / G