How Pros Compose

I've thrown together some posts on old threads that belong together and was hoping to engage in a dialogue about NOT how we compose necessarily but how pros do it. This will not be a discussion of electronic gear.

In submitting this, realize that I'm trying to show off as a newbie on this forum so that oldtimers will know who they're talking to. I've got five stars next to my name over on the music composition forum "Taxi." So, I've been doing this awhile. I'm also a member of 13 other composition forums. Anyway...

How Do Pros Compose?

I once read that Sting sometimes sets the sentences in his morning newspaper to music. This has the added benefit of suggesting possible lyrics once an adequate melody has materialized.

One article was saying that people who have composing careers tend to start with the big picture and work down toward the details. Amateurs do the opposite. Pros are able to conceive of and write several parts simultaneously taking into account how they interrelate.

Amateurs write one bar at a time.

Another point about complexity is instrumentation. Perhaps the reason pop music has only 5 or 6 parts is due to the "not enough instruments=boring, too many instruments=chaos" dichotomy. Pop music has the happy medium.

I was reading a scholarly article about complexity that basically said successfully creative people have personalities that love complexity, so they're able to crank out all this different stuff. However, complexity is not the same as popularity. The Beatles, it mentioned, got less and less popular, the more complex they're music became. The most popular music, they said, tends to have a moderate amount of complexity, not more or less. I read a related article on music perception/psychology which mentioned that well formed rhythms (what people expect to hear) have 2-6 events per 5 seconds (that must be the parameters for a moderate amount of complexity in that musical dimension). The tempo 100 beats per minute is in the center of the perceptual field (moderate complexity?). Scales also have well-formedness. They usually have 5 or 7 notes, not 4 or 6. 5 is a bit simple, the chromatic is a bit complex so, we mostly like 7 even if some of the notes are microtonal but most people are having none of this 31 divisions of the octave stuff etc... that computer music programmers are spewing out.

In Berklees book on melody writing it says the Chorus often starts on a "consonant" sounding place e.g. the downbeat. And ends in a position that feels final. Thats different for a 2 measure phrase than it is for a 4 measure phrase.

In a 2 measure phrase, (if I remember correctly) thats the third beat of the second measure.

In a 4 measure phrase thats the first beat of the 4th measure. I'll check my the book later to be certain.

Of course there are different places where you can end or start a chorus for sure and they all have different feelings of how much rhythmic "consonance" or finality there is.

The second most restful place for a 2 measure phrase to end is the first beat of the second measure etc..

Another point they brought up is that its often important to contrast starting and ending points between the sections. If the chorus starts on the downbeat then the verse shouldn't unless you have a reason for starting on the downbeat again such as its dance music or groove music or your mitigating that with a change of instrumentation or whatever.

You can also contrast melodic rhythms between sections or phrases. Long held notes on the chorus and shorter note values for the verse.

Or contrast phrase lengths within or between sections.

4bar phrase in the chorus 2 bar phrase in the verse or two 2bar phrases followed by a 4 bar phrase in the verse.

These are NOT the only possibilities.

Further, the consequent phrase within a section often reaches a greater height, has a wider leap, gives a greater dynamic or, or, or, etc...

Of course I'm sure you're aware of contrasting melodic outline between sections as well.

If you've got an upward curve in the verse then perhaps the chorus is a flatline or a descending line.

Varying your starting note is another tool. Does the chorus star on 1? Then the verse might start on a less stable note such as 2, 4, 6, 7, b2, #4, b6, b7. This is also affected by what chord is playing in the chorus.

These hints have helped me make better melodies.

In fact another one of my article reads said that pro musicians had a rhythmic concept behind their melody writing and amateurs didn't and only thought about a string of notes. Heck, most music just walks up and down the scale by neighbor notes. That aspect is hardly important most of the time for pop and rock music.

Another observation about classical music which surely applies to pop as well; An article on computer based music composition said that, after analyzing 1000s of classical compositions, the nostrum about hitting the climax only once was an old wives tale. That rule was broken so often that it could hardly be called a rule. The other old wives tale was reversing direction after a leap. Once I tried to follow that rule on everything I wrote and it was all sh*t. Lesson; don't believe everything your told.

Last thing; one technique is that the climax is approached by walking up the scale by step, the climax is a held or a repeated note and then leaps down or visa versa e.g. it leaps up then walks down. an example of the second is the foreign language part of "All Night Long" by Lionel Richie;

Jumbolitae Setemoya. Yeah Jambo Jambo. Way to Party Oh we goin, Oh Jambola. Jambolitae Setemoya (leap up)Yeah Jambo Jambo (Held) Yeah TEXT. All Night Long etc...

I related that experts who study creativity in professional composers said that pros have a rhythmic concept for writing melodies.

I just wanted to point out that experts also have a rhythmic concept for rhythms as well often making the parts played into rhythmic phrases.

In a very old article in Modern Drummer it said that in popular styles, the bass guitar follows the bass drum though won't necessarily be identical. They also said that the keyboard/guitar part follows the high hat. They also went on to say that the high hat part can be extracted from the bass part by using the retrograde rhythm of the bass as accents in the high hat part. This also suggests the guitar and bass part as they mentioned.

When thinking about each aspect of the music. Many musicians think about what they consider to be the best in that domain. Many drummers consider African percussion and the percussion of India to posess the height of excellence. "How is rhythm approachd in those systems?" is often what your favorite percussionists have answers to or wish they did.

In African music, master percussionists will make phrases from the rhythm (as you and I could be doing).

How is this accomplished you ask? The events occuring in the first half of the rhythm can be labeled "a" and the events in the second half can be labeled as "b."

Usually in African music only the permutations abab, abaa or abbb are used primarily because this is all an improvising musician's mind can keep track of. The other thing that they do is to only play half of a rhythm and then complete it much later.

Subfigurations can be added on one beat especially at important structural points primarily on the last beat. All of the possible permutations are possible on the last beat of four or eight measures.

These consist of 4 sixteenths, 3 rhtyhmic events in whatever configuration contrasts with what came previously, two rhythmic events that contrast and one rhythmic event.

Examples,

Possibilities of three rhythmic events at the level of the beat might be one eighth note and two sixteenths, two sixteenths and one eighth, one sixteenth plus an eighth note followed by another sixteenth etc...

Possibilites of two rhythmic events might be one sixteenth rest followed by an eighth note followed by a sixteenth note. A sixteenth rest followed by two sixteenth notes followed by another sixteenth note rest etc....

Possibilites of one rhythmic event might be one sixteenth note rest followed by a dotted eighth.

You get the idea.

Another concept employed as a structuring element is what's called turning the beat around as Gloria Estefan so passionately sings about.

The concept came to Cuban music from Africa.

African music has a bell pattern which has an "on the beat" first half and an "off the beat" second half.

The clave developed from African musicians trying to fit their bell pattern into western rhythms.

As you may well know, there are two kinds of clave 3:2 and 2:3.

The clave never changes but if all the rest of the accompanying instruments drop the last measure or play an extra measure or an odd number such as three (this functions as a transition) then the clave turns around to it's opposite which gives the next section of music a sense of newness.

There are actually several kinds of clave; cuban, brazilian, jazz, etc...

In popular music the two side of the clave is generally dropped and the three side isn't played on the clave. It's often in the bass. But permutations of this are also employed. The three side is a dotted quarter note with an eithth note tied to a quarter and then another quarter has it's own attack.

I'll leave it to you to write the permutations.

The other thing I forgot to mehntion is that in Africa they trade fours as well. In Korea they do this thing where they drop everything except the events in the first beat in one measure then in the next measure there is the first beat and the events occuring within the third beat. Then there's the first, third and fourth beat then in the next measure are all the events. Then they repeat one of the motifs in one or two of the rhythmic events for a measure, then they proceed to shove the rhythms tightly together. It really builds excitement.

Shoving is also used in Middle Eastern music.

If your pieces have choruses in the B section here's a list of possible forms to put it in which are in the pop format ballpark. So that A's =Verses, B's =Choruses, C's =Bridges, D's=make C's another section proper and the D is the bridge, T's=Transitions and R's=a rise to the chorus.

ABCDABCDABCD It doesn't end on the chorus. Good for smooth jazz or real Latin music.

ABCABCABC It doesn't end on the chorus. Might work for a special song.

AABAABCAAB Verses too long for unknowns to sell easily most likely.

ABBABBCABB Double choruses to lengthen a short song.

AABTAABTCAAB Double verses again.

ABBTABBTCABB. Transition after the chorus.

AARBAARBCAARB Rise to the chorus is very common in the postmodern era of popular songwriting. Those double verses plus the rise will kill you though. The exec's arent' gonna wait that long. Next!

ARBBARBBCARBB Rise to a double chorus.

AABTAABCAAB. If repeating the transition everytime makes the song too long leave it out the second time. There's a bridge anyway. Damn double verses again though. Bane of the unsigned!

ABBTABBCABB. No trans the second time. Double chorus.

AABAABTCAAB. Not much point in putting a transition before the bridge except such as where the band drops out 4 bars early while the improvisor solos before the start of the solo section as in jazz.

ABABCAB. This is good if you've got sections longer than 8 bars or a slow tempo. It will prevent you from going over 3.5 minutes.

ABTABTCAB The transition is repeated before the Bridge.

ABABTCAB. transition before the chorus.

ABTABCAB. Another good one for slow tempo or longer sections.

ARBTARBTCARB. Contemporary practice often has the rise to the bridge and a transition somewhere. Why not everywhere?

AARBTAARBTCAARB. Double verse AND a rise before the chorus. Forget it unless you're the Beatles or Elvis.

ARBARBTCARB. transition before the chorus.

ARBTARBCARB This could work. Sound perfectly postmodern.

AARBAARBTCAARB. Double verse and rise. Do you want to sell this song or not?

AARBTAAARBCAARB. Ditto

AABBAABBCAABB. Double verses and choruses. NEVER try more! Otherwise we'll stamp "Rank Amateur" on your forhead permanently.

AABBTAABBTCAABB Double trouble with transitions.

AARBBTAARBBTAARBB. Are you Marilyn Monroe or Princess Di reincarnated? If not, then don't try it.

Lyrical considerations aside, I had the idea, and I'm sure I'm not the first, to take a melody from your favorite song and write it out inverted, retrograde and the retrograde-inversion. This alone produces just kind of stilted melodies in my experience but when you combine them with just two or three other techniques it really works well. Decide which one sounds the best starting on the down beat and call that the chorus.

Then decide where in the form the chorus is going to go, in the A or in the B section. Then, transpose the remaining phrases to starting points below the chorus'. Perhaps so that sections B, C and D's starting notes progress upwards consecutively until they lead to the starting note of the chorus. (The highest notes may also be an important consideration since you probably don't want any other parts to reach higher than the chorus).

Then use rhythmic displacement on these non-chorus melodies so that none of the melodies start in the same place. Displace them forward or backward by either an eighth note or a quarter note.

I think if you did that to a lot of your favorite music or public domain stuff then you'd A) have a lot of melodies and B) find some gems.

Setting lyrics to existing melodies is another problem entirely. Not my specialty (yet).