Friday, May 26, 2006

The Wall-of-Sound Explained: Part 6

As a producer, maybe you don't plan to write songs. But you still need songwriting skills. When do you bring in certain instruments? What if you need to write a tiny melody to link two sections, or to fill unused space? These are all artistic decisions every bit as important as the composition of words and melodies. Both Phil Spector and Brian Wilson were songwriters and multi-instrumentalists.

What makes a good song? Lots of self-help books for songwriters have tried to answer that question. They talk about following A-B-A structures and keeping scrapbooks of ideas. Some even draw the notes out on a musical staff. My opinion is, those books waste your money.

The best way to improve your songwriting is to expand your musical vocabulary by listening to a lot of good music, and listening to them over and over again. There is no getting around this. It is the same as learning a language. You have to get to a point where you are so comfortable with melodies and chords that you can pull them out of the air. A snippet of melody from this song, a verse resolution inspired by another song... that's what the best songwriters have on their musical palettes.

I can't tell you what specific records to buy, but I will advise you to pay more attention to the music of the past. Don't take them for granted just because you heard them during your childhood in your parent's car radio. Listen to the records again, take them off your parents shelves if you have to, and I think you might be pleasantly surprised.

As for modern music, do not be too influenced by what's happening today, because you don't want to participate in a fad or a trend. Don't worry about getting on the radio. Renoir didn't try to make his paintings look good next to Monet's and Cezanne's. Making it in the music industry is a concern only for people whose music is borderline acceptable to the masses, and then they make a lot of noise about why the mainstream music industry ignores real musicians. The truth is, if you make a good, big song worthy of the world, people will hear it.

Now we have talked about some general ideas about songwriting. What about some specifics? Let's begin with some characteristics of good and bad pop songs.

The good pop song is fun for the listener. Unlike a bad pop song, it is never boring. It is not obnoxious. It does not waste the listener's time. Melodies are not slow and drawn out, like some Radiohead songs. A pop recording is not supposed to showcase a singer's voice. It is supposed to make people dance, make people sing along.

For instance, "Help Me Rhonda" is a great song to sing along to, because there are so many vocal parts. If you don't want to sing the lead vocals, there's a great harmony line with the equally-fun-to-sing "bow-wow-wow" part during the chorus. In contrast, "Exit Music (For A Film)" is not fun to sing along to. Too many whole notes. Too many empty spaces. Too slow. That's why it's not a hit single, despite being an otherwise well-written and melodic song.

At this point, I would like to say that I am not accusing bands like Radiohead of being bad musicians. Remember, this is an essay about pop music. There are many other styles of music out there, and I generally have nothing against them, even experimental music, as long as the artist is making a serious and sincere effort to be creative, considerate and meaningful. We should realize, too, that listeners do not always want to hear pop music. If you are in a bad mood, you might rather listen to Kid A than "Da Doo Ron Ron."

I am however against music that is unexciting either because it is formulaic or because it is overindulgent by the artist. A lot of classic rock and 70s pop falls into this category. Here is a list of things that waste a listener's time:
  • A line of melody repeated more than 3 times. This can take the form of a very repetitive chorus, or a verse that keeps using the same melody, such as the Raconteurs' "Steady As She Goes."
  • A melody of only one or two pitches that lasts more than one measure.
  • A guitar solo that lasts for more than 20 seconds, unless there are vocal harmonies during the interval, or if the guitar is playing a singable melody
  • Chords being played by a distorted guitar with nothing else happening except bass and drums. Contrary to popular belief, a distorted guitar does not necessarily project strength. The best way to project strength at any volume is to have multiple instruments playing the same thing. A hundred violins playing in unison is more powerful than Jimi Hendrix. Keep in mind that most people listen to music at a quiet volume, and the distorted guitar sounds pathetic when quiet.
  • The instrumental arrangement does not change throughout the song (a common flaw on Smiths songs)
  • Anything that could be described as "jamming," unless you are Bob Marley
  • Anything that could be described by the words "ethereal," "atmospheric," "hypnotic," or "deep." The point is to make people dance, not to make them contemplate life and fall asleep.
  • You plan to impress your audience by a "shock and awe" technique of producing chaos and cacophony, even if it's melodic.
  • The rhythm section is repetitive (e.g. the Postal Service, Death Cab for Cutie, Broken Social Scene). However, in certain rare occasions it is okay to have a repetitive backing track (Smashing Pumpkins' "1979").
  • Song length is more than 3 min 30 seconds. (but really, you should aim for 2 min 30 seconds.)
Regarding the last point, it is usually okay to create an epic song if it is under six minutes and contains more than three very distinct movements. A good example is "Bohemian Rhapsody." It is not okay to create an epic song of the same length that follows a predictable buildup to a climax. This is why it is a truth universally known and accepted that "Stairway To Heaven" sucks.

Does your song suck? A good way to find out is to listen to your song with a (sober) friend who's never heard it before. You will immediately be much more self-conscious about your work, and it will be easy for you to sense where the boring, even embarrassing, sections are. Of course, there is also the benefit of receiving much valuable feedback from your friend.

A good sign that you have a hit song is if your friend can remember the hook a few minutes after listening to it. Always strive to make an impression the first time. Here is a list of more things that make a song exciting for the listener:
  • The song changes about 3/4 of the way through (something to look forward to)
  • A single, unrepeated hook (for example, the link segments on Weezer's "Buddy Holly" and the Beach Boys' "God Only Knows"). It makes people go back to the beginning of the song when they are disappointed not to hear the hook again.
  • Drums that drop out and come back in
  • An exotic instrument (such as the theremin on "Good Vibrations" and the synthesizer on Del Shannon's "Runaway", but be careful about using faddish ethnic instruments such as the sitar)
  • A melody that spans more than one octave
Over the course of your career, you will discover more things that are exciting to people.

So now you have a good idea of what makes a good song. How do you put it all together? In the next section, we will talk about the best instrumental arrangements for the wall-of-sound.

1 comment:

Anonymous said...

Be careful of formulas, dude. The best art tends to violate, rather than adhere to, any kind of rigid formula -- that's why the exceptions you note (like 197A9) are some of the best songs. Mahler used "hundreds of violins" - does that really make him "more powerful" than Hendrix in say, "All along the watchtower"?

-AK