Ed Note: Dr. James C. Prodan, Assistant Professor of Oboe at The University of Akron, Akron, Ohio, received his D.MA. from The Ohio State University in 1976. This article is taken from his doctoral essay, submitted as a portion of the dissertation requirement for that degree.
The Tone is a castle in the air.
Imagination is its architect.
Nerves carry out the plans.
Muscles are the laborers.
The soul inhabits it.Stevens Hewitt, "Method for Oboe"
The oboe has been the subject of many investigations during this century. These studies have endeavored to improve the playing, teaching, construction, and understanding of its development and place in the world of music. Articles have been written in leading educational and professional journals to help the beginning oboist, the experienced performer, and the teachers of beginners, high schoolers, and college students. The majority of these efforts have been aimed toward problems of the instrument other than the reed. Embouchure fingerings, vibrato, intonation, tone quality, hand position, posture, technique and breath support have been discussed repeatedly by prominent players and teachers.
The lack of many technical studies on the reed is obvious yet understandable. The oboe reed is a very personal piece of equipment. There are probably as many different ways of making the reed as there are reed makers. Each reed maker constantly endeavors to duplicate or improve his best efforts, experimenting and changing his procedures in the hope of winning the battle with Arundo Donax.
Unlike most other wind instruments, which have some sort of permanent mouthpiece or embouchure plate, the whole "mouthpiece" of the oboe must be constructed of a material that is not of a permanent nature. The oboist attempts to make a reed that will satisfy his personal requirements but, because of the destructibility of the oboe reed cane, it is important to be able to duplicate each reed as closely as possible.[1]
Much of our knowledge of the reed has been passed from teacher to student, to his student and so on. Trial and error has been our main teaching method, yet is has served us well, however slowly. It is past time that the oboe reed receive the investigation awarded the rest of the instrument and its pedagogical problems.
A few books have been published that deal with the reed. There are, in English Sprenkle and Ledet's, The Art of Oboe Playing, published by Summy-Birchard (1961), Mayer and Rohner's, Oboe Reeds - How To Make and Adjust Them, published by The Instrumentalist (1953), and the recent book by Elizabeth and Peter Hedrick, Oboe Reed Making - A Modern Method published by Swift-Dorr (1972). The wider spread use of these manuals is evidence of the need of having texts that deal with reed making and adjusting. The small number of these books also shows that reed making is difficult to handle in book form. Most sources suggest as does Marjorie Jackson in her dissertation that "it is imperative that the study of this art be conducted under the supervision of a fine teacher and developed by the pupil through thousands of failures and much experimentation."[2] Most oboists agree that good reed making can be achieved only through many failures and much frustration. One of the difficult aspects of reed making for students is the pitch of the crow of the finished reed. The crow is the sound obtained by blowing through the reed without the instrument. It is recommended by almost all American texts including Sprenkle and Ledet (p. 82) in The Art of Oboe Playing and Hedrick (p. 23) in Oboe Reed Making - A Modern Method that the pitch of the crow of the reed be at C natural in octaves. Stevens Hewitt in Book One of his new Method published by American Composers Alliance (1972) says the reed must crow in octave C naturals. Teachers with whom I have studied including William Baker of the Ohio State University, John Mack of the Cleveland Orchestra, and James Caldwell of Oberlin Conservatory of Music, adhere to this principle. Yet students are very willing to settle for much less in their reed making (anything from B flat up to the C natural) hoping that the embouchure will make up the difference in intonation and that the tone quality will not be affected. This of course, creates a host of problems in developing a good embouchure, good tone, and good support. Students often fail to realize the need for a well-pitched reed early in their reed making efforts.
Before the actual experiment is discussed it will be helpful to explain the construction of my reeds and what properties I search for in a good reed. The discussion in Chapter II is not intended as a complete guide, but as a summary of my reed making and a general guide to the American cut reed. Chapter III will discuss the tone quality of the oboe. This discussion will not conclusively define a good oboe tone quality. It will give many descriptions of oboe tone and show the difficulty of defining a good oboe tone due to the many variations in tone quality around the world
The purpose of this project is to show that the pitch of the crow of the reed does affect the tone quality of the instrument, that a reed with a crow of C natural will be a better sounding reed than one whose crow is flatter than C natural.
This project is intended not as an exhaustive investigation, but as a preliminary study of just one important facet of the chief concern of all oboists - the reed. Research on this subject is long overdue. It is hoped that this effort will raise questions that will spur others to investigate the reed further.
For my own playing, I use a standard American cut reed. It is the combined and individualized product of my teachers: William Baker, James Caldwell, and John Mack.
The cane measures between 10.25 and 10.5 mm. in diameter and is gouged at .60 mm. in the center and .45 mm. on the sides at the edge of the shaper. The shaper is from William Brannen (Evanston, Illinois) and is his "TYPE-X." The shaped cane is wrapped on a 47 mm. nickel-silver tube in such a way that the cane extends 27 mm. beyond the tube giving the reed blank a total length of 74 mm. The Hedrick manual explains the wrap well.
Holding the cane over the tube between thumb and index finger. . . and holding the spool of thread in the other hand, make three winds of the thread around the cane so that the top of the wind is exactly even with the end of the tube. The blade of the piece of cane should then be slightly displaced in the direction opposite to the pull of the string This way the blades will be tightly bound against each other, thus promoting vibration and helping to prevent leakage of air between the blades.[3]
The displaced or slipped blades should always show the left blade over the right blade as the reed is observed from the side. (see Illustration 1.) The reed should never be wrapped beyond the end of the tube because this will not permit the reed to crow up to the pitch of C natural.
The configuration of the slipped blades allows for an adjustment in response and intonation without the knife, a characteristic that is very helpful in a rehearsal situation. By slipping the blades further apart the opening will become smaller, the pitch of the reed will go up and the response, will be easier. By slipping the blades less, the reed will be more open, flatter in pitch, and less responsive (see Illustration 2 and 3.)
These details are stressed because the main thrust of this study concerns the intonation of the reed.
The finished reed will have the approximate measurements, shown in Illustration 4. The scrape of the back of the reed begins four millimeters above the, string. The side of the tip begins 19 millimeters above the string and the center of the tip begins between 20 and 21 millimeters above the string. This configuration forms the upside down "V." The tip extends upward to between 22 or 23 millimeters. Because no two reeds are exactly alike, the measurements will vary slightly for each reed. These measurements however, provide a guide for the reeds used in this project.
It is difficult to describe in words the final appearance of a reed. Pictures can provide a much more accurate conception. The photographs found in the Hedrick manual serve as good examples of my reeds (see Illustrations, 5, 6, and 7). It is hoped that the corners of the tip that are missing in these photographs would be retained. Reeds with missing corners often do not respond well in the low range due to the air leaking through the reed before an attack.
As Illustration 5 shows, the tip is very thin, especially at the upper edge and its cross section is shaped as an inverted V or U. The blend area gradually thickens before the hump. The hump is somewhat thicker than the remainder of the back of the reed, yet the spine is clearly visible. Bark should be visible on both rails or sides of the reed (not just on the left as is shown).
As stated earlier, the ideal reed should crow at the pitch of C natural in three octaves. The reed's construction should permit it to vibrate freely enough for easy attacks in the low register, but it should have enough resistance to produce a stabilized tone in the very high register. The resulting sound should be full and round with good projection when playing throughout the dynamic range. Tonguing should be easy. An oboe reed having all these qualities is of course, usually the product of several attempts and many adjustments. Inexperienced reed makers must spend even more time at the reed table, and usually their results are unsatisfactory in one or more qualities of the reed, especially intonation.
Oboe tone quality varies greatly around the world. Any discussion about oboe tone quality without actually hearing an oboe or several different players will still leave the question of "what is a good oboe tone" unanswered. The following is a collection of several definitions or explanations of the oboe tone. The statements vary greatly and will certainly not answer the question of oboe tone quality. They will, however, show the complexity of the issue.
. . . the quality or timbre of a musical sound is determined, chiefly, by the number, intensity, and distribution of the partials that enter into its composition.[4]
The oboe's conical bore allows all the harmonics to be available in a tone, as opposed to the clarinet's cylindrical bore and the stopped column effect which allows only the odd numbered harmonics (fundamental, third, fifth, etc.) in the tone to predominate.
The formant theory has been advanced as a factor determining tone quality. F. A. Saunders offers the following concerning the lack of effect that formants have on tone quality.
There is a widely held belief that musical instruments owe their individual qualities to the presence of formants. A formant may be defined as a region of frequency such that every partial tone that falls within it is relatively strong, no matter what fundamental it belongs to. Such a reinforcement could hardly be caused by any action but resonance. The resonances of an air column are fairly sharp, especially at high frequencies, if we define the sharpness as the width of the response in semitones. It is doubtful if a single sharp resonance could exert a marked effect on tone quality; at least, this does not happen in violin tones. [5]
Saunders' analyses of wind instruments show the lack of strong formants which could contribute to tone quality in woodwind instruments. "There is no fixed arrangement of relative strength of partials in what we call oboe tone."[6]
Many factors contribute to timbre. In his tests on the recognition of timbre, Kenneth Berger showed that when the attack and release of a tone were removed, and when all the upper partials were removed, the recognition of timbre of a wind instrument became much more difficult.
Of the several factors investigated in this experiment the listeners correctly recognized wind-instrument tones best when both rise and decay and partials were present. Unaltered tones were correctly recognized by listeners in 178 of 300 cases. Tones filtered of all but the fundamental component were recognized correctly in only 55 out of the total of 300 cases. That the attack and release of the tone seem to give the listener some important recognition clues of the instrument is revealed by correct recognition scores better than the filtered condition but worse than the unaltered condition.[7]
Realizing that volume changes cause timbre changes, Berger attempted to remove this variable by asking his performers to play at the level of 80dB which they could monitor visually on a General Radio sound level meter. Further, Philip Bate says "the source of instrumental timbre is now known to lie in the shape and dimensions of the bore and its array of note-holes plus the cooperation of various properties inherent in the reed."[8]
Walter Wehner's tests on the intonation of four different profiles of reeds is based on the premise that intonation is the prime concern of the young oboist in an ensemble. His tests show that different profiles of reeds offer very different intonation levels and characteristics. In his tests no concern was demonstrated for tone quality and none of the reeds were of the American style cut. They were of French and German cuts. Of course, proper intonation is very important in an ensemble, but what ensemble leader is not concerned at the same time about tone quality?
It has been shown that many factors contribute to tone quality, such as the bore; the number, strength, and distribution of the partials; attack and release, volume and the reed. A problematical question that has not been discussed yet concerns the definition of "good oboe tone." In his article in Grove's Dictionary of Music and Musicians[9] Bate presents a graphic analysis of an oboe tone by F. A. Saunders. In tests at Harvard, Saunders used a Loree instrument played by Dr. A. Sprague Coolidge. In the article Bate does not attempt to judge the quality of the tone, nor does Saunders (see Illustration 8). The graph is the only presentation on quality. Saunders, in another source, does mention several points about the tone.
The fundamental rises in importance at higher pitches, and in the top notes it is the strongest No. 2 is second, No. 3 third; in other words the number of the partial tells its relative loudness. Assuming, as is likely, that this arrangement is common with average string tones, then we may say of the oboe,and of all the other instruments tested, that the highest tones have a string quality. In the lowest oboe tones partials Nos. 6 and 5 are strongest of all; with rising pitch No 4. takes the lead, then No. 3, and in the upper half of the range No. 2 dominates until the fundamental rises to overcome it. The highest oboe partial (with a strength of 10 dB) was No, 18. The lowest tone showed 17 partials; an octave above there were 9; another octave only 5. The fundamental showed a marked weakness in the region between the first and second registers, which seems to be a common defect. Partial No. 3 was unaffected in this region, and 2 and 4 were also fairly strong throughout. [10]
Here are several statements on tone quality by practical performers. These statements are much less scientific, yet are colorful and descriptive.
When a tone is pure, the lower harmonious overtones are heard. When the tone is forced or spread, the higher discordant harmonics predominate, causing a hard metallic quality. When overtones are lacking, tone is hollow, stuffy, and wooden. [11]
I don't feel limited to finding enjoyment in only one particular kind of sound, but there is one characteristic that leaps to mind here, and that is something I would call integrity. Conviction in the sound. Most all the best players seem to have that quality - even though their tones may vary from one player to another quite a bit. It sounds corny, but I like to say tone is a matter of life and depth; to have these two characteristics in your sound. [12]
The bitter-sweet oboe which is first heard marshalling the orchestra to tune, continues as the music proceeds, to assert its small but inexpressively poignant voice whether it is heard singing plaintively to a hushed accompaniment or whether under the passionate surge of the strings it is heard calling, as it seems, from the innermost secret places.[13]
"While the oboe produces a 'pastoral,' slightly quaint and nasal sound of fairly constant quality . . ."[14]
For a more specific description we turn to Dr. Bernard Hague who presents the following:
A typical tonal spectrum for the oboe is shown . . . with the following remarkable features: (i) there are no important overtones at any part of the compass with a frequency exceeding 7,000 cycles per second; (ii) the low and middle registers are rich in overtones up to about the sixteenth harmonic clearly explaining why Helmholtz failed to imitate oboe tone by the synthesis of only eight harmonic terms; (iii) in the low register the first five overtones are approximately equal to the fundamental. [15]
The graph shown in the illustration represents tones played at a lower pitch level (A=435) than is used now (see Illustration 9).
Harry Olson explains the oboe tone (see 1 Illustration 10) in this manner:
The acoustic spectrums of three tones of the oboe are shown . . . The major portion of the energy resides in the fundamental and overtones in the frequency range from 500 to 1,500 cycles. The tone of the oboe is bright and reedy and sometimes like a flute in brilliance. The low tones rich in powerful harmonics therefore are reedy in character. The tone can also be very incisive. [16]
These discussions and illustrations still do not give a clear definition of the good tone. Perhaps John Mack's statement "I don't feel limited to finding enjoyment in only one particular kind of sound . . . " should be taken as a clue that there can be no clear definition since national styles of playing and personal tastes differ so greatly. It is not the purpose of this project to attempt to define good tone, but simply the qualities of sound made by | reeds of different intonation characteristics.
The tests were conducted in a laboratory of the Bio-Medical Engineering Center at 2015 Neil Avenue on the campus of The Ohio State University. The apparatus was set up and calibrated by Richard M.Campbell of the Center. The equipment used consisted of a model UA-14A Ubiquitous Spectrum analyzer made by Federal Scientific, a Type 502 Dual-Beam Oscilloscope made by Tektronix, Inc., a model 322 Dual Channel DC Amplifier-Recorder made by Sanborn, a model 635A microphone made by Electro-Voice, and a model 675 Stroboconn Scanning Unit made by C. G. Conn, Ltd.
The complete project consisted of two separate experiments. The first test was performed on only one instrument, thus removing the possibility that differences in quality and pitch could be caused by different instruments. The test was completed in a short period of time in order to minimize the effects of changing weather conditions on the reeds. The reeds were, of course, on different tubes, however, they all were the same length (47 mm.) and fit my mandrel alike. Attack and release were not considered in the test. The tones were played into the apparatus and held for a short period, then frozen and recorded. Variations in volume were limited by playing at an easy mezzo-forte level throughout the test. It was found that variation in the proximity of the microphone greatly changed the response registered in the apparatus. An attempt was made to keep the microphone at a constant distance from my instrument, approximately four inches from the bell. Any changes of pitch also affected the harmonic pattern. Pitch variation was minimized by the use of the stroboconn. The tone was not frozen into the apparatus until the pitch was true.
The first test consisted of playing a two-octave D major scale on each of four reeds and recording the sound spectrum of each tone. As mentioned previously, the tones were played at an easy mezzo-forte level and checked with the stroboconn. Each note was held a few moments before freezing the tone into the oscilloscope for recording on graph paper. Four different reeds were played in the test. The first was a "good" reed, new but broken in. It crowed two C naturals (see Illustration 11) and played with a rich full tone quality.
The second reed had been used in a few rehearsals. It crowed one C natural (see Illustration 11). It still played fairly well, but was old and "tired." The third reed played crowed two B naturals (see Illustration 11). It is a very good example of the reeds students play. It required much embouchure pressure to play it in tune, and the sound was not as rich and flexible as the first reed. A fourth reed was played to show the difference between the custom-made reed and the commercial reed. This commercial reed was made by Selmer and had a French cut. Some adjustments had to be made before the test to get it to respond at all. The crow of the reed was wild, having no constant pitch center. It was very difficult to make this reed play in tune. This reed's basic pitch wavered between B flat and B natural with other pitches appearing and disappearing, also (see Illustration 11). Each reed was soaked in water for a few minutes and warmed up before the test.
The sound spectrum graphs of test I demonstrate a great difference in the tone quality generated by the various reeds. The reeds do show similarities for a given pitch, but a closer scrutiny gives evidence that the partials definitely form different patterns.
Before looking closely at the graphs, a few general statements should be made of the results. Perhaps these statements will prove more valuable than a detailed explanation of each pitch, due to the fact that no two reeds will produce exactly the same sound spectrum.
In general, the array of partials in reed one is more balanced than in the other reeds.[l7] In other words the array of the good reed has more partials making a significant contribution to the tone quality. This can be seen easily in the graph of the lowest tone played, the D natural, (see Graph No. I) and in the high A natural (see Graph No. 12, page 47). The more balanced array certainly seems to contribute to the richer sound of the good reed.
Another point related to the balance of the array is the number of very strong partials in the sound. The number of strong partials is greatest in the good reed, and the lesser reeds have only one very strong partial or just a few. This is supported by statements on tone quality in Chapter Three. The graphs on pages 43, 44 and 45 and the statements by Dr. Bernard Hague and Harry Olson on page 44 show that many strong partials are present in the tones, especially in the lower register. The good reed in some cases has up to four or five partials making a strong contribution to the sound. This is more evident in the upper octave but still is a strong consideration in the lower octave. The French cut reed should be especially noted here. In the lower octave it usually had one or two partials very strong while the remainder were very weak, such as F sharp (see Graph No. 3, page 47) and A (see Graph No. 5, page 48). In the upper register, however, it was very similar to the American cut reeds.
In no instance was the fundamental the prominent partial in the sound. Usually the third, fourth, fifth, or even sixth partial was the strongest of the tone. In the upper partials a marked tapering off in strength occurs, but there are definite exceptions to this such as G natural, F sharp2, B natural2, and C sharp2. In these instances a high partial shows strength over some neighboring partials.
In all these points the illustration of Saunders (Illustration 8, p. 43) and that of Hague (Illustration 9, p. 44) confirm my observations. In some cases, the spectra of the lesser reeds conform nicely to that of the good reed. However, it is important to this study that the degree of conformity be observed. For instance, the graph of F sharp1 (see Graph No. 3, page 47) shows that all four reeds are superficially similar. But it is easily seen that the first reed has more very strong partials and that even the upper partials are much stronger than in reed two or three. The graph of A1 (see Graph No. 5, page 47) also shows an apparent similarity. Reed one, however, has more partials making strong contributions.
Timbre is largely determined by the strength of the major partials combined with the number and strength of other contributing partials. The difference in strength and number of partials in the reeds indicates that the intonation of the reed's crow does affect the tone quality. Looking at the tones individually supports the above general statements. Tables 1-15 identify the partials TABLES HERE in each reed which were strong enough to reach the tenth, fifteenth, or twentieth mark on the spectrum graphs. The totals at the right show the number of partials reaching the tenth, fifteenth, or twentieth mark on the spectrum graphs. In most cases these tables confirm the observations made above. The first exception is G natural1. The first reed does show more partials contributing to the tone quality at mark ten. But by mark fifteen the third reed tuned to a B natural1 (see Table 4, p. 49) has more contributing partials.
TABLE 1
CONTRIBUTING PARTIALS OF d' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 9
Reed 2 6
Reed 3 7
Reed 4 6
Mark 15 or more Reed 1 5
Reed 2 4
Reed 3 3
Reed 4 2
Mark 20 or more Reed 1 2
Reed 2 1
Reed 3 1
Reed 4 2
TABLE 2
CONTRIBUTING PARTIALS OF e' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 8
Reed 2 7
Reed 3 7
Reed 4 10
Mark 15 or more Reed 1 7
Reed 2 3
Reed 3 6
Reed 4 7
Mark 20 or more Reed 1 4
Reed 2 1
Reed 3 1
Reed 4 4
TABLE 3
CONTRIBUTING PARTIALS OF f#' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 8
Reed 2 8
Reed 3 6
Reed 4 9
Mark 15 or more Reed 1 6
Reed 2 4
Reed 3 3
Reed 4 3
Mark 20 or more Reed 1 3
Reed 2 3
Reed 3 2
Reed 4 2
TABLE 4
CONTRIBUTING PARTIALS OF g' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 8
Reed 2 8
Reed 3 7
Reed 4 8
Mark 15 or more Reed 1 3
Reed 2 4
Reed 3 4
Reed 4 3
Mark 20 or more Reed 1 3
Reed 2 2
Reed 3 4
Reed 4 2
TABLE 5
CONTRIBUTING PARTIALS OF a' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 6
Reed 2 5
Reed 3 4
Reed 4 4
Mark 15 or more Reed 1 6
Reed 2 4
Reed 3 3
Reed 4 3
Mark 20 or more Reed 1 5
Reed 2 4
Reed 3 3
Reed 4 2
TABLE 6
CONTRIBUTING PARTIALS OF b' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 7
Reed 2 5
Reed 3 7
Reed 4 6
Mark 15 or more Reed 1 5
Reed 2 5
Reed 3 4
Reed 4 5
Mark 20 or more Reed 1 4
Reed 2 3
Reed 3 2
Reed 4 2
TABLE 7
CONTRIBUTING PARTIALS OF c#' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 5
Reed 2 6
Reed 3 6
Reed 4 6
Mark 15 or more Reed 1 5
Reed 2 5
Reed 3 6
Reed 4 6
Mark 20 or more Reed 1 3
Reed 2 4
Reed 3 4
Reed 4 3
TABLE 8
CONTRIBUTING PARTIALS OF d' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 7
Reed 2 6
Reed 3 6
Reed 4 6
Mark 15 or more Reed 1 5
Reed 2 4
Reed 3 4
Reed 4 3
Mark 20 or more Reed 1 5
Reed 2 3
Reed 3 2
Reed 4 3
TABLE 9
CONTRIBUTING PARTIALS OF c2 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 10
Reed 2 7
Reed 3 2
Reed 4 8
Mark 15 or more Reed 1 5
Reed 2 2
Reed 3 1
Reed 4 3
Mark 20 or more Reed 1 4
Reed 2 1
Reed 3 1
Reed 4 2
TABLE 10
CONTRIBUTING PARTIALS OF f#2 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 7
Reed 2 4
Reed 3 6
Reed 4 3
Mark 15 or more Reed 1 4
Reed 2 4
Reed 3 2
Reed 4 3
Mark 20 or more Reed 1 2
Reed 2 3
Reed 3 2
Reed 4 1
TABLE 11
CONTRIBUTING PARTIALS OF g2 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 9
Reed 2 5
Reed 3 8
Reed 4 5
Mark 15 or more Reed 1 6
Reed 2 3
Reed 3 4
Reed 4 3
Mark 20 or more Reed 1 4
Reed 2 3
Reed 3 3
Reed 4 3
TABLE 12
CONTRIBUTING PARTIALS OF a2 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 8
Reed 2 5
Reed 3 4
Reed 4 5
Mark 15 or more Reed 1 6
Reed 2 2
Reed 3 2
Reed 4 4
Mark 20 or more Reed 1 3
Reed 2 1
Reed 3 1
Reed 4 2
TABLE 13
CONTRIBUTING PARTIALS OF b2 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 5
Reed 2 6
Reed 3 5
Reed 4 4
Mark 15 or more Reed 1 5
Reed 2 4
Reed 3 4
Reed 4 3
Mark 20 or more Reed 1 3
Reed 2 3
Reed 3 3
Reed 4 3
TABLE 14
CONTRIBUTING PARTIALS OF c#' AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 6
Reed 2 7
Reed 3 5
Reed 4 6
Mark 15 or more Reed 1 3
Reed 2 3
Reed 3 4
Reed 4 6
Mark 20 or more Reed 1 3
Reed 2 2
Reed 3 4
Reed 4 4
TABLE 15
CONTRIBUTING PARTIALS OF d3 AT MARKS 10, 15, AND 20 ON THE SPECTRUM GRAPHS
Total*
mark 10 or more Reed 1 6
Reed 2 7
Reed 3 4
Reed 4 7
Mark 15 or more Reed 1 5
Reed 2 4
Reed 3 4
Reed 4 5
Mark 20 or more Reed 1 3
Reed 2 3
Reed 3 3
Reed 4 5
Another similar exception is C sharp1 (see Table 7, p. 49). This tone, even at mark ten shows five strong partials in the good reed and six in the bad reed. The octave above (see Table 14, p. 50) shows six strong partials in reed one and five strong partials in reed three at level ten, but by mark fifteen only three strong partials remain in the good reed, with four partials contributing to reed three.
All remaining tones agree with the statements made earlier in this chapter. The first reed tuned to octave Cs contributes more strong partials to the tone quality than the other reeds tested. In most cases the second reed shows better results than the third reed. Tables 1-15 on pages 48-50 show that at mark ten on the spectrum graph reed one was best on nine out of the fifteen possible or 60%. Reed two was best on six notes or 33.3%, reed three best on two or 13.3% and reed four best on five or 30%. The possibility of an equal number of strong partials in various reeds allows for more than one reed to show best results in the tables and for the percentages to total more than 100%. At mark fifteen reed one is best twelve times or 80%, reed two is best four times or 26.7%, reed three is best three times or 20%, and reed four is best six times or 40%.
These tables again show that the reed crowing octave Cs by far exceeds the other reeds in the number of strong partials contributing to the tone quality. The second reed also usually shows better results than reed three. The surprise was that the French cut reed, number four, had shown so many strong partials at various marks in spite of its wild crow, poor response, and intonation problems. This reed also sounded very wild, bright, and uncontrollable. It was far from the tone quality of reed number one.
In the second phase of this project the format was changed. Instead of having just one player there were three oboists. They were professional players and teachers. Each played his or her own instrument and reeds. All three players used Loree oboes. The serial numbers were BG17, DI38, and DT01. The equipment used for this test was identical to that of the first test, except for the addition of a model 1022 reel to reel tape recorder made by Magnacord and a playback speaker which had no identifiable markings. The tape recording of the events of test II are in the holdings of the Ohio State University Library.
Each player was asked to bring three reeds: one pitched high, one pitched low and one pitched at what was felt to be in tune. Each expert made the reeds for the test. The players were asked to warm up on all three reeds and to declare which reed was high, low, or good. Each was then asked to crow first the high reed then the low reed, then the good reed for measurement on the stroboconn. The experts were not looking at the strobe. This removed the chance of unconsciously adjusting the crow to the stroboconn. Table 40 shows these results. The players Table 40 here then played a D major arpeggio through the apparatus in the same manner as the first test. Each tone was played into the stroboscope until the pitch was true and frozen into the oscilloscope. It was then recorded on the graph paper. Each player used this procedure on all three reeds.
TABLE 40 PITCH OF REEDS EXPERT No. 1 Reed 1 high-pitched reed C+25 cents Reed 2 low-pitched reed B+50 cents Reed 3 well-pitched reed C+10 cents EXPERT No. 2 Reed 1 high-pitched reed C+34 cents Reed 2 low-pitched reed C-26 cents Reed 3 well-pitched reed C-20 cents EXPERT No. 3 Reed 1 high-pitched reed C+48 cents Reed 2 low-pitched reed B+40 cents Reed 3 well-pitched reed C-27 cents
TABLE 19
CONTRIBUTING PARTIALS OF d1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 5
Reed 2 6
Reed 3 4
Mark 10 or more Reed 1 2
Reed 2 4
Reed 3 2
Mark 15 or more Reed 1 0
Reed 2 0
Reed 3 2
TABLE 20
CONTRIBUTING PARTIALS OF f#1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 3
Reed 3 4
Mark 10 or more Reed 1 1
Reed 2 2
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 2
TABLE 21
CONTRIBUTING PARTIALS OF a1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 4
Reed 3 4
Mark 10 or more Reed 1 2
Reed 2 2
Reed 3 3
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 2
TABLE 22
CONTRIBUTING PARTIALS OF d2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 4
Reed 3 5
Mark 10 or more Reed 1 1
Reed 2 1
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 1
Reed 3 2
TABLE 23
CONTRIBUTING PARTIALS OF F#1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 2
Reed 3 1
Mark 10 or more Reed 1 2
Reed 2 1
Reed 3 1
Mark 15 or more Reed 1 2
Reed 2 1
Reed 3 1
TABLE 24
CONTRIBUTING PARTIALS OF a2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 3
Reed 3 2
Mark 10 or more Reed 1 2
Reed 2 1
Reed 3 2
Mark 15 or more Reed 1 2
Reed 2 1
Reed 3 2
TABLE 25
CONTRIBUTING PARTIALS OF d3 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 2
Reed 2 2
Reed 3 2
Mark 10 or more Reed 1 1
Reed 2 1
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 1
Reed 3 1
TABLE 26
CONTRIBUTING PARTIALS OF d1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 3
Reed 3 5
Mark 10 or more Reed 1 1
Reed 2 2
Reed 3 4
Mark 15 or more Reed 1 0
Reed 2 1
Reed 3 3
TABLE 27
CONTRIBUTING PARTIALS OF f#1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 6
Reed 2 5
Reed 3 3
Mark 10 or more Reed 1 3
Reed 2 3
Reed 3 2
Mark 15 or more Reed 1 2
Reed 2 2
Reed 3 2
TABLE 28
CONTRIBUTING PARTIALS OF a1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 5
Reed 2 5
Reed 3 4
Mark 10 or more Reed 1 3
Reed 2 2
Reed 3 3
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 2
TABLE 29
CONTRIBUTING PARTIALS OF d2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 4
Reed 2 2
Reed 3 3
Mark 10 or more Reed 1 3
Reed 2 1
Reed 3 2
Mark 15 or more Reed 1 2
Reed 2 1
Reed 3 1
TABLE 30
CONTRIBUTING PARTIALS OF f#2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 2
Reed 2 2
Reed 3 2
Mark 10 or more Reed 1 2
Reed 2 2
Reed 3 2
Mark 15 or more Reed 1 2
Reed 2 1
Reed 3 2
TABLE 31
CONTRIBUTING PARTIALS OF a2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 2
Reed 3 2
Mark 10 or more Reed 1 2
Reed 2 2
Reed 3 1
Mark 15 or more Reed 1 1
Reed 2 2
Reed 3 0
TABLE 32
CONTRIBUTING PARTIALS OF d3 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 2
Reed 2 2
Reed 3 1
Mark 10 or more Reed 1 2
Reed 2 2
Reed 3 0
Mark 15 or more Reed 1 0
Reed 2 1
Reed 3 0
TABLE 33
CONTRIBUTING PARTIALS OF d1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 4
Reed 2 6
Reed 3 4
Mark 10 or more Reed 1 2
Reed 2 3
Reed 3 2
Mark 15 or more Reed 1 0
Reed 2 1
Reed 3 1
TABLE 34
CONTRIBUTING PARTIALS OF f#1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 4
Reed 2 5
Reed 3 4
Mark 10 or more Reed 1 1
Reed 2 2
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 2
Reed 3 2
TABLE 35
CONTRIBUTING PARTIALS OF a1 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 4
Reed 2 5
Reed 3 4
Mark 10 or more Reed 1 3
Reed 2 3
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 1
TABLE 36
CONTRIBUTING PARTIALS OF d2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 4
Reed 2 3
Reed 3 3
Mark 10 or more Reed 1 1
Reed 2 2
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 1
TABLE 37
CONTRIBUTING PARTIALS OF f#2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 2
Reed 3 2
Mark 10 or more Reed 1 2
Reed 2 2
Reed 3 2
Mark 15 or more Reed 1 1
Reed 2 1
Reed 3 2
TABLE 38
CONTRIBUTING PARTIALS OF a2 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 2
Reed 2 4
Reed 3 2
Mark 10 or more Reed 1 2
Reed 2 1
Reed 3 1
Mark 15 or more Reed 1 2
Reed 2 1
Reed 3 1
TABLE 39
CONTRIBUTING PARTIALS OF d3 AT MARKS 5, 10, AND 15 ON THE SPECTRUM GRAPHS
Total*
mark 5 or more Reed 1 3
Reed 2 2
Reed 3 3
Mark 10 or more Reed 1 3
Reed 2 0
Reed 3 3
Mark 15 or more Reed 1 1
Reed 2 0
Reed 3 1
Again in this test all the reeds showed a great difference in their spectra. The tables 19 through 29 show how the reeds Tables 19-39 go here of each expert ranked in number of partials at each mark on the graphs. In all cases reed three of expert number one showed the most partials at each mark. At mark 5 reed three was best 57.1% of the notes tested, at mark 10 reed three was best 71.4% of the notes tested and at mark 15 it was best 85.7% of the notes tested. Expert number two's reeds were less consistent. Reed one was best 85.7% of the notes tested at mark 5 and 10. However, at mark 15 reed three showed the most partials contributing to the sound 57.1% of the notes tested. Expert number three's reeds were also inconsistent. Reed two was best at mark 5 on 57% of the notes tested. At mark 10 and 15 reed three was best 85.7% of the notes tested. At mark 15 on the graphs reed three of each expert showed the most strong partials in the sound.
In the first test the French cut reed showed very well on the graph but in my opinion sounded bad. In an attempt to be more objective in the second test, the tape recorder and playback speaker were employed. The entire playing process of the second test was recorded. After their performances the experts listened to a playback of all the notes. The experts then judged which reed of each player sounded the best. In all cases the voting was unanimous. Reed one of expert one was voted the best of her reeds. Reed three of expert two was voted the best of his three reeds and reed three of expert three was voted best. Experts two and three had each considered their third reed their best, however, expert one's reed that was voted best was considered by her to be pitched high. In two of the three cases the experts agreed that the reed considered to be pitched well sounded the best. Both reed one and reed three of expert one were pitched high. It was noted that expert one had to open her embouchure considerably to get those reeds down to the proper intonation level.
The reed of expert one voted best by the group (reed one) was best on the graphs at mark five 42.9% of the notes tested. At mark ten reed one was best 28.6% of the notes tested. At mark 15 reed one was best 42.9% of the notes tested. It is interesting to note that the reed expert one considered well pitched at the beginning of test II (reed three), showed the best results on the graphs but was not voted best by the group. In no case was the reed voted best by the group the best on the graphs at mark five. At mark 10 the reed voted best of expert three was best on the graphs. Finally, at mark ten the reeds voted best of experts two and three were also best on the graphs.
The stated purpose of this study was to determine if the intonation of the crow of the reed has an effect on the tone quality of the instrument. The project was in two parts. The first experiment consisted of playing two octaves of the D major scale into an apparatus which recorded the sound spectrum of each note played. The apparatus consisted of a microphone, sound analyzer, oscilloscope, and an event recorder. Intonation was controlled by the use of a stroboscope.
Four reeds of differing intonation characteristics were used in the first test. Reed one crowed octave C naturals, reed two crowed only on C natural. Reed three crowed octave B naturals and reed four was a French cut commercial reed with a wild crow fluctuating between B flat, B natural and other pitches.
The first test was performed on one instrument, removing the possibility that differences in quality could be caused by different instruments. The tones were played into the apparatus and held for a short period after which they were frozen and recorded. Usually the third, fourth, fifth or even sixth partial was the strongest of the tone. At mark ten on the sound spectrum graphs reed one was the best on 60% of the notes tested. At mark fifteen on the graphs reed one was best 80% of the notes tested and at mark twenty it was best 66.7% of the notes tested. The results showed that tone quality is affected by the intonation of the reed. The sound spectrum of reed one was in most cases superior in number and strength of the contributing partials to the tone quality. The surprise of test I was the French cut reed. At the various marks on the graphs it was either second or third best of all the reeds. Yet the sound of the reed was very bright and definitely not an American concept of sound.
In the second experiment three professional oboists were asked to play three reeds each through the apparatus. Each was asked to make and bring a reed that in his or her opinion was pitched too high, a reed too low, and a reed that was well in tune. A D major arpeggio was played through the apparatus and recorded. In addition to being recorded on graph paper each reed was recorded on magnetic tape. The tape is on file in the Ohio State University Library. After all reeds were played the tones were played back and the experts were asked to judge which of the three reeds for each player sounded the best. Again in this test all the reeds showed a great difference in their spectra. In all cases reed three of expert one showed the most partials at each mark; 57.1% at mark five, 71.4% at mark 10 and 85.7% at mark fifteen for the notes tested.
The reeds of expert two were less consistent. Reed one was best at 85.7% of the notes tested at mark five and ten. At mark fifteen reed three sowed the most partials contributing to the sound, 57.1% of the notes tested. The reeds of expert three were also inconsistent. Reed two was best at mark five 57.1% of the notes tested. At mark ten and fifteen reed three was best 85.7% of the notes tested. At mark fifteen on the graphs reed three of each expert showed other partials in the sound.
In all cases the results of the spectrum graphs showed that the reed considered to be in tune by each player had the highest number of strong partials contributing to the tone quality. When the players judged the playback of the tones tested they chose the reed in tune as best for two of three experts. The high pitched reed of expert one was judged best. The reed of expert one voted best by the group (reed one) was best on the graphs at mark five 42.9% of the notes tested. At mark ten reed one was best 28.6% and at mark 15 best 42.9% of the notes tested. In all judging the vote was unanimous for the selection made.
Although these results cannot be considered totally conclusive, the weight of the evidence indicates that the pitch of the crow of the reed does affect the tone quality, and that a reed that crows a C natural or a close as possible to a C natural will deliver the best tone quality. There are many variations on the American concept of oboe tone quality, however, the C natural pitch of the crow of the reed is one constant.
Due to limitations and scope of this project several questions have been unanswered and need further research. The first of these is the French cut reed. Although the spectrum graphs were similar in many respects to the American reeds, the sound was very different. Second, two of expert one's reeds were sharp to the stroboscope. This necessitated a change in embouchure to bring the notes in tune. What effect does varying amounts of embouchure pressure have on the tone quality or even on the pitch of the crow? Third, what gouges of cane and shapes produce the best results? Fourth, noting that the proximity of the microphone to the oboe alters the sound spectrum graphs, studies should be done in large concert halls and more intimate recital halls to determine what differences and similarities are found and what alterations, if any, are needed in reed making, to produce the best results. These are just a few of the questions about the oboe and the reed that need consideration and possible research.
FOOTNOTES: