## 18. Arranging and Instrumentation

Arranging is the decision about who plays what and where: which band of the pitch range each part occupies, how many independent lines are in the air at once, which notes are worth sounding twice, how a handful of players is laid out across a chord, and what each family of instruments can and cannot reach.

Source: https://musictheory.center/course/orchestration
Level: Advanced · Applied · 6 sections
Before this: Voice Leading, Form and Structure

### Register and Range

https://musictheory.center/course/orchestration#register-and-range

> Play the first three. Each sounds the same three note names together, and nothing about which notes they are changes from one clip to the next — only how far up the range they sit. Play them in order and say which of the three is hardest to pick three separate notes out of, and which sounds thinnest. Then play the fourth: five notes, a silence, and the same five notes again higher up. Ask yourself what is identical in the two halves and what is not.

**Plays:** Three notes together (C2 E2 G2) · The same three names, further up (C4 E4 G4) · The same three names, further up again (C6 E6 G6) · Five notes, then the same five higher (C4 D4 E4 G4 C5 D5 E5 G5)

The staff draws the second of the four clips: C, E and G sounded together, the C being middle C. The first clip is those same three names two octaves lower and the third is those same three names two octaves higher, and this page draws only a treble staff, so the first would need eight ledger lines below it and the third four above — which is why the rest of this section names bands instead of counting lines. The table does that naming. Read its last column down and the pattern is that what changes as you climb is not which notes are available but what they are any good for: weight at the bottom, tune in the middle, sparkle at the top.

- **The three chords are the same three note names** — The first clip sounds C2, E2 and G2 together; the second C4, E4 and G4; the third C6, E6 and G6. Each is two octaves above the one before it, and all three are [C](https://musictheory.center/chord/c-maj).
- **The tune moves; the distances inside it do not** — The fourth clip plays C4, D4, E4, G4, E4, then a silence, then C5, D5, E5, G5, E5. Every interval inside the tune is identical in the two halves. Only the pitch the whole thing sits at is different, and that is the only thing you are being asked to hear.

| Band | Roughly | What it carries well |
| --- | --- | --- |
| Sub-bass | C1 up to C2 — about 33 to 65 Hz | Weight and size. Nothing down here is followed as a tune; it is felt as much as heard, and a second part sounding with the first simply thickens it |
| Bass | C2 up to C3 | Chord roots and the bass line. One part at a time down here reads clearly; two parts close together turn into a blur |
| Low-mid, or tenor | C3 up to C4 | Inner harmony and the warm end of a melody. Chords fill out here without going muddy, which is why keyboard left hands live in it |
| Mid, or alto | C4 up to C5 | Most melodies and most singing. This is the band a listener attends to first, and the one every part in an arrangement is competing for |
| High, or soprano | C5 up to C6 | A tune that has to be heard over everything else, and counter-lines placed deliberately above the tune |
| Brilliance | C6 up to C8 | Attack, air and sparkle. Few parts have their fundamental this high; what usually lives up here is the upper partials of notes sounding lower down |

**register**

**Definition.** A register is a band of the pitch range treated as a region — the notes lying between one stated pitch and another, considered together for what they can carry rather than for which notes they happen to be.

Chapter 2 introduced the word loosely — the part of an instrument’s low-to-high reach a sound comes from — and let **range** stand in for it whenever the whole of that reach was meant. Everyday talk still does, and nothing here says that usage is wrong. An arranger needs the tighter version, because the whole point is to name a band without naming an instrument. Used strictly, then: a **range** is the whole span something can produce, from its lowest note to its highest, which is what Chapter 13 measured a [melody](https://musictheory.center/course/melody#what-makes-a-melody) by — a cello’s range runs from C2 to about C6. A **register** is a slice of the pitch range itself, fixed at stated pitches, so the same slice means the same thing whoever is playing in it: the bass band from C2 to C3 is the bass band for a cello, a trombone and a bass guitar alike. And a part’s **tessitura**, the third word Chapter 13 introduced, is the slice it actually spends its time in — a fact about a piece of writing rather than about the instrument, so that a cello part sitting mostly between G3 and G4 has a low-mid tessitura whatever the instrument could reach if asked.

The three chords at the top of this section are the same chord. The lowest of them is [C](https://musictheory.center/chord/c-maj) exactly as much as the highest is, and no note in any of them is different from the corresponding note in the others by anything but [octaves](https://musictheory.center/course/pitch-and-notes#the-octave). What changes is how easy the chord is to hear as three things. Low down the three notes crowd into each other and the result reads as one thick sound; higher up they separate, and higher still the chord thins out until it is more colour than harmony. An arranger who moves a chord an octave has therefore not made a small decision, and the reason is acoustic rather than stylistic.

The working use of the word is as an address. Saying that a part sits in the bass register says where it is and therefore what job it can do, without naming a note and without naming an instrument, and two people can agree on it across a piano, a bass guitar and a trombone. Every later section of this chapter is a question about that address: which band a line occupies, how many lines are in one band at once, and what happens when two of them want the same one.

**Good to know**

- (Fact) A note’s frequency doubles with every octave, so the same interval spans a different number of hertz in different registers. Middle C is about 262 Hz, which puts C2 at about 65 Hz and C6 at about 1047 Hz. A major third is a major third everywhere — but C2 to E2 is a gap of about 17 Hz while C6 to E6 is a gap of about 272 Hz. The ear separates two tones only when the gap between them is wide enough relative to the region of the inner ear they both excite, the **critical band**, and that band is roughly 100 Hz wide at the bottom of hearing and widens far more slowly than frequency climbs. A third low down therefore falls inside one critical band and is heard as roughness; the same third two octaves up falls well outside one and is heard as two notes.
- (Convention) The band names in the table are labels, not measurements, and the labels differ by trade. Recording and mixing use sub-bass, bass, low-mid, mid, high and brilliance, with the boundaries drawn wherever a given desk or a given writer draws them. Choral and orchestral writing uses bass, tenor, alto and soprano for overlapping bands, borrowing the names of the four voice parts. Nothing in the sound marks a boundary: the bands are a shared vocabulary for pointing at regions, and two people using different sets of names are not disagreeing about anything.

**You already know this**

- The octave from Chapter 2 — the distance at which a note recurs with the same letter name and its frequency has doubled — which is the unit every move between the bands above is measured in — https://musictheory.center/course/pitch-and-notes#the-octave
- Register from Chapter 2 — the part of the low-to-high span a sound comes from, whether it sits near the bottom of an instrument’s reach, near the top or somewhere in the middle — which this section pins to stated pitches, so that two people can name the same band without either of them naming an instrument — https://musictheory.center/course/pitch-and-notes#high-and-low
- Voice leading from Chapter 12 — reading a passage as several lines running left to right, of which a chord is only a vertical slice — because it is the individual lines, and not the chords, that each need somewhere of their own to sit — https://musictheory.center/course/voice-leading#voices-not-chords

**In real music**

- Edvard Grieg — “In the Hall of the Mountain King” — the theme is stated at the very opening by the bassoons and low strings, about as low as the piece goes, and comes back again and again as the music builds, higher and with more players each time. Follow one statement of the tune to the next and almost nothing about the notes changes — what changes is where they sit and who is playing them

**Try it on your instrument:** Play one chord three times on your instrument or in your DAW, using the same three note names each time: once as low as the instrument comfortably reaches, once around the middle of its range, and once as high as you can still count three separate notes in it. Then play a five-note tune twice, moving every note of the second pass up twelve half steps. Write one sentence about what the lowest version of the chord was good for and one about what the highest version was good for.

**Check:** Three notes sounded together, once. Decide where they sat.

- Around the middle — at or just above middle C **(correct)** — Correct. The three notes were D4, F4 and A4 — a D minor chord sitting directly above middle C, in the mid band of the table. The lowest of the three was one whole step above middle C and the highest was a major sixth above it.
- Low down — at the bottom of a bass guitar — The bottom of a bass guitar is around E1, nearly three octaves below the lowest note you heard. The lowest note in this clip was D4, a whole step above middle C; nothing in it came near the bass band at all.
- High up — two octaves above middle C — Two octaves above middle C is C6, where the third chip at the top of this section sits. This clip topped out at A4, a major sixth above middle C — a minor third short of the high band, which starts at C5, and more than an octave short of the brilliance band at C6.
- Spread apart — one note low, one middle, one high — All three notes were inside a single perfect fifth: D4 at the bottom, then F4, then A4 at the top. A spread setting would have put an octave or more between at least two of them, and there is no gap of that size anywhere in this chord.

### Texture

https://musictheory.center/course/orchestration#texture

> Play the first: seven single notes, one after another, and nothing else. Play the second: the same seven notes, with three more notes struck underneath every one of them at the same moment. Play the third: the same seven notes again, with one note underneath each. Play the second and the third back to back and ask how many things you were following in each — and in the one where the answer is more than one, which of them you would call the tune.

**Plays:** One line, and nothing else (E4 F4 G4 A4) · The same line, with three notes under each of its notes (C3 E3 G3 E4 D3 F3 A3 F4 G2 B3 G4 F2 A4 C4) · The same line, with one note under each of its notes (C3 E4 A2 F4 G2 G4 F2 A4)

Nothing is drawn here. The difference between the three clips is not a difference in which pitches are printed — all three have the same seven notes on top, and a staff would show that and hide everything else. What differs is how many independent shapes are in the air and whether any one of them is in charge, and the table is the better picture of that. One warning before you read it: the notes underneath the tune in the second clip are struck again on every step rather than left to ring, which is how this clip is written and not something forced on it. That is not a limitation here. Chords restruck in the tune’s own rhythm is precisely what the second row of the table calls chorale style.

- **What the second clip plays** — On top, E4, F4, G4, A4, G4, F4, E4. Underneath, three notes to a step: [C](https://musictheory.center/chord/c-maj), then [Dm](https://musictheory.center/chord/d-m), then [G](https://musictheory.center/chord/g-maj), then [F](https://musictheory.center/chord/f-maj), then [C](https://musictheory.center/chord/c-maj), [Dm](https://musictheory.center/chord/d-m) and [C](https://musictheory.center/chord/c-maj) again. They change when the tune changes and never between, and every note of the tune belongs to the chord underneath it — in the fifth step the tune supplies the chord’s fifth itself, the three notes below it being C3, E3 and C4.
- **What the third clip plays** — The same seven notes on top, and underneath them C3, A2, G2, F2, G2, A2, C3. The upper line rises and falls; the lower line falls and rises. They go opposite ways at every single step, which is what keeps the ear able to hold both at once.

| Texture | What is going on | Heard in |
| --- | --- | --- |
| Monophony | One line and nothing else. Any number of performers, so long as every one of them is on the same notes | Gregorian chant; an unaccompanied folk song; a whistled tune |
| Homophony | One line carries the tune and everything else supports it. The support may move in the tune’s own rhythm — chorale style — or have a rhythm of its own, which is what "melody and accompaniment" names | A four-part hymn; a singer with a strummed guitar; very nearly all popular music |
| Polyphony | Two or more lines sounding at once, each with its own shape, and usually its own rhythm, none of them merely supporting another | A Bach fugue; a round such as "Frère Jacques" sung in parts; the front line of a New Orleans jazz band |
| Heterophony | Several performers on one line at once, each ornamenting or timing it a little differently, so that a single melody is heard with a blurred edge rather than several melodies | Javanese and Balinese gamelan; an Irish session where every player takes the same tune their own way |

**texture**

**Definition.** Texture is the relationship between the things sounding at once in a passage: how many independent lines there are, whether they move together or go their own ways, and which of them the listener is meant to follow.

The four names in the table are the standard classification, and they answer one question each. Monophony: is there only one line? Homophony: is one line in charge and the rest in support? Polyphony: are there several lines, each with a life of its own? Heterophony: is there one line being played several ways at once? A near-synonym worth keeping separate is **density**, which is how much is going on — three parts or thirty. Density and texture vary independently: a thirty-piece string orchestra playing a single line in unison is dense and monophonic, and two recorders playing a round are thin and polyphonic.

What decides which of the four a listener actually hears is not the label an arranger had in mind. Notes that begin at the same instant and move in the same direction fuse into one perceived sound; notes that begin at different instants or move in different directions separate into distinct streams the ear can follow apart. That is why the second clip above reads as one tune with support — its lower notes move at exactly the moments the tune moves, and never otherwise — and why the third reads as two, having given its lower line the opposite direction at every step. **Texture is made by rhythm and direction at least as much as by how many notes are present.**

Naming a texture is describing a moment, not a whole piece. Most music changes texture constantly: a verse that is [melody](https://musictheory.center/course/melody#what-makes-a-melody) and accompaniment, a chorus where backing vocals move in the tune’s rhythm, an instrumental break where two lines chase each other. The practical use of the four names is as a question to ask about a bar you are writing — how many things am I asking to be followed here, and have I made that possible — rather than as a box to put a piece in.

**Good to know**

- (Fact) The ear groups sound into streams using onset and motion. Tones that start together and change together are heard as one event with a composite colour; tones that start at different moments, or move in different directions, or lie far apart in pitch are heard as separate streams that can be attended to one at a time. This is auditory stream segregation, and it is a property of hearing rather than of any style, notation or instrument — which is why the same four written parts can be heard as one thing or as four depending only on how their rhythms line up.
- (Convention) Monophony, homophony, polyphony and heterophony are classroom categories built from Greek roots, and real music sits between them constantly. A hymn with one voice briefly breaking into its own line is neither purely homophonic nor properly polyphonic, and no analysis is improved by forcing the choice. The names are useful for pointing at what a passage is doing; they are not a taxonomy the music was written to satisfy.
- (Tendency — Western art music since roughly 1600 and the popular music descended from it. The century before that was built on polyphony, where no line was meant to be the tune, and heterophonic and polyphonic traditions elsewhere have never had this default at all) Homophony is the default: one line carries the tune, and everything else exists to support it. A listener who is given several lines with no obvious hierarchy will usually take the highest as the tune until something makes them do otherwise, and an arrangement that wants to be understood on one hearing works with that rather than against it.

**You already know this**

- The melody from Chapter 13 — a succession of single notes the ear follows as one thing, with a shape of its own — which is the line all three clips above have in common and the thing every texture is a texture around — https://musictheory.center/course/melody#what-makes-a-melody
- Smooth voice leading from Chapter 12 — moving each part as little as the next chord allows, so that every part keeps a path of its own — which is what makes several lines followable at once rather than merely simultaneous — https://musictheory.center/course/voice-leading#smooth-motion

**In real music**

- Queen — “Bohemian Rhapsody” — the first twenty seconds, before the piano comes in: several vocal parts, all of them changing note at the same moments, with the top one carrying the words you follow. Try to pick out a second tune underneath and there is none — everything below the top line moves when it moves
- Johann Sebastian Bach — “Fugue in G minor, BWV 578” — the first thirty seconds: one line completely alone, then a second entering underneath with the same tune while the first carries on with something different. Count how many separate lines you can still follow a minute in, and notice that no one of them is the accompaniment

**Try it on your instrument:** Play a short tune on your instrument or in your DAW three times. First on its own. Second with a chord struck under every note of it, changing exactly when the tune changes and at no other moment. Third with a single second line underneath that goes down where the tune goes up and up where the tune goes down. Play all three to somebody else and ask them, after each one, how many things they were listening to. Write down where their answer differed from what you intended.

**Check:** Five notes on top and more notes underneath, played once. Decide what the relationship between them is.

- One line with supporting notes moving in step with it **(correct)** — Correct. The tune was G4, F4, E4, D4, C4 descending, and under each of its five notes three more were struck at the same instant: C, E and G under the first; D, F and A under the second; C, E and G under the third; G, B and D under the fourth; C, E and G under the last. The lower notes change when the tune changes and never between, so the ear hears one tune with support rather than several tunes.
- One line and nothing else — Four notes sounded at every step, not one. Play the first chip at the top of this section, which is a tune with nothing underneath it at all, and then this clip — the difference is exactly what a single line does not have.
- Two or more lines, each with its own shape — Two independent lines would each have a shape of their own — the third chip at the top of this section has a lower line that goes down wherever the tune goes up. Here the three lower notes move as a block, in the tune’s direction and only when the tune moves, so there is nothing underneath with a shape of its own to follow.
- Several versions of the same line at once, decorated differently — Heterophony would mean the notes underneath were the same tune as the one on top, ornamented or timed a little differently. The tune here is G4, F4, E4, D4, C4, and what is underneath is C, E, G, then D, F, A, then C, E, G, then G, B, D, then C, E, G — chords built under the tune, not further copies of it.

### Doubling and Blend

https://musictheory.center/course/orchestration#doubling-and-blend

> Play the first: seven single notes. Play the second and the third: the same seven notes each time, with one extra note added beneath every one of them — a different extra note in the two clips. Play those two back to back and say which of them you would still call one thing and which you would call two. Then play the fourth: one chord sounded twice with four notes each time, the two versions differing only in their top note, and say which note name you hear twice in each version.

**Plays:** Seven single notes (G4 A4 B4 C5) · The same seven, with a note added below — first way (G3 G4 A3 A4 B3 B4 C4 C5) · The same seven, with a note added below — second way (E4 G4 F4 A4 B4 C5) · One chord, filled out two ways (C3 E3 G3 C4 E4)

No staff here either: what this section turns on is which note appears twice, and a staff shows that no more plainly than the list beside it does. The table sorts the choice by what is being doubled and what each choice buys. Read the middle column down and one distinction runs through the whole of it — doubling a note the chord already contains changes the chord’s balance and nothing else, while adding a note the chord does not contain is no longer doubling at all but harmony, which is why the last row is in the table mainly to be told apart from the rest of it.

- **What the second clip adds** — The line is G4, A4, B4, C5, B4, A4, G4. The second clip puts G3 under the G4, A3 under the A4, B3 under the B4 and C4 under the C5 — the same letter every time, twelve half steps down. No note that was not already in the line has been introduced anywhere.
- **What the third clip adds** — The third clip puts E4 under the G4, F4 under the A4, G4 under the B4 and A4 under the C5. Every added note is two steps down the C major scale, which makes E4 to G4 a [Minor Third](https://musictheory.center/interval/minor-third) and F4 to A4 a [Major Third](https://musictheory.center/interval/major-third). The key decides which of the two you get, not the arranger.
- **The two versions of the chord** — Both are [C](https://musictheory.center/chord/c-maj). The first sounds C3, E3, G3 and C4 — the C twice. The second sounds C3, E3, G3 and E4 — the E twice. Only the top note is different between them.

| What is doubled | What it does | What to watch |
| --- | --- | --- |
| The root | Weight and certainty. The root is the note that tells you which chord this is, and two of it makes the chord sit down | Very little. In a triad set for four parts it is the safe default, and it is what the spare voice usually takes |
| The fifth | Almost nothing in either direction. The fifth is the note a chord can most afford to lose, so doubling it neither adds colour nor removes any | It leaves only one third in the chord, which in a large ensemble can make the harmony sound hollow and hard to place as major or minor |
| The third | Colour and warmth. The third is the note that makes the chord major or minor, and two of them push that quality forward | Two thirds can make a small group sound top-heavy — and in common-practice four-part writing the third of a dominant chord is the leading tone, which that tradition does not double |
| The same line at the [Octave](https://musictheory.center/interval/octave) | Reinforcement with no new note. The line is in two places twelve half steps apart and is still one line: louder, with a wider bottom or a brighter top depending on which end the copy is added at | It is parallel octaves by definition, which is exactly why Chapter 12’s rule has nothing to say about it here — the fusion the rule is designed to prevent is the whole point |
| The same line at the unison | Blend. Two sources on the same pitch make a composite colour neither has alone, because their tuning and their attacks never coincide exactly | It costs two performers to produce one line, which in a small ensemble is usually one performer too many |
| The line in thirds or sixths | A second, parallel line. This is not really doubling: a third below the tune adds a note the tune does not contain, so what comes out is harmony | The added line follows the key rather than a fixed distance, so the interval alternates between major and minor and the harmony changes underneath as it does |

**doubling**

**Definition.** Doubling is the practice of giving the same note, or the same line, to more than one part at once — at the same pitch or twelve half steps apart — so that one thing in the music is carried by two or more sounding sources.

Two words that travel with it are kinds of doubling and not other names for it. **Octave doubling** is a line copied twelve half steps above or below itself; **unison doubling** is a line copied at the same pitch on a second instrument or a second voice. Doubling a line in thirds is a third thing again and, strictly, not doubling at all: a third below a tune is a note the tune does not contain, so the result is a second line in harmony rather than the same line twice. The distinction matters because the first two change nothing about the chords and the third changes the chords.

What unison doubling buys is blend, and the reason is that two performers never coincide. Their tuning differs by a few cents and their attacks by a few milliseconds, and the sum of two slightly mistuned, slightly misaligned copies of one note is a tone that is thicker, slower to speak and audibly alive in a way neither is alone. That is the effect a string section is built on and the effect a producer reaches for when they record the same guitar part twice and pan the two copies apart. It is also why doubling costs something: it uses a player to get a colour rather than a note.

Inside a chord the question is narrower — which of the chord’s own notes gets sounded twice. Doubling the root adds weight; doubling the fifth adds almost nothing; doubling the third pushes the chord’s major or minor quality forward and can make a small group top-heavy. The one place a tradition turns this into a prohibition is the [leading tone](https://musictheory.center/course/scales#scale-degrees), and the reasoning is worth knowing even if you never write a chorale: two parts on a leading tone both want to rise a half step to the tonic, and if they both do, they arrive in parallel octaves. **Doubling is a decision about balance, and it is almost always reversible — which is why it is the cheapest thing to experiment with in an arrangement.**

**Good to know**

- (Fact) Two sources sounding the same written note never produce the same waveform. Their fundamentals differ by a few cents, which makes the two waves drift in and out of phase at a rate equal to the difference in their frequencies — slow beating, of a few cycles a second. Their attacks differ by a few milliseconds, which smears the onset. The sum is a single perceived pitch with a broader, slowly shifting spectrum and a softer front edge, and no amount of volume on one source reproduces it.
- (Constraint — four-part common-practice writing — chorale-style exercises and the repertoire they are drawn from. It is not a rule about doubling in general: a rock arrangement doubling a guitar line an octave down, or a horn section doubling a vocal, breaks nothing, because nothing in those textures depends on four parts staying separately audible) The leading tone is not doubled, and neither is the third of a dominant chord, which in a major key is the same note. Two parts carrying it would each be obliged to rise a half step to the tonic, arriving in parallel octaves at precisely the [cadence](https://musictheory.center/course/diatonic-harmony#cadences) the rule exists to protect; and whichever of the two did not rise would leave the chord’s strongest pull hanging. The fifth may be doubled, the root is preferred, and the third of a non-dominant chord is allowed with care.
- (Tendency — Western popular, film and big-band arranging, and orchestral writing from the nineteenth century onward. It is a habit of practice rather than a rule, and it is a habit about melody specifically — the same move applied to an inner part usually just makes the middle of the arrangement thicker) Where a [melody](https://musictheory.center/course/melody#what-makes-a-melody) has to be heard over a busy accompaniment, arrangers double it at the octave rather than adding a harmony note to it. An octave copy adds the line’s own strength without putting a second pitch class into the chord, so the tune gets louder and more present while the harmony underneath is left exactly as it was.

**You already know this**

- Parallel octaves from Chapter 12 — two parts a perfect octave apart moving in the same direction to another octave, which four-part writing avoids because the two parts fuse and stop being heard as two — which is the same motion this section calls octave doubling and uses on purpose for exactly that fusion — https://musictheory.center/course/voice-leading#parallels-and-why-they-are-avoided
- The triad from Chapter 10 — a chord of three notes built by stacking thirds, and only those three notes however many are actually sounded — which is what makes a fourth note in a chord a decision about which of the three to repeat rather than a decision about which note to add — https://musictheory.center/course/chords#stacking-thirds

**In real music**

- Ludwig van Beethoven — “Symphony No. 5, first movement” — the first two seconds. Several instruments play those four notes at once, on the same pitches and at octaves of them, and nothing harmonises them. Listen for a chord in the opening figure and there is none to find — which is what makes it land as one sound rather than as harmony

**Try it on your instrument:** Take a four-note melody you can play from memory. Play it once as it stands. Play it again with every note doubled twelve half steps below. Play it a third time with a note two scale steps below each melody note, staying in the [key](https://musictheory.center/course/keys-and-key-signatures#what-a-key-is), so that some of the added intervals come out major and some minor. Then play any major chord twice with four notes each time: once with the root appearing twice and once with the third appearing twice. Say which of those two chords you would put under a loud passage and which under a quiet one, and write down your reason.

**Check:** Four pairs of notes, then the clip stops. In every pair the lower note is two steps of the C major scale below the upper one. The line’s next note would be G4. Predict the note that would sit underneath it.

- E4, a minor third below **(correct)** — Correct. Two steps down the C major scale from G is E — count G, F, E. The four pairs you heard were C4 over A3, D4 over B3, E4 over C4 and F4 over D4, each lower note the same two steps below its upper one, so the pair after F4 over D4 is G4 over E4. E up to G is a minor third, because the key supplies E and G with only F between them.
- D4, a perfect fourth below — D is three scale steps below G — G, F, E, D — which is a perfect fourth, and none of the four pairs in the clip spanned one. The last pair you heard was F4 over D4, and D is two steps below F rather than three; a fourth below F4 would have been C4.
- F4, a whole step below — F is one scale step below G, a whole step. Every pair in the clip spanned two scale steps: C4 down to A3 is C, B, A. A note one step below the upper one would make the pair a second, which is a distance the clip never used and which would sound rough rather than harmonised.
- E♭4, a major third below — E♭ is not in C major, and every note in the clip was. The lower line was A3, B3, C4, D4 and the upper was C4, D4, E4, F4, with no accidental anywhere. Two steps below G in this key gives E natural, and E natural up to G is a minor third — E♭ would make it a major third and put a flat into a key that has none.

### Dynamics

https://musictheory.center/course/orchestration#dynamics

> Play the first, then the second. The notes are identical — the same four chords, the same bass underneath, the same tune on top, at the same speed. Something else is different in the second, and it is different by the end rather than at any one moment you could point at. Say what it is in your own words before you read on. Then play the third, where that same something changes again — except that this time you can point at exactly where.

**Plays:** Four chords, every one struck the same (C3 G3 E4 C5 F2 A3 F4 G2 B3 D4 B4) · The same four, and something grows (C3 G3 E4 C5 F2 A3 F4 G2 B3 D4 B4) · The same four, and something changes all at once (C3 G3 E4 C5 F2 A3 F4 G2 B3 D4 B4)

The second clip written out: four bars, one chord to a bar, held for the whole of it. The chords are C major, F major, G major and C major; the tune along the top is C5, C5, B4 and C5; the bass underneath falls from C3 to F2, rises to G2 and returns to C3. Under the first bar stands a small italic p and under the last a small italic f, and between them a long wedge opens from left to right, narrow beneath the second bar and at its widest beneath the fourth. The p asks for the first chord to be soft, the f for the last to be loud, and the wedge asks for everything in between to travel from one to the other gradually rather than in a single step. Nothing about the note heads differs from the first clip, and that is the whole point of the picture: none of this is written in the notes. The table below is the row of marks the p and the f belong to, softest at the top; press any row to hear that chord at that level.

- **Two marks ask for a change rather than a level** — A **crescendo** gets louder and a **diminuendo** gets softer. Both are written either as the word, usually abbreviated to cresc. and dim., or as a wedge that opens in the direction the sound is going — which is why players call them hairpins.
- **Subito means the change is not gradual** — Written before a mark — subito p, often abbreviated sub. p — it asks for the new level to arrive at once and with no approach to it, which is the third clip above. Without it, a new mark after a hairpin is simply where the hairpin was heading.
- **A hairpin needs a level at both ends** — On its own a wedge says only "louder", not louder than what or louder to what. A mark before it and a mark after it are what turn a direction into an instruction, which is why the staff above carries a p, a wedge and an f rather than a wedge alone.

| Mark | Italian | What it asks for |
| --- | --- | --- |
| pp | pianissimo | Very soft — as quiet as the player can be and still speak clearly |
| p | piano | Soft |
| mp | mezzo-piano | Moderately soft. Mezzo means half, and it halves the word after it |
| mf | mezzo-forte | Moderately loud — the same halving applied to forte |
| f | forte | Loud |
| ff | fortissimo | Very loud. Past pp and ff a composer simply adds more letters |

**dynamics**

*also called*: nuances (French teaching and French editions, where les nuances is the ordinary word for a score’s dynamic markings)

**Definition.** Dynamics are how loud music is, and the marks that ask for it: a short row of Italian words, abbreviated to one or two letters, each naming a level relative to the other levels in the same piece rather than any measurable quantity of sound.

The marks are an order, not a scale of measurement. There is no loudness that f names, and no composer has ever meant one. A flute at ff and an orchestra at ff are nowhere near the same amount of sound; what they have in common is that each is at the top of what it has, and that each is louder than the mf in the bar before. Six marks cover nearly everything ever printed, and a composer who wants more extreme simply adds letters — ppp, ffff — which means only "further in that direction than the last one".

What a listener actually follows is the change rather than the level. The ear settles into a constant loudness within a few seconds and stops reporting it, which is why a piece held at one level for three minutes sounds flat however loud that level is, and why the two wedge-shaped marks in the panel above do more work than the six letters do. The third clip is the other way of changing: a new level arriving with nothing leading to it, which lands as an event precisely because the ear had settled.

For an arranger the question is never how loud a part is but how loud it is **against the others**, and that is where this section joins the three before it. A tune that cannot be heard has three fixes. Move it into a register nothing else is occupying, which is where this chapter began. Double it, which is the section immediately before this one. Or ask for it louder and for the accompaniment softer, which is this one — and that third fix is the weakest of them, because the first two are true of the arrangement whoever plays it and this one is a request a player has to grant.

**Good to know**

- (Fact) Loudness is how far the source swings and pitch is how often it swings, which is Chapter 2’s distinction and the reason dynamics is a dimension of its own. The same pitch, held for the same length, on the same instrument, can be played at any loudness the instrument has — so loudness cannot be derived from any of the other things this course describes, and nothing in Chapters 5 to 14 predicts a single one of the marks in the table above.
- (Fact) Loudness does not add the way players do. Sound intensity adds, but the ear’s response to intensity is roughly logarithmic: ten times the intensity is heard as about twice as loud. Ten players are therefore about twice the sound of one rather than ten times, which is why adding one more to a large section changes almost nothing audible, and part of why an orchestra carries sixteen first violins against four horns.
- (Convention) The marks are Italian, abbreviated, and read as a sequence — pp, p, mp, mf, f, ff — with mezzo meaning half. That vocabulary settled in Europe alongside the instruments that could act on it, and it is a convention of one notational tradition: it says nothing about loudness that a performer in a tradition without the marks is failing to do.
- (Tendency — Western printed music and the recorded popular music descended from it. The generalisation is about notation rather than about sound: unmarked repertoire — plainchant, most folk singing, a great deal of music outside Europe — is shaped in loudness by performers continuously, and the absence of a symbol is not the absence of the thing) Dynamic marks are sparse before the eighteenth century, dense in the nineteenth, and largely absent from popular music, where the level of each part is decided at the mixing desk rather than by the player and then deliberately flattened by compression. A pop record is not undynamic; its dynamics are simply not the performer’s to write and not the score’s to record.

**You already know this**

- Loudness from Chapter 2 — how far the source swings rather than how fast, and the one property of a sound named there that this course had no section of its own for until this one — https://musictheory.center/course/pitch-and-notes#sound-and-vibration
- Doubling from earlier in this chapter, which makes a line louder by adding a player to it rather than by asking for more from the one already there — the structural version of what this section asks for with a mark — https://musictheory.center/course/orchestration#doubling-and-blend
- Chapter 16’s rule that a letter records material and nothing else — not key, not scoring, not dynamics — which is why a quiet verse and a loud chorus built on one chord loop are the same letter, and why what tells them apart is this section rather than that one — https://musictheory.center/course/form-structure#analysing-form

**Where this goes**

- Hearing a cadence arrive before it arrives is what lets a performer put the dynamic on it rather than just after it, which is where these marks stop being print and start being playing — https://musictheory.center/course/applications

**In real music**

- Joseph Haydn — “Symphony No. 94 in G major, "Surprise", second movement” — the theme is stated quietly and then immediately repeated even more quietly, and a fortissimo chord for the whole orchestra, timpani included, interrupts the end of that repeat. Not one note of the tune has changed at that moment and the harmony is the plainest in the movement; the only thing that happens is loudness, which is why the joke works on people who cannot read a note
- Edvard Grieg — “In the Hall of the Mountain King, from Peer Gynt” — the same short tune from beginning to end, stated quietly in the low strings and repeated until the whole orchestra is playing it fff. It is the plainest long crescendo in the repertoire, with one honest complication worth hearing for itself: it speeds up as well as getting louder, so listen to the opening statement and the closing one and try to separate the two changes
- Nirvana — “Smells Like Teen Spirit” — the verse and the chorus are built on the same four chords in the same order. The verse is played with the guitar barely present and the chorus at full volume, and that difference alone is what tells a listener which section they are in — the clearest possible case of loudness doing work that the notes are not doing

**Try it on your instrument:** Take a four-chord [phrase](https://musictheory.center/course/melody#phrase-antecedent-consequent) you can already play without thinking about it. Play it four times over and change nothing but how hard you play. First, keep all four chords at exactly the same loudness, which is harder than it sounds and worth finding out. Second, start as quietly as your instrument will speak and arrive on the loudest chord you can play cleanly. Third, do that backwards. Fourth, play three chords quietly and the last one as loudly as you can, with nothing in between. Then play the second version to somebody and ask them to raise a hand the moment it starts getting louder — most people raise it much later than you started.

**Check:** The same four chords once more, with not a note altered. Which of the four shapes was it this time?

- It got quieter from first chord to last **(correct)** — Correct. The first chord was the loudest and each one after it was softer, which is a diminuendo — printed as the word dim. or as a wedge closing from left to right. Because the loudest moment is the one you hear first, a phrase shaped this way sounds as though it is withdrawing rather than arriving, which is the opposite of the second clip at the top of the section.
- It got louder from first chord to last — That is the second clip at the top of this section, and it is the natural thing to expect because the great majority of printed hairpins open rather than close. Play the two back to back and listen only to the first chord of each: here it was the loudest thing in the clip, there it was the quietest.
- It stayed at one level throughout — The first clip is the one at a single level, and it is the useful comparison to make: play it straight after this one and the difference is obvious, because nothing in it ever recedes. What you heard arrived at its quietest chord at the end rather than its loudest.
- It stayed quiet and then one chord was loud — A sudden change is the third clip, where three chords sit at one level and the fourth is somewhere else entirely with nothing leading to it — the change happens between two chords and you can point at which two. Here every chord differed from the one before it by roughly the same amount, so there is no single place to point at.

**Check:** A passage for solo flute is marked ff. Somewhere else in the same piece a passage for the whole orchestra is marked ff. Before you play anything, decide which of the two is the louder sound.

- The orchestra **(correct)** — Correct, and by an enormous margin. A mark names a level relative to the levels around it, not a quantity of sound. Both passages are asking for the top of what the players have, and the top of what eighty players have is nothing like the top of what one flute has. The mark tells each of them to give everything; it never promises that two everythings match.
- Neither — the mark is the same, so the loudness is the same — That is what the mark would mean if it were a measurement, and it is much the commonest misreading of the system. It cannot be one: if ff named a fixed loudness, a piece for solo flute could never reach it and a piece for full orchestra could never get below it, and both kinds of piece are marked with the whole range all the time.
- The flute — A flute at full pressure is genuinely loud and bright enough to carry over a large accompaniment, which is exactly why flutes are given solos to play over one. But bright enough to be picked out and louder than the whole orchestra are different claims, and only the first of them is true.

#### The instrument that could not get louder, and the one named after fixing it

A harpsichord plucks its strings with a quill, and how hard the key is pressed changes nothing about how hard the quill plucks. A harpsichordist can play one note before another and one note longer than another, and cannot make one note louder than another by pressing its key harder. The instrument that dominated two centuries of European keyboard music had no dynamics in the sense this section means.

What it had instead was levels. A second manual, a different set of strings, a stop engaged or released: whole blocks of music at one loudness and then a switch to another, with nothing in between. That is why Baroque dynamics are called terraced, and why a Baroque score rarely marks more than a bare p and f — a hairpin would have been an instruction the instrument could not carry out.

The replacement was named after the thing it could do. Bartolomeo Cristofori built a keyboard around 1700 whose hammers struck rather than plucked, so that a harder press produced a louder note, and called it a gravicembalo col piano e forte — a harpsichord with soft and loud. The phrase shortened to pianoforte and then to piano, so the instrument most of this course is demonstrated on is named after its dynamic range and nothing else. The vocabulary grew with the instruments: pp and ff are enough for most of the eighteenth century, the nineteenth is full of ppp and fff, and Tchaikovsky asks a bassoonist for pppppp in the first movement of his sixth symphony — six letters for a loudness no notation can really specify and every player understands anyway.

### Writing for an Ensemble

https://musictheory.center/course/orchestration#writing-for-an-ensemble

> Play the first: four chords, all four crowded into the space of one hand. Play the second: the same four chords with their notes pulled far apart — one note well below, two in the middle, one well above. Play the third: the same four chords again, arranged a third way. Play the second and the third back to back and say which of them lets you follow a tune, and where in the third one the tune has gone.

**Plays:** Four chords, all in one handful (C4 E4 G4 B3 D4 A3 F4) · The same four, arranged another way (C2 E3 G3 E5 G2 B3 D4 D5 A2 C4 E4 C5 F2 A3 A4) · The same four, arranged a third way (C2 E2 E3 G3 D2 G2 B3 D4 A2 C4 E4 F2 A3)

No staff again: what these three clips differ in is where each part sits, and a four-part spread over four octaves is exactly the thing a single staff shows worst. The second clip is the arrangement the table below describes — one part on the bass, two filling the middle, one on top with the tune. The third is the same four chords with the tune dropped down into the bass’s own territory, which is the mistake the second row of the table exists to prevent. Read the last column down: each entry names the sound a missing decision makes, and three of them are audible in the third clip — the tune is no longer above the other parts, the bass part is no longer the lowest sound in two chords out of four, and the first and last chords put a major third down at the very bottom. The chip on each row plays a setting in which that row’s decision is missing, except the third row, which plays the decision kept.

- **The harmony is identical in all three clips** — Every clip plays [C](https://musictheory.center/chord/c-maj), then [G](https://musictheory.center/chord/g-maj), then [Am](https://musictheory.center/chord/a-m), then [F](https://musictheory.center/chord/f-maj), in that order. No note name differs anywhere between them. Everything you can hear change is placement.
- **Where the parts sit in the second clip** — The bass is C2, G2, A2, F2. The middle pair is E3 and G3, then B3 and D4, then C4 and E4, then A3 and C4. The tune on top is E5, D5, C5, A4 — four notes descending, with a sixth or more of clear air under every one of them: a major thirteenth over the G3 of the first chord, an octave over the D4 of the second, a minor sixth over the E4 of the third and a major sixth over the C4 of the fourth.
- **What the third clip changes, and what it does not** — Only the tune moves — E5, D5 and C5 down three octaves to E2, D2 and C2, and the A4 down two, to A2, which is as low as this chapter goes. The bass and the middle pair are untouched. In the second chord the tune’s D2 now lies below the bass’s G2, and in the third the tune’s C2 lies below the bass’s A2 — so the lowest sounding note is no longer the note the arrangement assigned to the bass, and in two chords out of four the root is not at the bottom at all.

| The decision | What it means | What it sounds like when it is missing |
| --- | --- | --- |
| Assign the bass | One part, and only one, owns the lowest note of each chord. Everything else stays above it | Two parts competing at the bottom, and a harmony whose root the ear cannot locate |
| Give the tune its own space | Keep the melody clear of the other parts — in Western practice, usually above all of them — so that nothing is sounding in the band it occupies | A tune inside the accompaniment’s own range, which disappears into it and has to be played louder to be heard at all |
| Fill the middle without crowding | The inner parts supply the notes the bass and the tune leave out, and no more than that | A complete chord in every part, four copies of the same harmony, and nothing for the ear to hear through |
| Space wide at the bottom, close at the top | The interval between adjacent parts gets smaller as you climb: an octave or a fifth at the bottom, thirds and seconds at the top | Mud. Two notes a third apart down in the bass band beat against each other and blur the chord above them |
| Leave holes | Not every part plays all the time, and a part that stops has somewhere to come back from | A texture with nothing missing, which cannot grow, cannot arrive anywhere, and sounds the same in the last chorus as the first verse |

**arranging**

**Definition.** Arranging is the work of deciding who plays what: which part carries the tune, which supplies the bass, which fills the harmony between them, and where in its own range each of them does it.

Arranging and orchestration are used interchangeably, and there is a real distinction underneath them. **Orchestration** is strictly the assignment of already-written music to instruments — the same notes, distributed. **Arranging** is the larger job, which may also change the harmony, the form, the [key](https://musictheory.center/course/keys-and-key-signatures#what-a-key-is) or the accompaniment pattern before deciding who plays it. In practice nobody keeps the line clean, and an orchestrator who moves an inner part down an [octave](https://musictheory.center/course/pitch-and-notes#the-octave) to stop it clashing has just made an arranging decision. The useful half of the distinction is the reminder that placement alone is a craft: the second and third clips above differ in nothing but placement, and only one of them works.

The physical constraint the craft is built around is that the low end of the pitch range holds much less than the high end. A minor third between two parts sounds like harmony in the middle of the range and like roughness two octaves down, for the reason §1 gave — the hertz gap at the bottom is small enough to fall inside one critical band. The practical rule that falls out is to space the parts wide at the bottom and close them up as they rise, which is not an arbitrary aesthetic: it is the spacing a single vibrating string already produces above its own fundamental.

The rest of the craft is subtraction. A beginner’s arrangement is four parts each playing a complete chord, which is four times as much information as the listener needs and leaves nowhere for the music to go. **An arrangement is mostly a set of decisions about what not to play.** Give the bass to one part and keep the others off it; give the tune a band nothing else is in; let the middle supply only what the outer parts have not already said; and take a part out entirely for a section so that its return is an event.

**Good to know**

- (Fact) The overtone series above any note is widest at the bottom and narrows as it rises. Above a fundamental the intervals between successive partials are an octave, then a perfect fifth, then a perfect fourth, then a major third, then a minor third, each smaller than the last. A chord spaced wide at the bottom and close at the top is therefore spaced the way a single vibrating string, column of air or voice is already spaced — the arrangement and the acoustics of one note agree.
- (Constraint — conventional acoustic arranging in the Western tradition — orchestral, choral, big-band and most acoustic pop. Electronic production has other ways of keeping low parts distinct, including filtering, sidechaining and stereo placement, and dense low clusters are a deliberate resource in a good deal of it) Keep the interval between the two lowest sounding parts wider than the intervals higher up: an octave or a fifth at the bottom, and thirds and seconds only once you are above the bass band. A third between two parts in the bottom octave will be heard as roughness rather than as harmony, and it blurs whatever is sounding above it as well.
- (Tendency — Western popular and classical practice, where the highest part is the default melody. It is an expectation rather than a law, and composers have exploited it deliberately for centuries: a tenor melody under moving upper parts works precisely because the ear has to be persuaded to go and find it) The tune goes on top. A listener hearing several parts at once takes the highest as the [melody](https://musictheory.center/course/melody#what-makes-a-melody) unless something stops them, so putting the tune anywhere else — in an inner part, or in the bass — is a decision that has to be paid for with dynamics, rhythm or timbre to make the tune findable.

**You already know this**

- The bass line from Chapter 12 — the lowest sounding part, read as a line in its own right, whose note decides how settled each chord sounds — which is the first thing an arrangement assigns and the one part nothing else is allowed to sit under — https://musictheory.center/course/voice-leading#bass-lines-and-inversions
- The phrase from Chapter 16 — a stretch of music that makes one complete statement and then stops — because an arrangement changes what is playing at those joins and rarely anywhere else — https://musictheory.center/course/form-structure#from-phrase-to-section

**In real music**

- Maurice Ravel — “Boléro” — neither of its two themes changes for fifteen minutes, and neither does the snare drum pattern underneath them. Listen to the first statement — one flute, low and quiet — and then to any statement ten minutes later. Every difference between them is an arranging decision: who is playing, how many of them, and in which part of their range

**Try it on your instrument:** Take a [four-chord loop](https://musictheory.center/course/contemporary#pop-and-loop-based-harmony) you already play and arrange it for three parts on your instrument or in your DAW. Part one plays only the root of each chord, as low as the instrument comfortably goes. Part two plays two notes of each chord somewhere around middle C, and never the same two notes twice in a row if you can avoid it. Part three plays a four-note tune above both. Play it through. Then move part three down two octaves, leave the other two exactly where they are, and play it again. Write one sentence about what you can no longer hear, and one about which of the five decisions in the table above the second version broke.

**Check:** A setting of the same four chords you have not heard yet, played once. One of the five decisions in the table above was broken first, and it caused the others. Decide which one.

- The middle parts have been crowded down onto the bass **(correct)** — Correct, and the fourth row of the table is the one that broke. The bass is C2, G2, A2 and F2 as before and the tune is still E5, D5, C5 and A4, but the two inner parts have come down an octave to sit directly on the bass: E2 and G2 over the C2, B2 and D3 over the G2, C3 and E3 over the A2, A2 and C3 over the F2. That puts a major third at C2–E2 in the first chord and another at F2–A2 in the last, which is the spacing the fourth row calls mud — and because the inner parts are down there, the third row goes too: in the second chord there is nothing at all sounding between D3 and D5.
- The chords are not the ones the other clips played — The chords are identical. Every clip in this section plays C major, then G major, then A minor, then F major, and this one contains exactly those notes: C, E and G, then G, B and D, then A, C and E, then F, A and C. The harmony was never touched — only where each of its notes was put.
- The whole thing has been moved into another key — Nothing has been transposed. The bass is still C2, G2, A2 and F2 and the tune is still E5, D5, C5 and A4, exactly as in the second chip at the top of the section. What moved is the pair of notes between them, and each of those moved by an exact octave — E3 to E2, G3 to G2, B3 to B2, D4 to D3, C4 to C3, E4 to E3, A3 to A2 — which changes no note name and therefore no key.
- There is nothing underneath the tune at all — There is a bass, and it is the same one: C2, G2, A2 and F2, the lowest note of every chord and the root of every chord. The trouble is not that the bottom is empty but that it is overcrowded — two more parts have been stacked straight on top of those four notes instead of being left up in the middle of the range.

### The Instrument Families

https://musictheory.center/course/orchestration#the-instrument-families

> Play the three clips in order. Each is the same four notes in the same order, and the only thing that differs between them is how far up they sit. One warning before you start: every clip on this page is played by the same sampled piano, so nothing here will show you what a violin or a trumpet sounds like — that half of the subject cannot be demonstrated by this dock at all. What these three clips can show is the other half. Ask yourself, for each one, which instruments you know of could reach it and which could not.

**Plays:** A four-note phrase (G2 A2 B2) · The same four notes, further up (G4 A4 B4) · The same four notes, further up again (G6 A6 B6)

The dock plays everything in this chapter with one sampled piano, so nothing you have heard demonstrates timbre — the difference between a violin and a clarinet on the same note is precisely what these clips cannot give you. What they do give you is range, and range is the half of the subject that settles most arranging questions before timbre is reached: the three clips are the same four notes in three places, and a double bass could play only the first of them, a viola only the second, and a violin the second and the third. The table below is this chapter’s single reference for the families. Every range in it is a sounding range — what you hear, not what the player reads — rounded to the standard textbook figure, and every one assumes a competent player rather than a specialist. Individual instruments and individual players go further at both ends, and the notes at both ends are the hardest to control.

- **What the three clips play** — The first is G2, A2, B2, G2. The second is G4, A4, B4, G4 — the same four notes two octaves higher. The third is G6, A6, B6, G6, two octaves higher again.
- **The tunings behind the string ranges** — Violin, viola and cello are each tuned in fifths — every pair of adjacent strings a [Perfect Fifth](https://musictheory.center/interval/perfect-fifth) apart. The violin’s four are G3, D4, A4 and E5; the viola’s sit a fifth below that and the cello’s an octave below the viola’s. The double bass is tuned in fourths instead — a [Perfect Fourth](https://musictheory.center/interval/perfect-fourth) between adjacent strings — and the guitar in fourths but for one third, which is why the bass sits slightly apart from the other three in fingering as well as in range.

| Family | Members, with sounding ranges | How the sound is made | Role in an ensemble |
| --- | --- | --- | --- |
| Strings | Violin G3–A7 · viola C3–E6 · cello C2–C6 · double bass E1–G4 | A bow drawn across a stretched string, or the string plucked with a finger | The orchestra’s backbone. Melody at the top, inner harmony in the middle, the bass line at the bottom — and the only family that can play continuously without stopping to breathe |
| Woodwinds | Flute C4–D7 · oboe B♭3–A6 · clarinet in B♭ D3–B♭6 · bassoon B♭1–E♭5 | A column of air inside a tube, set going by blowing across a hole or through a single or double reed | Solo colour and agility. Fast lines, and melodies meant to be heard as one distinct voice rather than as a blended section |
| Brass | Horn in F F2–F5 · trumpet in B♭ E3–C6 · trombone E2–F5 · tuba D1–F4 | The player’s lips buzzing into a mouthpiece, setting a long coiled tube of air vibrating | Weight and carrying power. Fanfares, sustained chords that can be heard through a full orchestra, and the bottom of the harmony in the tuba |
| Percussion | Timpani about D2–A3 across a set of four drums; snare drum, bass drum, cymbals and triangle have no definite pitch | A surface struck, shaken or scraped | Time, accent and impact. The pitched members can also take a bass note or a melody, which is what a xylophone or a marimba is usually doing |
| Keyboards | Piano A0–C8 · organ and synthesizers vary with the instrument and the stop or patch selected | A key mechanism operating hammers, pipes, reeds or an electronic oscillator | A whole ensemble under one pair of hands: bass, harmony and melody at once, which is why arrangements are sketched here before they are distributed |
| Voices | Soprano C4–C6 · alto F3–F5 · tenor C3–C5 · bass E2–E4 | Air from the lungs setting the vocal folds vibrating, the result shaped by the throat, mouth and nose | The tune, nearly always, and the only family that carries words — which is why everything else in a song arrangement is placed around it |
| Rhythm section | Electric guitar E2–E6 · bass guitar E1–G4 · drum kit, mostly unpitched · keyboards as above | Plucked strings read by a magnetic pickup and amplified, and struck drums and cymbals | The engine of a band. Bass and drums fix the harmony’s foundation and its time between them; guitar and keys fill the middle and leave the top for the voice |

**instrument family**

*also called*: section (in orchestral usage, where "the string section" names the same group of players)

**Definition.** An instrument family is a group of instruments that produce sound the same way — a bowed string, a column of air, a pair of buzzing lips, a struck surface — and which therefore share a manner of speaking, a way of being played, and a set of things none of them can do.

The orchestra is conventionally described as four of them, and the table above adds three more that a description of the orchestra leaves out and an arranger cannot. Keyboards are a family by construction rather than by sound, and they matter because a keyboard is where most arrangements are written. Voices are a family with a capability nothing else has, which is words. And the rhythm section of popular music — bass, drums, guitar, keys — is a working ensemble with a division of labour as fixed as any orchestral one, and no place in a four-family scheme at all.

What a page can carry about a family is range, and what it cannot carry is timbre. The table gives you the first half honestly: a cello stops at about C6 and no amount of asking will get it higher, so a line written above that has to go somewhere else. The second half — why an oboe cuts through a texture a clarinet blends into, why a muted trumpet and an open one are practically different instruments — is not something reading about it will teach, and this chapter’s clips cannot teach it either, because they all sound through the same sampled piano. The only way to learn it is to listen to the families one at a time with the score in front of you.

Use the table in that order. Check the range first, because a range is a hard limit and settles the question before any other consideration arises. Then check the role, which tells you what the family is used to being asked for and therefore what an arrangement will sound conventional or surprising for doing. Then, and only then, worry about colour — and go and listen rather than read. **A part that is out of range is not a stylistic choice; it is a part nobody can play.**

**Good to know**

- (Convention) A full score lists the families in a fixed order from the top of the page downward: woodwinds, then brass, then percussion, then harp and keyboards, then voices, then strings at the bottom. Nothing about the sound requires this, and it does not reflect who is loudest or who is most important. It is a printing convention, old enough that everybody has learned it, and its whole value is that a conductor can find a part by where it sits on the page.
- (Convention) Several instruments sound a different pitch from the one their player reads. A clarinet in B♭ and a trumpet in B♭ each sound a major second below the written note; a horn in F sounds a perfect fifth below; a double bass and a guitar sound an [octave](https://musictheory.center/course/pitch-and-notes#the-octave) below. The reason is fingering rather than acoustics — one set of fingerings serves a whole family of differently sized instruments — and the consequence is that "range" is an ambiguous word unless it is qualified. Every range in the table above is a **sounding** range.
- (Tendency — acoustic orchestral and band instruments as described in orchestration texts. Individual instruments and individual players vary a great deal, some instruments invert the pattern — the clarinet’s lowest register is one of its strongest — and electronic instruments have no such constraint at all) The extremes of an instrument’s range are not neutral territory. The top of most acoustic instruments is loud, effortful and hard to play quietly; the bottom is weak, slow to speak and hard to control. An arranger who needs a part to blend keeps it out of both ends, and an arranger who wants strain, brilliance or a struggling sound puts it there on purpose.

**You already know this**

- The instrument you picked in Chapter 1 — the one this site draws for you instead of assuming a keyboard — whose range is the one row of the table above you can check by hand rather than take on trust — https://musictheory.center/course/fundamentals#pick-your-instrument
- Note names with octave numbers from Chapter 2 — the digit after the letter, counting upward from each C — which is the notation every range in the table above is written in and the only reason those ranges are comparable — https://musictheory.center/course/pitch-and-notes#naming-notes

**Where this goes**

- Deciding which instrument plays which part of a song you are analysing — or writing — is where every chapter of this course arrives at once, and it is the work the last chapter turns into a routine — https://musictheory.center/course/applications

**In real music**

- Benjamin Britten — “The Young Person’s Guide to the Orchestra” — the first four minutes. One theme is played by the whole orchestra, then by the woodwinds alone, then the brass, then the strings, then the percussion, then everyone again. The notes barely change between the statements — except the percussion statement, where only the rhythm of the theme survives — so almost everything you hear change is the family playing them — which is the demonstration this chapter’s dock cannot give you

**Try it on your instrument:** Look up the range of the instrument you play and write it down with octave numbers: lowest note, highest note. Play both of them. Then play the note a perfect fourth above your lowest and the note a perfect fourth below your highest, and compare how much effort each of the four took and how easy each was to play quietly. Finally, take the three-part arrangement you made in the previous section and say, for each part, which families in the table above could play it exactly as you wrote it and which could not — and for the ones that could not, what you would change.

**Check:** Four notes in turn, played once. Decide which part of the pitch range they sat in.

- Up where a violin or a flute plays, past the top of a cello **(correct)** — Correct. The four notes were A6, C7, B6 and A6 — the lowest of them, A6, a major sixth above the C6 at the top of a cello’s range and a perfect fourth above the E6 at the top of a viola’s, so neither instrument could play a note of it. A violin reaches about A7 and a flute about D7, and both can play all four. That is what a range table is for: it rules families out before it rules any family in.
- In the middle, where a viola or a trombone sits — A viola runs from C3 to about E6 and a trombone from E2 to about F5. Every note in this clip is above both of those ceilings: the lowest of them, A6, sits a perfect fourth above the viola’s top and a major tenth above the trombone’s.
- Down where a double bass or a tuba plays — A double bass sounds from E1 up to about G4 and a tuba from D1 up to about F4. The lowest note in this clip was A6, more than two octaves above the top of either. Nothing in the clip is within reach of them.
- Above the highest note on a piano — A piano runs from A0 to C8. The highest note here was C7, exactly one octave below the top of the keyboard, and the lowest was A6 — high, but comfortably on the instrument with a whole octave of keys still above it.

### Key Takeaways

https://musictheory.center/course/orchestration#key-takeaways

- A register is a band of the pitch range treated as a region, and the same band means the same thing whichever instrument is playing in it.
- Range is everything an instrument can produce, register is a slice of the pitch range, and tessitura is the slice a particular part actually occupies.
- The same chord moved up or down two octaves is the same chord, but the low version crowds and the high version thins, because an interval spans far fewer hertz at the bottom of the range than at the top.
- Texture is how many independent lines are sounding and how they relate: monophony, homophony, polyphony and heterophony are the four standard answers.
- The ear decides which texture it hears from rhythm and direction — notes that start and move together fuse, notes that start or move differently separate.
- Doubling sounds one note or one line from more than one source; octave doubling and unison doubling are kinds of it, and doubling in thirds is harmony rather than doubling.
- Unison doubling produces blend because two performers never coincide exactly in tuning or attack, and the mismatch is the colour.
- In four-part common-practice writing the leading tone is not doubled, because two parts on it both want to rise a half step and would arrive in parallel octaves.
- An arrangement assigns the bass to one part, gives the tune a band of its own, fills the middle with only what the outer parts left out, spaces wide at the bottom and close at the top, and takes parts out so they can come back.
- The natural spacing of a chord — wide at the bottom, close at the top — is the spacing the overtone series of a single note already has.
- An instrument family is a group that makes sound the same way, and a range table settles what a family can play before any question of colour arises.
- Ranges are ambiguous unless qualified: a clarinet in B♭ sounds a major second below what its player reads, and a double bass sounds an octave below.

### Questions

**What is register in music?**

A register is a band of the pitch range treated as a region — the bass band, the mid band, the high band — named for what it carries rather than for its notes. Range is the whole span an instrument can produce; tessitura is the part of that span a particular part actually occupies.

**Why do low chords sound muddy?**

Because the same interval spans far fewer hertz low down. C2 to E2 is a gap of about 17 Hz while C6 to E6 is about 272 Hz, and the ear separates two tones only when their gap is wide enough relative to the critical band. Low thirds fall inside one band and are heard as roughness.

**What are the four musical textures?**

Monophony is one line and nothing else. Homophony is one line carrying the tune with everything else supporting it. Polyphony is several lines, each with a shape of its own. Heterophony is one line played simultaneously by several performers, each decorating or timing it differently.

**What does doubling mean in arranging?**

Doubling gives the same note or the same line to more than one part at once — at the same pitch, or an [octave](https://musictheory.center/course/pitch-and-notes#the-octave) apart. It adds weight and blend without adding harmony. Doubling a line in thirds is a different thing: the added notes are not in the line, so the result is harmony.

**What is the difference between arranging and orchestration?**

Orchestration is strictly the assignment of already-written music to instruments: the same notes, distributed. Arranging is the larger job and may also change harmony, form, [key](https://musictheory.center/course/keys-and-key-signatures#what-a-key-is) or accompaniment before deciding who plays it. In practice the words are used interchangeably, and most orchestrating involves small arranging decisions.

**What are the families of the orchestra?**

Strings, woodwinds, brass and percussion, grouped by how each makes its sound: a bowed string, a column of air, buzzing lips, a struck surface. An arranger needs three more that the scheme leaves out — keyboards, voices, and the bass-and-drums rhythm section of popular music.

### Other languages

- https://musictheory.center/course/orchestration/llms.txt
- https://musictheory.center/fr/cours/orchestration/llms.txt
- https://musictheory.center/es/curso/orquestacion/llms.txt