Scales: Just vs Equal Temperament

(and related topics)

The "Just Scale" (sometimes referred to as "harmonic tuning" or "Helmholtz's scale") occurs naturally as a result of the overtone series for simple systems such as vibrating strings or air columns. All the notes in the scale are related by rational numbers. Unfortunately, with Just tuning, the tuning depends on the scale you are using - the tuning for C Major is not the same as for D Major, for example. Just tuning is often used by ensembles (such as for choral or orchestra works) as the players match pitch with each other "by ear." The "equal tempered scale" was developed for keyboard instruments, such as the piano, so that they could be played equally well (or badly) in any key. It is a compromise tuning scheme. The equal tempered system uses a constant frequency multiple between the notes of the chromatic scale. Hence, playing in any key sounds equally good (or bad, depending on your point of view). There are other temperaments which have been put forth over the years, such as the Pythagorean scale, the Mean-tone scale, and the Werckmeister scale. For more information on these you might consult "The Physics of Sound," by R. E. Berg and D. G. Stork (Prentice Hall, NJ, 1995). For an interesting discussion about the historical development of the equal tempered scale, you might read "How Equal Temperament Ruined Harmony," by Ross W. Duffin (W.W. Norton & Co., NY, 2007). For a very complete list of historical temperaments, see the book by Owen Jorgensen listed at the bottom of this page. A table showing a comparison of one meantone temperament with equal temperament can be found here. The table below shows the frequency ratios for notes tuned in the Just and Equal temperament scales. For the equal temperament scale, the frequency of each note in the chromatic scale is related to the frequency of the notes next to it by a factor of the twelfth root of 2 (1.0594630944....). For the Just scale, the notes are related to the fundamental by rational numbers and the semitones are not equally spaced. The most pleasing sounds to the ear are usually combinations of notes related by ratios of small integers, such as the fifth (3/2) or third (5/4). The Just scale is constructed based on the octave and an attempt

to have as many of these "nice" intervals as possible. In contrast, one can create scales in other ways, such as a scale based on the fifth only.

Ratio to Fundamental Ratio to Fundamental Just Scale Equal Temperament Unison 1.0000 1.0000 Minor Second 25/24 = 1.0417 1.05946 Major Second 9/8 = 1.1250 1.12246 Minor Third 6/5 = 1.2000 1.18921 Major Third 5/4 = 1.2500 1.25992 Fourth 4/3 = 1.3333 1.33483 Diminished Fifth 45/32 = 1.4063 1.41421 Fifth 3/2 = 1.5000 1.49831 Minor Sixth 8/5 = 1.6000 1.58740 Major Sixth 5/3 = 1.6667 1.68179 Minor Seventh 9/5 = 1.8000 1.78180 Major Seventh 15/8 = 1.8750 1.88775 Octave 2.0000 2.0000 Interval

You will note that the most "pleasing" musical intervals above are those which have a frequency ratio of relatively small integers. Some authors have slightly different ratios for some of these intervals, and the Just scale actually defines more notes than we usually use. For example, the "augmented fourth" and "diminished fifth," which are assumed to be the same in the table, are actually not the same. The set of 12 notes above (plus all notes related by octaves) form the chromatic scale. The Pentatonic (5-note) scales are formed using a subset of five of these notes. The common western scales include seven of these notes, and Chords are formed using combinations of these notes. As an example, the chart below shows the frequencies of the notes (in Hz) for C Major, starting on middle C (C4), for just and equal temperament. For the purposes of this chart, it is assumed that C4 = 261.63 Hz is used for both (this gives A4 = 440 Hz for the equal tempered scale).
Equal Difference Note Just Scale Temperament C4 261.63 261.63 0 C4# 272.54 277.18 +4.64 D4 294.33 293.66 -0.67 E4b 313.96 311.13 -2.84

E4 F4 F4# G4 A4b A4 B4b B4 C5

327.03 348.83 367.92 392.44 418.60 436.05 470.93 490.55 523.25

329.63 349.23 369.99 392.00 415.30 440.00 466.16 493.88 523.25

+2.60 +0.40 +2.07 -0.44 -3.30 +3.94 -4.77 +3.33 0

Since your ear can easily hear a difference of less than 1 Hz for sustained notes, differences of several Hz can be quite significant! Listen to the difference: The first second of this WAV file contains a major triad starting on F# (F# A# - C#) using the Just scale appropriate for C Major. The last part of the file contains the same triad but using the Just scale appropriate for F# Major. (This is one of the worst case situations). Tuning Shift WAV file. Here's another example to test your ears. The following WAV file has two "players" playing a C major scale. One of the players is using the Just Scale, the other the Equal Tempered scale. Both start on exactly the same pitch. See if you can hear the notes where the pitches are different by listening for the beats. Major scales in different temperaments

Equal Tempered Scale - Table of frequencies

For a detailed list of historical temperaments see: "Tuning: containing the perfection of eighteenth-century temperament, the lost art of nineteenth-century temperament, and the science of equal temperament, complete with instructions for aural and electronic tuning," by Owen H. Jorgensen, Michigan State University Press, East Lansing, MI 1991. (This is an expensive reference book -- you might wish to look for a copy in your library rather than in a book store)

Frequencies for equal-tempered scale

This table created using A4 = 440 Hz Speed of sound = 345 m/s = 1130 ft/s = 770 miles/hr ("Middle C" is C4 )
Note Frequency (Hz) Wavelength (cm) C0 16.35 2100. # b C 0/D 0 17.32 1990. D0 18.35 1870. # b D 0/E 0 19.45 1770. E0 20.60 1670. F0 21.83 1580. # b F 0/G 0 23.12 1490. G0 24.50 1400. # b G 0/A 0 25.96 1320. A0 27.50 1250. # b A 0/B 0 29.14 1180. B0 30.87 1110. C1 32.70 1050. # b C 1/D 1 34.65 996. D1 36.71 940. # b D 1/E 1 38.89 887. E1 41.20 837. F1 43.65 790. # b F 1/G 1 46.25 746. G1 49.00 704. # b G 1/A 1 51.91 665. A1 55.00 627. # b A 1/B 1 58.27 592. B1 61.74 559. C2 65.41 527. # b C 2/D 2 69.30 498.

D2 D 2/Eb2 E2 F2 # F 2/Gb2 G2 # G 2/Ab2 A2 # A 2/Bb2 B2 C3 # C 3/Db3 D3 # D 3/Eb3 E3 F3 # F 3/Gb3 G3 # G 3/Ab3 A3 # A 3/Bb3 B3 C4 # C 4/Db4 D4 # D 4/Eb4 E4 F4 # F 4/Gb4 G4 # G 4/Ab4 A4 # A 4/Bb4 B4 C5 # C 5/Db5 D5 # D 5/Eb5 E5 F5 # F 5/Gb5
#

73.42 77.78 82.41 87.31 92.50 98.00 103.83 110.00 116.54 123.47 130.81 138.59 146.83 155.56 164.81 174.61 185.00 196.00 207.65 220.00 233.08 246.94 261.63 277.18 293.66 311.13 329.63 349.23 369.99 392.00 415.30 440.00 466.16 493.88 523.25 554.37 587.33 622.25 659.26 698.46 739.99

470. 444. 419. 395. 373. 352. 332. 314. 296. 279. 264. 249. 235. 222. 209. 198. 186. 176. 166. 157. 148. 140. 132. 124. 117. 111. 105. 98.8 93.2 88.0 83.1 78.4 74.0 69.9 65.9 62.2 58.7 55.4 52.3 49.4 46.6

G5 G 5/Ab5 A5 # A 5/Bb5 B5 C6 # C 6/Db6 D6 # D 6/Eb6 E6 F6 # F 6/Gb6 G6 # G 6/Ab6 A6 # A 6/Bb6 B6 C7 # C 7/Db7 D7 # D 7/Eb7 E7 F7 # F 7/Gb7 G7 # G 7/Ab7 A7 # A 7/Bb7 B7 C8 # C 8/Db8 D8 # D 8/Eb8
#

783.99 830.61 880.00 932.33 987.77 1046.50 1108.73 1174.66 1244.51 1318.51 1396.91 1479.98 1567.98 1661.22 1760.00 1864.66 1975.53 2093.00 2217.46 2349.32 2489.02 2637.02 2793.83 2959.96 3135.96 3322.44 3520.00 3729.31 3951.07 4186.01 4434.92 4698.64 4978.03

44.0 41.5 39.2 37.0 34.9 33.0 31.1 29.4 27.7 26.2 24.7 23.3 22.0 20.8 19.6 18.5 17.5 16.5 15.6 14.7 13.9 13.1 12.3 11.7 11.0 10.4 9.8 9.3 8.7 8.2 7.8 7.3 6.9

(To convert lengths in cm to inches, divide by 2.54)

Conical vs Cylindrical Bores

Why do they sound different?

A soprano saxophone and a clarinet are about the same size and are both single reed instruments, but they sound completely different. In addition, the range of a clarinet is rougly one octave lower. While there are other more subtle differences, the main difference is that a clarinet has a cylindrical bore and a saxophone has a conical bore. This difference causes a large difference in the resonant modes, and hence the overtones which can be excited. For a cylindrical bore, the amplitude of the pressure variations for resonant modes are well described by sine waves. At the ends of an open finite cylinder (neglecting end effects), the pressure variations should be zero (i.e. the ends are a pressure node). At a closed end, the pressure variations should be a maximum (i.e. an "anti-node"). For a conical bore, the amplitude of the pressure variations are not simple sine waves, but are described by sin(x)/x, where x represents a distance (in appropriate units) along the cone, and x = 0 is the apex. As is the case for the closed cylinder, a pressure anti-node must be present at the closed end of the cone, which occurs automatically for the function sin(x)/x, and a pressure node should be present at the open end. The figure below illustrates (schematically) the pressure variations for a cylinder open at both ends, an open cone, and a cylinder closed at one end. The end conditions for the cylinder closed at one end can only be met by the odd harmonics, and hence all even harmonics are missing from the sound. On the other hand, the cone and the cylinder open at both ends contain both the odd and even harmonics and when the same length, will have the same resonant frequencies. The frequency for the nth harmonic is fn = n*f1. In addition, note that to obtain the same fundamental frequency, f1, the cylinder closed at one end is 1/2 the length of the cylinder open at both ends. Hence, if the two were the same length, the closed cylinder would play an octave lower than the open cylinder.

To a first approximation, the clarinet can be considered as having a cylindrical bore closed at one end (by the reed/mouthpiece), whereas a saxophone has a (truncated) conical bore (with the apex near the mouthpiece).

For more detail, including the case of a partial cone with openings at both ends, see: "The conical bore in musical acoustics," R. D. Ayers, L. J. Eliason, and D. Mahgerefteh, American Journal of Physics, Vol 53, No. 6, pgs 528-537, (1985).

Flute Finger Hole Locations

Determining approximate finger hole locations for a simple 6 hole flute is somewhat complicated. Much of what follows is due to the analysis of A. H. Benade (see "Fundamentals of Musical Acoustics," A. H. Benade, Oxford Univ Press, or "The Physics of Musical Instruments," N. H. Fletcher and T. D. Rossing, Springer-Verlag.)

Localizao dos orifcios tonais em uma flauta

Determinar as localizaes aproximadas dos orifcios tonais para uma flauta simples de 6 orifcios um pouco complicado. Muito do que se segue devido anlise de A.H. Benade (consulte "Fundamentos de Acstica Musical," AH Benade, Oxford University Press, ou "A Fsica de Instrumentos Musicais", NH Fletcher e TD Rossing, Springer-Verlag.) The wavelength of the sound produced is determined by the flute tube and tone holes. For a pipe of length L, open at both ends, and ignoring end effects, the wavelength of the sound is twice the length of the tube. The frequency produced is given by dividing the speed of sound (345 m/s) by the wavelength. O comprimento de onda do som produzido determinado pelo tubo da flauta e pelos orifcios tonais. Para um tubo de comprimento L, aberto em ambas as extremidades, e ignorando os efeitos finais, o comprimento da onda sonora o dobro do comprimento do tubo. A frequncia produzida dada dividindo a velocidade do som (345 m / s) pelo comprimento de onda. The method described here is to estimate an effective length for a real (cylindrical) tube taking into account end effects, the size of the tone holes, etc. O mtodo descrito aqui consiste em estimar um comprimento efetivo para um verdadeiro tubo (cilndrico), tendo em conta os efeitos finais, o tamanho dos orifcios tonais, etc If the fundamental (all tone holes closed) corresponds to an effective length Leo, then the desired effective lengths for a six (or fewer) hole flute to produce notes of the (Western major) scale are given by:

Se a fundamental (todos os orifcios tonais fechados), corresponde a um efetivo comprimento Leo, ento os comprimentos efetivos desejados para que uma verdadeira flauta tubular de seis (ou menos) orifcios, produza notas da escala (ocidental maior), so dados por:
Nota Nmero 0 1 2 3 4 5 6 Comprimento Nota Unidade de Leo Do 1 Re 8/9 Mi 4/5 Fa 3/4 So La Ti 2/3 3/5 8/15

The actual physical length of the flute will be shorter than Leo and the distance to the tone holes (from the blow end) will be shorter than the values calculated in the table. O real comprimento fsico da flauta ser mais curto do que Leo e a distncia para os orifcios tonais (a partir da extremidade do sopro), ser menor do que os valores calculados na tabela. As described by Benade, the end effects (at both the blow hole and at the first open hole) act (approximately) as an additional length which must be subtracted from the calculated effective lengths to get the physical lengths desired. As long as you don't deviate too much from typical flute proportions, these approximate calculations should get you pretty close. If you are serious about getting your flute in tune, plan on making at least two flutes. First, make your flute with no tone holes and adjust the length to match the desired lowest note. The blow hole is typically 1/2 to 2/3 the tube's inner diameter, and the stopper will be about an equal distance from the center of the blow hole. Then get out your ruler and measure the following: t = wall thickness of the tube Lo = length of the tube from the center of the blow hole to the open end. a = inside radius of tube (2a = inside diameter), and d = distance from center of blow hole to stopper (see below) Then, Leo = (Lb + Lo + 0.6 a) where

0.6 a = approximate end correction at the open end and Lb is the length correction at the blow end, which is not known. Lb can be determined approximately by: b = radius of blow hole Heff = (t+1.7b) Lb = Heff (1 - e) (a/b)2 where e = fraction of blow hole covered by the player's lip e is typically 0.25 to 0.33 Note: using a single Lb for all notes is a reasonable approximation IF d is close to 0.37 Lb Now use your value of Leo and the table above to get desired effective lengths for each of the hole positions, measured from the center of the blow hole to the center of the tone hole. Call these Lei where "i" corresponds to the first column in the table. (e.g. Le1 = 8/9 Leo, etc). From each Lei, subtract (Lb + 0.6a) to get your first approximate hole positions, Li. The actual position for each hole will depend on the tone hole size and the position of any other open tone holes. Hence, you have some leeway to choose one or the other of these. Smaller tone holes will give a mellower sound and larger holes give a brighter (and louder) sound. Your holes do not need to be all the same size, so it is possible to make some choices which affect the ergonomics of your flute. Below I will assume that all the tone holes are fixed in size, and only their positions are to be adjusted. If this results in tone hole positions which are uncomfortable (or unusable) then adjust one or more of the tone hole sizes and recompute. Now compute the first corrections: Linew = Lei - Lb - Lc where if i = 1 (for Re) then Lc = (t+1.5b)/( (b/a)2 + (t+1.5b)/D ) where D = Lo - L1 else (for all other holes) Lc = s [ ( 1 + 2(1.5b + t) (a/b)2 /s )1/2 - 1 ] where b = tone hole radius (2b = diameter) of first open hole from the blow end, s = 1/2 of the spacing between first and second open holes from the blow end [i.e. s = ( Li-1 - Li )/2 ] endif

For each of these calculations, you will need to repeat them several times. For example, for the first tone hole, an initial value of D is used which was computed from the previous approximate positions. Once the correction is found, compute the new value of D and recompute the correction. Continue until the answer doesn't change much. For the other holes, recompute s using the corrected position to obtain a better correction. Do this one hole at a time. If you correct all the holes, then go back and redo the calculations for all the holes, you can get into numerical trouble in some cases. Now drill your holes. If a note is a bit flat, you can enlarge the first open hole (from the blow end) a bit to sharpen it. If you drill your holes a little small to begin with, you might be able to bring your flute into tune without having to make a second one. Note: if you have particular thick walls on your flute (like some wood flutes) you will need to add another correction. Use Li = Lei - Lb - Lc - Lt where Lt = t (b/a)2 /4 times the number of closed holes between the first open hole and the blow end. You will also need to shorten your tube a bit as the presence of the closed holes will flatten the fundamental as well. Now check the tuning of your flute, and estimate adjustments to make your second flute. That is, if your note is 3% flat, move the hole 3% closer to the blow hole, etc. A frequency counter is best for this, but if necessary, you can do it "by ear" (e.g. by comparing to another instrument which is in tune and listening to the beats). If you see systematic problems (e.g. Re is a little flat, Me a little flatter, and by the time you get to Ti it is very flat) then your value of Lb can be adjusted to fix all the holes at once. For a fipple flute (e.g. a recorder) the correction at the mouth can be approximated by Lb = 2.3 a2 / (area of fipple opening)1/2 though this is only a rough approximation. If you just want to calculate, or need to check your calculations, then see the flutomat javascript by Pete Kosel