One of the great things about college a cappella groups is that they attract people who have a love of music regardless of whether they have any formal training. Some of the people I’ve most enjoyed listening to have done most of their singing in the shower. Inevitably though, written music is essential for communicating musical ideas, and so a natural conflict arises between people’s desire to contribute creatively and their ability to coordinate the group to sing what they are thinking.
Thankfully a little guidance and music notation software puts arranging within the reach of even those who are musically illiterate. Having an instant playback of whatever you put down is invaluable. I know enough notation to get by, but I’ve never been good at reading it. In this text I intend to write up everything that worked for me over the years producing arrangements that we enjoyed singing. I originally wrote this for my group at Carnegie Mellon, so excuse the physics diversions and the in-jokes here and there.
It’s a great and soulful experience to sing with a group of friends, and it’s even better if they are singing something you wrote. Don’t be intimidated! Dive in!
Listening and Picking out Notes
Picking out notes is hard and a lot of the difficulty comes from how our ears and brains hear pitch. This will get a little technical, but knowing this will help you understand the ways that you can mishear a note, and how to correct for it.
Pitch can be defined as the sensation that our brains create when interpreting repeating patterns of sound. Contrary to what you might think, there is no physical phenomenon that directly maps to pitch. There is a rough correspondence between the frequency of a sound wave and pitch, which is what most people associate with it. Higher frequencies sound like higher pitches, and lower frequencies sound like lower pitches. The A above middle C for instance, corresponds to a pattern of sound that repeats at 440 Hz. However, if you hear a person sing an A, there is much more to the sound than just a pure tone at 440 Hz. The folds of your throat and the shape of your mouth create resonances at other frequencies that are called harmonics. This is what allows us to hear the difference between a human, and say, a flute, even if they are playing the same note perfectly in tune. The shapes and materials of the instruments color the sound by adding harmonics to it. These effects are referred to as the “timbre” of the note.
Interestingly, studies have shown that these harmonics have more to do with our perception of the pitch of a sound than does the presence of the root frequency of the note. It’s possible to correctly hear a note as the A above middle C even when the 440 Hz frequency isn’t in the sound at all! When an instrument is resonating at a certain frequency, the harmonics that it produces have a consistent relationship with that frequency. As a result, even if you can’t hear the original frequency in the sound, your brain can still determine the pitch of the note. In a sense, the harmonics “point” to a certain root frequency that produced them. What we perceive as pitch is our brain’s guess as to the frequency that best explains the pattern of harmonics. When we hear more than one note at a time, the ways in which these harmonics interact can play with our perception of pitch in counterintuitive ways.
When chords are perfectly in tune, it is difficult to mentally separate them into distinct notes. When two notes are in tune, their harmonics are also in tune, and this can confuse your brain’s pitch detection apparatus by pointing to a different note, or causing your brain to assume that one note is just a resonance of the other. This is particularly true when the notes are played by the same instrument, or when they are in a pop song where everything is blurred together into a robust sound soup. In particular, it is very easy to confuse a note with one that is an octave or a fifth apart from it. Sometimes, you can even convince yourself that what you are hearing is a single note instead of a chord. If the notes you are writing aren’t sounding enough like the original song, look out for this. Make sure that everything you thought was a single note actually was, and make sure that the notes you wrote aren’t a fifth or a third off from the actual notes in the song.
You can observe this effect quite a bit in hard rock or heavy metal music. The most commonly used chords in this type of music are known as “power chords,” which consist of a root note, the fifth above it, and the root note again an octave up. When you put these chords through some heavy distortion, the result ends up sounding like a beefier version of the root note instead of 3 distinct notes. Listen to Nirvana’s “Smells like Teen Spirit ” and pay attention to the parts of the song where the distortion is on strong. You can hear the 3 distinct notes in each chord when he plays them at the very beginning without distortion, but after he turns on his fuzz box they blend into a single beefy note.
When you hear a loud sound, your ability to hear soft sounds happening during the loud sound is diminished. This effect is exploited by audio formats like mp3 to achieve their small file size. When a loud sound is occurring, the encoder stores the quieter sounds with less precision, because when you are listening to it, you won’t be able to tell the difference. However, this makes compressed audio formats worse to use for arranging because it can become difficult to distinguish the background melodies that make up the body of the music.
It is harder to determine the pitch of short notes than long notes. This is actually a fundamental property of all waves as any quantum physicist will tell you, but it has particular implications when arranging. On your first attempt at transcribing a part it is easy to put in transitional notes incorrectly. Because the notes are short, there is a good chance you won’t notice the problem when you listen to the arrangement. You may find yourself thinking that something doesn’t sound right but you can’t figure out exactly what is wrong. If this happens, double check those notes.
- GET A HIGH QUALITY ENCODING OF THE SONG!!!! Seriously, it warrants that many exclamation points. If what you are working with is a 128kbps mp3, you are going to have a much harder time than if you are working with the original CD, or a high quality encoding. Music formats like mp3 compress the audio into a smaller file by storing the subtle harmonics with less precision, which are exactly what you need if you are trying to precisely determine pitch. If you ever try to arrange a song from a wma file, you are likely to get an aneurysm.
- Get a good pair of headphones. It will make your life simpler, and will keep you on good terms with your roommate after you’ve listened to the same 3 seconds of music for the 47th time.
- Previous notes can be good cues for future notes. If you are trying to hear the note an instrument is playing at one point, start listening a few seconds before that point, and focus on that instrument. Trace the notes through in your head. Try to hear the part in your head independent of the rest of the piece.
- Find tablature, sheet music, or a midi. Beware: if you find these things online, most of the people making them aren’t any better at it than you and are probably worse. Furthermore, they probably weren’t intending the tracks of a midi to be sung by human beings and didn’t set it up to make good vocal parts. You are best served by treating these resources as other people’s ideas and suggestions about the song rather than an exact transcription. For instance, I’ve been trying to learn to play the U2 song “One” on guitar off and on for a number of years, and I have yet to find a tablature online that sounds remotely like the song other than having the same key and basic rhythm.
- Find sections of the song where parts you are interested in play by themselves. Often as a song progresses some of the background parts will be featured without other layers. Use these to figure out parts you are having trouble with elsewhere in the song.
- Find a live version of the song. Usually live recordings are mixed differently than studio recordings, and this can be a big help for listening to parts that are difficult to hear in the studio version. If it is a band that improvs in their live sets, sometimes you can also get good ideas for embellishments to add to the parts.
Figuring Out the Key
If you have a real knowledge of music theory, ignore this part and do it however it is really supposed to be done, or if the key signature is immediately obvious, put it in. If you are like me however, and don’t have any real knowledge of music theory, keep reading.
- Choose whether you are going to use sharps or flats for the song. Use only whichever accidental you pick.
- Start arranging, adding in accidentals when you need to.
- You may get to a point where you need to use a sharp or a flat and then negate it with a natural one or two notes later, or where you need to insert a sharp or flat in the middle of a measure where you’ve already inserted the natural version of that note. This isn’t a problem at the moment, but you can see how it will get in the way of giving the piece a key signature later. There are usually two possible causes of this:One of the possibilities is that you chose the wrong type of accidental 2 steps ago. You don’t need to change this immediately because finale’s key changing functionality is fairly smart, but this will narrow down the possibilities of keys for your song. Take note of it and keep going.The other possibility is that the notes you are working with are part of a chromatic scale. This is characterized by a series of notes that are all a half step apart. These happen fairly often in rock and jazz, but particularly often in blues. You’ll often see 3 notes that are a half step apart, and the middle one will be a “color” note. If the notes are part of a chromatic scale the key signature will never be able to remove all of the accidentals, so keep that in mind when it comes time to pick the key signature.
- Once you’ve gotten a significant theme for your song arranged, you can probably pick the key signature fairly accurately. The goal here is to pick one that minimizes the number of accidentals in your piece. Finale has a key change tool that you can use to test out the effects of different key signatures. Be careful when you use it however, because by default it will transpose your notes into the new key, rather than holding them to their original pitches and imposing the key on it. When you change the key of the song, be sure to select “Hold notes to original pitches: Enharmonically.” If you do this a lot you might see notes marked with oddly shaped X’s in your song. This indicates a double sharp if you haven’t seen it before. Don’t worry about these. You can clean them up by erasing the accidental, and raising the note up a whole step. These are the result of Finale making a mistake in thinking you actually knew what you were doing when you wrote the note in the first place. The same thing applies for notes with two flats next to them.
- Some songs, particularly rock songs, don’t have a well defined key. In particular, it is common to have a minor pentatonic bass line and a major key guitar part. You can recognize this if you are finding it difficult to choose between having the song in the minor or major versions of the same key, or between a major key, and the major counterpart of its relative minor. (A major and C major, for instance) To make things even more confusing, sometimes the guitar will join in with the bass line and emphasize the minor pentatonic part and go back to the major key a measure later. Take a look at Cake’s “Short Skirt/Long Jacket”, or U2’s “Mysterious Ways”, for good examples of this.
- Some songs don’t even really pay lip service to a traditional key. Bemoan the loss and do what you can. These are rare however, so don’t give up too quickly.
Composition/ Assigning Parts
When you are arranging a song you have a minimum of two goals: you want the song to be fun to sing, and you want the song to be fun to listen to. These goals are more independent from one another than you would think. It’s possible to write an arrangement that sounds great to the audience but has individual parts with no continuity or direction. It is a bit harder to write a piece that is fun to sing but sounds bad to the audience, because fortunately we can hear ourselves.
- Basses are comfortable between a G and an A. You can stretch that two or three notes in either direction. Much below G it will be inaudible, and higher than a C will be falsetto and we’ll probably hate you.
- Tenors are comfortable between a D and an E. Historically the ranges of our tenors vary a lot, so ask them what works.
- Altos are comfortable between an A and a C.
- Sopranos are comfortable between a C and an E. You can push the low end down to an A, but don’t count on volume in those areas. On the high end, you can use notes up to a G or A, but don’t count on them blending with the group, and they aren’t going to be quiet. Think screeching birds diving from the heavens.
- Zavos are comfortable with any note you want. Unleash the Beast.
- It is easy to sing high notes loudly and low notes softly. If you want something loud don’t put it anywhere near the bottom of the range for that voice part. If you want something quiet don’t put it anywhere near the top of the range for that voice part. This second rule doesn’t apply as stringently to male parts as it does to female parts, because guys have the option of singing very high notes in falsetto which can be soft more easily.
- Staying in half of a voice part’s range for an entire song usually makes the song no fun to sing.
- Complex chords in low notes tend to sound muddy. This is generally why the basses carry the root notes of chords, and rarely split into multiple parts. This isn’t a strict rule, and you can produce neat things when you violate it, but think about what it will sound like first.
- It’s easy to find yourself getting wrapped up in working on the notes of a piece and forgetting the syllables until it comes time to print it out and you realize that no one will know what to sing. If two thirds of an arrangement are made up of the rhythm and the pitch, the syllables are another full third. Good syllables can give an arrangement its polish. Strong syllables make an arrangement easier to remember and sound more cohesive and professional. “Sing the part with the jumjiggidows,” is much faster in rehearsal than, “ok, bottom of page 3 third measure where the chorus. . . is that the chorus? . . . yeah, where the first chorus starts . . .” Strong syllables will make your arrangement equally more memorable for the audience.
- Think of syllables as a way to shape the sound. You’ve put in the notes and the rhythm, and now you get to sculpt them.
- Unlike rhythm and pitch, the syllables have only a small chance of sounding or being “wrong” with respect to the original song, so you have a lot of freedom. The instruments in the original song can provide good cues however. Pay particular attention to where they begin and stop sharply, and where they trail off.
- Different vowel sounds naturally promote different volumes depending on how open your mouth is when you make them. You can use this to change the balance of the parts or to emphasize the dynamics in the song. “Ah”s are the loudest I can think of, “uh”s are a bit softer, “oh”s are softer still, and “oo”s are the quietest. You can also use this to create rhythmic patterns in the volume.
- When you want a well defined end to a note, but in a consonant at the end of it. “m”s are good for a soft stop, “n”s are slightly harder, and syllables that end with “p” invariably become staccato.
- “l” and “j” are underrated consonants.
- When assigning syllables to rapid notes, you can make singing them less taxing by alternating consonant sounds that are made in different parts of the mouth. For instance, “na-ma-na-ma” is easier to say than “na-na-na-na” or “ma-ma-ma-ma,” and “da-ba-da-ba” is easier to say than “da-da-da-da” or “ba-ba-ba-ba.” Sometimes the composition really requires syllables with an exactly consistent sound, but think about whether you really need it, because alternating the syllables can make it much simpler to sing without changing how it sounds all that much.
- Echoing lyrics in the syllables can produce cool effects.
- You absolutely must include your name somewhere in the syllables for your song. Otherwise it can never be truly yours.
- Know the song backwards and forwards, or in other words, arrange songs you like. A good exercise for familiarizing yourself with the song is to listen to it several times, and each time pick a particular instrument to follow throughout it. Sing along with that part. This will make it much easier to assign parts without getting stuck awkwardly halfway through, because you will intuitively know what themes in the song are continuous enough to make good vocal parts before you even start arranging. The better you know a song the more you’ll understand what makes it work musically, what parts bring out the changes in mood, what notes are essential, and you’ll be comfortable enough with it to adapt it and really transform it.
- Think of every part both as its own melody and as a complement to the other parts.
- Learn every part. Sing them in your own octave. This ensures that all of the parts are comfortably singable, and that when people have questions in rehearsal, you’ll know what to tell them. You’ll find that you intuitively have a sense of melody, and you’ll be able to spot problematic areas easily.
- A good way to start arranging a song is to pick out the different instruments, and pick one to assign to each voice part. You’ll find however that most instruments have a wider range than most singers, so often you will have to reassign parts along the way. Instruments can also produce sequences of notes that humans can’t. A complicated arpeggio can be trivial for a guitar player because each string can operate somewhat independently. Humans unfortunately don’t have 5 independent sets of vocal chords (as awesome as that would be.) If you encounter parts like this, try to come up with a good way to split the notes between parts in a way that is singable and melodic, even though the rhythms may get complicated.
- When you are arranging a song, think about the voices we have in the group. You probably know what the various voice parts sound like. Don’t arrange your song with some idealized hypothetical a cappella group in mind. Think specifically of the people who are singing it. Make sure that there is at least one person that could pull off the solo, and make sure that all of the parts will sound right with the people we have.
- Play the parts off of each other, and don’t forget the soloist.
- Think about the direction of each part of the song.
- Breaking from a syncopated rhythm to a straight rhythm or vice versa is a good way to introduce a dramatic transition. If every part has a separate rhythm, bringing them together onto the same rhythm can have the same effect.
- When you establish a pattern in an arrangement, don’t break from it where you don’t need to. When you do break from it, have everyone break from it at the same time. The hardest songs to learn are those that change in unpredictable places.
- If a song repeats the same chord progressions that doesn’t necessarily mean that your arrangement needs to repeat. If a song repeats the same structure on each verse, consider making the first verse close to the song and take liberties with the rest. That said, do try to establish patterns. Your computer is a lot better at remembering what you’ve written than we are.
- If you can accurately figure out the chord progression of the song you’ll have a much easier time taking liberties with it, because you’ll know what notes have to be there to make it sound right. Tablature can be a good resource for this, but see my earlier caveat about getting things online.
- Ask for feedback on your arrangement. Get people to look at it, see if people like it. Positive feedback is boring. Negative feedback is good. I’ve found that people have to like something to feel that it is worth their effort to constructively criticize.
- Keep working on it. If you don’t have a deadline for it, open it up periodically and tweak parts. Polish it.
- Be nice.
- Have fun. 🙂