Editor's Note: Bruce Haynes is at the moment one of the few professors of baroque oboe in the world--he teaches at the Royal Conservatory in the Hague, Holland.
Haynes has played oboe professionally since 1960, having studied originally with Raymond Duste in San Francisco, and briefly with John de Lancie. In 1964 he was awarded a Fellowship from the University of California to study with the Dutch recorder virtuoso Frans Brueggen. After graduating cum laude in 1967, Haynes began a two-year apprenticeship with Friedrich von Huene in Boston. He later set up his own workshop in California as a builder of baroque oboes
In 1972 Haynes was invited to replace his former teacher at the Royal Conservatory during a sabbatical leave. He has remained since then in Holland. As a player of baroque oboe and recorder, he is heard regularly in Europe and America with many ensembles and orchestras, and is a featured soloist on the Telefunken Bach cantata series on original instruments, as well as several recent Seon/ ABC recordings of Italian oboe concertos and other works by Couperin, Bach, and Hotteterre. In addition to playing and teaching, Haynes is involved in comprehensive research on the repertoire and playing technique of his instrument, and has written a number of articles on the early oboe.
As every double-reed player knows, writing about reeds is difficult. This article is an adaptation of one I wrote several years ago. At that time, I didn't expect much of it to be useful for very long because of constant new discoveries in this field. I am surprised, however, how much of it seems to have endured the test of even this short time. I offer it now as final and definitive, but in the hopes that it will be of some use to fellow IDRS members. Personally, I am continuing to experiment, and do not regard the information given here as anything more than a starting point for further discoveries.
This article limits itself to aspects of reed-making which are different than those generally used for the modern oboe. As readers of this Journal know, there are several good books on making modern reeds. [1] For those who want to make baroque reeds with no previous experience in reed-making, I would recommend combining information from this article with lessons from a modern oboist.
To define my terms, I call the 'baroque oboe' the instrument which was invented about 1660 and gradually mutated up till about 1760; it generally had three keys and elaborate turning, and was pitched at about a'= 415 (1/2 tone lower than modern orchestral pitch.) The dimensions I will be giving do not necessarily apply to later types of oboe. [2]
Unfortunately, there seem to be no known surviving baroque oboe reeds. Reeds from the late 18th century do exist, but although they are invaluable in giving us clues for reconstructing earlier reeds, they were made for an oboe with a bore and tone-holes of significantly different proportions than those of the standard baroque oboe type, and for music of a different artistic and technical nature.
For a number of reasons, frustratingly little was written by contemporaries about early reeds. In no case is there any information on scraping or finishing, and what little is said about dimensions is not to be trusted unquestioningly. Pictures are our best early source, although oboe reeds hardly form important parts of compositions, and are consequently rarely depicted in detail. [3]
There is an ever-present danger that we modern oboists coming to the baroque oboe, like pianists approaching a harpsichord, will bring with us unconscious expectations and assumptions which are not relevant. For this reason we must constantly reread and refer to the few original sources that have come down to us. On the other hand, our best key to reconstructing early reeds, though sometimes misleading, remains the practical and empirical experiments each player makes with his reed in one hand and his knife in the other.
The differences in reeds are naturally a reflection of differences in instruments. The baroque oboe is considerably larger in overall dimensions than the modern oboe; the bore is much wider usually 1.5 to 2 mm.) and the pitch is lower usually one half tone or more.) To accommodate this bore, the staple and reed must be proportionally wider. [4] A wide reed tends to play low notes more easily, helps the response of the cross fingered notes like the b-flat, g# and f#, causes less squeaking, and has a generally rounder and sweeter tone. Beyond a width of 10 mm., however, the high notes stop responding.
The size and shape of the tone holes is also quite different. On the baroque oboe, these are much smaller and proportionally longer. This causes a greater resistance in the general response of the instrument, so that reeds must be scraped softer and freer. The resistance that the modern oboist often consciously creates in his reed, for the sake of tone quality, is automatically built into the early oboe. Softer reeds allow more of the dynamic nuance and tonal range demanded by baroque music.
The cane used on modern woodwinds comes from the plant genus Arundo, to which references are made in sources dating from at least the late 17th century. [5] It is difficult to imagine using any other reed material on an instrument as sensitive in its reed as the oboe has always been.
If one makes a wider reed than the modern one, the diameter of the cane when in tube form must also be larger, to help keep the reed reasonably closed (i.e., without too much 'arch'). Modern oboe cane is generally 11-12 mm. in diam. for a tip width of about 7.5 mm.; modern English horn cane is 12-13 mm. It follows that cane for a reed 10 mm. wide needs to be about 15 mm. in diam. Cane of about 13 mm. diam. can be used (large modern English horn cane, in other words), but it may create a reed which is too open.
Good quality cane of 14-15 mm. diam. is available in tubes directly from some growers in southern France. A list of these can be found at the end of this article. As cane often comes incompletely aged from suppliers, it can be cured by being left in the sun (on a windowsill, for instance) for several months, being turned occasionally. This gives it a rich yellow-brown colour.
Buying ready-gouged unshaped English horn cane eliminates this step, but may limit one's choice of gouge thickness or cause the reed to be too open.
Using a larger diameter of cane means that a gouging machine has to be adapted. Both the bed and the blade need to be larger. This gives you the opportunity to experiment, of course, with gouging the sides thinner than the middle if you wish. (With my shape and staple dimensions, I find that this gives a better tip opening and saves scraping on the sides of the reed. See fig. 4)
Reeds were probably gouged by hand in the baroque period, as many oboists in Holland and other countries continued to do until recently.
This involves using either a tool similar to those shown in Diderot's and Garnier's plates, [6] or a hand-carving gouge (a kind of roundbladed chisel) with the ground (sharpened) side on the bottom, and a diam. approximately equal to that of the cane used. [7] (Gouges for hand-carving work better than the similar but more heavily built ones used with a hammer.) (Fig. 2)
The cane is first sawn to length and split lengthwise in four with a knife, and then soaked in water for an hour or two. Sawing slightly longer than the bed length, allows one to hold the cane while gouging, by pushing on the near end with a finger.
During the gouging process, the cane is supported by a wooden bed of somewhat larger diam. than the cane, about 110 mm. long, with a blind end toward which one gouges. (Fig. 3) In order to keep both hands free, the bed can be clamped to a table. In gouging, try to make a continuous cut from one end of the cane to the other, in order to keep a consistent thickness. The gouging tool should be kept quite sharp. Leave more wood in the middle than on the sides, so that the cane has a slight crescent shape when seen from the end. (Fig. 4)
After gouging, the cane is scraped smooth with a tool similar to Garnier's 'grattoir' [8] or scraper. (Fig. 5) Similar tools are still available commercially, the English horn end being most appropriate. Such a tool can also be made fairly simply. Other materials which can be used for finishing the inside surface of the cane are Dutch rush, and fine 'wet-or dry' sandpaper wrapped around a dowel or pencil.
Cane thickness is one of those variables which must be matched to a particular scrape, shape, etc. and ultimately depends on personal preference. Since the hardest cane is near the outer surface (or bark), the more material that can be removed from the inside surface of the cane by gouging, the harder, more resilient, and long-lasting the reed. This must be balanced by the fact that thicker cane generally gives a fuller and more desirable tone. A standard thickness does not guarantee consistent results, however. This is because cane varies in hardness, and the softer pieces need to be thicker to produce the same amount of resistance. One old tried-and-true method of measuring proper thickness is to hold the piece of cane between the thumb and forefinger of each hand and twist slightly: you quickly develop a sensitivity for the proper feel (i.e. thickness) of the cane.
Just before shaping, a final dry scraping smooths and polishes the inside surface, assuming that the scraper is sharp. With practice, this method of gouging takes scarcely more time than using a machine.
Baroque staples, aside from their dimensional differences, varied from modern ones in two basic ways: they did not fit into a 'well' or tenon with shoulders or corners at the top of the bore, but rather into a reverse cone or tapered counterbore. (Fig. 6) They were also made from a piece of flat metal stock which was wrapped around a mandrel; the seam was generally left unsoldered. (Fig. 7) The staples were then wrapped with thread to make them airtight. Early staples were usually made of brass, of an average thickness of 0.4 mm.
One can choose different solutions to the problem of staples. A handy and practical one is a multipiece staple. This consists of an upper section made from a cut-off modern oboe d'amore or Viennese oboe staple (on which the reed is tied) combined with one or more conical brass tubes which fit into the oboe. This telescoping staple is especially useful for experimenting with different types of staple, or for trying out a new instrument when one is unsure of the type of staple it needs. The reed, of course, effectively remains the same for various 'different' staples. (Fig. 9)
Another solution is to use no staple at all. Diderot in the Encyclopedie, [9] next to his plate depicting an oboe, shows a bassoon like reed apparently made without a staple. Similar reeds have been found on early musettes, instruments closely related to the baroque oboe. Perhaps both stapled and unstapled reeds were used in Diderot's time, although Garsault [10] clearly shows the stapled type. Interesting experiments are at present being made with Diderot type reeds, and for some oboes, at least, they may be useful. [11]
Staple making from brass stock is time consuming, especially if one is experimenting with designs and cannot re-use old staples, but the possibilities which it opens for control of intonation, response, and tone will repay the effort. Because the modern staple is more or less standardized (as is the modern oboe compared to earlier ones), the art of staple design has nearly been lost to us. [12] Rediscovering it is perhaps the longest and most complicated aspect of baroque reed making, and certainly one where much experimentation is still needed.
I think it helps when thinking about staple dimensions to consider five variables as important: the three "active" ones are top diameter, relative conicity, and 'exposed length', or the distance the staple extends beyond the end of the oboe. The bottom diam. and total length of the staple are 'passive' variables, as they are determined after the other three are fixed.
The top diam. of the staple affects the response of the extremes of high and low registers: generally, for low notes the bigger the better; for high notes vice-versa. For an average baroque oboe, the range lies between about 2.4 and 3.2 mm. Since the top end of the staple is usually somewhat flattened, it should be kept in mind that the flatter the final opening, the wider the original round diam. can be, as the area of the opening becomes smaller as it flattens (to the extreme of completely flat and no area), and it is this area which is significant to us. Also the design of the top of the staple will have a direct influence on the shape of the reed (at the throat.)
The second variable, conicity, affects the relative intonation of high and low registers. The more conical the staple, the wider the intervals.[13] If the high notes, for instance, are sharp in relation to the low notes, a less conical staple will lower them. This works, of course, in both directions. Likewise, if the high notes do not seem to be flat, but are thin in tone, it is possible that they are being pinched too much in order to bring them up to pitch; making the staple more conical will raise the high notes so that they can still be played in tune but with a looser embouchure, thus opening up their tone quality. Conicity can therefore have a secondary effect on tone. The practical range of conicity in baroque oboe staples seems to lie between about .4 and .7 mm. expansion per cm. of length. One is advised to begin by trying .5-.6.
The last active variable, exposed length, determines the basic pitch of the reed. To control this variable, one must of course maintain a constant basic shape, approximate cane thickness, and scrape. The further the staple is exposed, the lower the pitch. 30-40 mm. is average. Generally a mm. or two more or less in the length of the staple has little effect, although this same difference in the length of the cane itself is quite noticeable.
The only dimensions as yet undetermined are the total length of the staple and its bottom diam. The bottom diam. will be fixed by the final length, of course, (if the staple bore is a straight cone) and the total length is simply the sum of the exposed length described above, plus whatever amount is necessary inside the oboe to bring the sides of the staple into contact with, or at least close to, the sides of the oboe bore. In order to avoid leaking, the bottom of the staple should come as close as possible to the oboe bore walls. (Leaking is a major problem on the baroque oboe, and is the main cause of squeaking. If the staple is too short, it creates turbulence in the bore just below the reed, the direct result of which is squeaking on the middle d'', e-flat'', and e''.) When this end point has been found, it also shows the final length of the staple and the diameter of the staple bore at its bottom (N.B.: because the staple walls also have a certain thickness, the inside diam. of the staple is of course smaller than than outside diam., and this must be taken into account.)
The most accurate and exact way to find this end point, and to design the staple in general, is on graph paper. If one determines a centre-line through the bore axis, marks the exposed length and top diam.-on one side and the bore of the top of the oboe on the other, it is only necessary to draw a straight line (i.e. the conicity desired) between the top end of the staple (which is already marked) and some point on the oboe bore. (Fig. 12) As this line represents the inside bore of the staple, one must then allow for brass thickness (about .4 mm.). (Fig. 13) To use this method, it is necessary to know the dimensions of the top of the oboe bore. If this is not possible to measure accurately, (in the case of modern copies, measurements of this part of the bore should be obtainable from the maker) one must make a staple which is first too long, and gradually file or saw away the bottom until it is short enough to project the desired length from the top of the oboe.
Two warnings are in order here: the staple must project a certain distance into the oboe bore (about 20 mm.) to avoid a squeaking e-flat'', which is very sensitive where the staple and oboe are coupled. If one has designed a staple which does not project at least this far into the oboe, about 5 mm. should be removed from the bottom end of the staple, allowing it to extend further in. To compensate for the lost 'exposed length', make the cane part of the reed 2-3 mm. longer. In this way one will end up with a reed which squeaks less, but retains the same conicity and pitch. Secondly, extreme care should be taken (especially when working with original instruments) that the staple does not cause any damage to the oboe bore when it is inserted. Sharp corners or rough handling can easily create a groove or hole in the bore which affect the playing qualities of the instrument.
When the round dimensions of the projected staple are known, they must then be transformed into flat dimensions, because one begins with a flat piece of brass.[14]The length, of course, remains the same, but the diameters have to be converted to widths. This is done basically by multiplying them by pi, or 3.14. However, to compensate for the thickness of the brass, it is necessary to add .4 mm. to the diam. before multiplying by pi. For example, to calculate from a diam. of 6.0, add .4 and multiply by pi, or: 6.0+0.4 = 6.4x3.14 = 20.096 or 20.1 Thus the end of the staple that will have a diam. of 6.0 when round must be cut first, when flat, to a width of 20.1 mm.
In designing a staple, begin with about the mean of the practical ranges given above, and alter the design from this point as indicated. It is advisable to experiment with only one variable at a time!
A staple with unsoldered seam can be made in the following way:
1. Cut out a piece of brass with metal cutters (scissors) to the dimensions calculated. Cutting should be accurate to within .1 mm. tolerance. A file can be used after cutting to help touch up the dimensions.
2. Anneal (soften) the piece of brass by making it red-hot and then cooling it immediately in water. This can be done by holding the brass with pliers over a flame, such as that of a kitchen stove. When the brass is annealed but still flat, scribe horizontal lines in it from side to side with an old knife, which will later help to keep the thread and cane in place.
3. Make an initial bend down the middle (lengthwise) of the annealed piece of brass by pressing it into a groove with a mandrel. (the gouging bed works well for this--fig. 3) Then bend it roughly around the mandrel with your fingers. (The brass should be soft enough so this is quite easy.)
4. Avoid any rough hammering of the staple, so that the bore will be as smooth as possible when finished. Tap it along the seam so that it forms a butt-joint but does not overlap. (Fig. 14) Use a wooden or plastic hammer, on a wooden surface. Even this small amount of hammering will "work-harden" the brass sufficiently. The bottom of the staple must be perfectly round when finished, in order to fit exactly into the oboe bore.
5. File both ends of the staple straight (i.e. at right angles to the axis of the staple). (Fig. 15) File the bottom rim of the staple at about 45, so that there are no sharp corners to damage the oboe bore when the staple is inserted. (Fig.16)
6. Tightly wrap the staple with beeswaxed linen or cotton thread of about .15 mm. thickness, beginning near the top but leaving room to tie on the reed, and progressing nearly to the bottom. Be sure the seam closes completely over its entire length. Before wrapping, the staple should be inserted into the oboe to determine if it extends far enough into the bore, if it does not, check (1) if it is absolutely round, (2) if it closes completely at the seam, and (3) if the dimensions are still accurate.
7. Continue to wrap the staple with thread so that it forms an airtight seal when inserted into the oboe. This is critical. Regularly check, while wrapping, to be sure that the staple remains projecting the proper length out of the oboe.
8. Flatten the upper end of the staple with pliers as desired. In the interests of avoiding leaks, put the seam on one of the flat sides so that the cane will seal it when it is tied to the staple.
As mentioned before, for the sake of tone quality and easy response of cross-fingerings, it helps to use a reed of 9-10 mm. tip width. For the same reasons, the width of the bottom part of the cane, the 'throat', should also be as wide as possible. Although it is often stated that early reeds were 'fishtail' in shape, early evidence is not clear on this subject, and this was probably an oversimplification (note the shape of the reed in the painting detail reproduced here; compare also the difference in shape between Garsault's oboe reed (top) and the bassoon or crumhorn reed (bottom)).
As the throat of the reed depends for its dimensions on the size of the top of the staple, this can be a starting point in deciding on a shape. The sides of the reed should be wide enough to close together just as one ties up to the top of the staple. If they close too soon, there may be too much pressure on the sides of the reed, causing it to remain too open, but if they never quite close, the reed will leak along the sides, causing squeaks on the low g', g#' and a', and general unresponsiveness on the low notes.
One must experiment with shaping by hand until a usable shape is found. In practice, this will not take long. After a shape is decided upon, it helps for the sake of consistency to make a rudimentary shaper of wood or soft metal along the lines of that shown in Fig. 17. When shaping, the folded cane can be held against the shaper by its ends with the fingers. Using wood for a shaper has the advantage that the grain holds the wet cane in position better than metal, but wood wears out more quickly. Eventually one can have a special tip made for a standard commercial shaper with interchangeable tips.
As the shape often affects the opening of the reed the occasional or regular use of a wire in connection with a given shape should be considered. Wires do affect the freedom of vibration of the reed, but this may be controlled to one's advantage. They also add support. They were evidently used even in the early 18th century, as Hotteterre apparently refers to them. [15]
When using unsoldered seamed staples, it is important that the thread be carefully wrapped so that there are no leaks, as these cause squeaks and unresponsiveness. A lacquer or glue should be applied to the thread after tying, to help seal it.
It should be noted that surviving early reeds were often tied as shown in Fig. 18, so that the width of the throat is greater than necessary to fit the top of the staple. This may have two effects: it allows the shape of the reed to remain very wide at the throat (which primarily affects the tone), and it implies a loose thread pressure (as tightly binding such a shape will cause the reed to be too open). Experiments with this method of tying might therefore be worthwhile.
As there are at least as many opinions about how to scrape a reed as there are oboists, little need be said here. It may be helpful to read Garnier's instructions on scraping. [16] Baroque reeds are generally scraped more lightly than modern ones, to facilitate the response of the crossfingerings and to compensate for the resistance caused by long narrow tone holes.
As the opening of the reed is important, this can also be controlled by using the old bassoonist's trick of pinching the staple with pliers (Fig. 19) at points 'A' for closing, or points 'B' for opening the reed. (With unsoldered staples, it is wise to do this with the mandrel inserted, or the staple may collapse.)
Goldbeater's skin is sometimes necessary to assure an airtight seal on the sides of the reed. Squeaks on the low g', g#' and a' are generally caused by leaking along the sides.
It goes without saying that reeds are only a means to good playing, not an end in themselves. Therefore, shortcuts are in order when first learning to play the baroque oboe. For cane, use gouged, unshaped English horn cane, which will work quite reasonably and is easy to find. Better yet, for cane and ready-made staples which are good for many types of baroque oboe, write to Pat Grignet, Music Dept., Western Kentucky University, Bowling Green, Ky. 42101, USA, who supplies these materials. Readers who may know of other sources for baroque oboe reed supplies should let us know via the IDRS newsletter.
Oboe cane of 14-15 mm. diam. can be ordered by the kilo from: Francois Alliaud, 224 rue de la Republique, 84310, Morieresles-Avignon, France. (Highly recommended.) Dante Biasotto, Route de Boron 83, Frejus, France. Albert Glotin, 15 rue du Progres, 95460 Ezanville, France. (14 mm. only)
[1] For Instance: Bas, Louis. Methode Nouvelle de Hautbois, Enoch and Cie, Paris. Rothwell, E. Oboe Technique, Oxford University Press, London. Sprenkle, R. and D. Ledet. The Art of Oboe Playing Summy-Birchard. Stein, K. Rohrbau für Oboen, Bote und Bock, Berlin. Stotijn, J. DeKunst van het Maken van Hoborieten. Molenaar, and Hedrick, Peter and Elizabeth. Oboe Reed Making--A Modern Method, Swift-Dorr, 1972. [return]
[2] Such as Delusse, Milhouse, and Grenser oboes (the so-called Rococo and Classical oboes), whose bores and toneholes are smaller, and which play higher In pitch. [return]
[3] The most useful and detailed of written sources are Banister J. The Sprightly Companion, 1695. Brod, H. Methode pour le Hautbois, c. 1826. Garnier, J.-F. Methode Raisonee pour le Hautbois, c. 1800 (For a very interesting review of Garnier's reedmaking instructions and a reproduction of his plate of tools, see Thomas Warner's 'Two Late Eighteenth-Century Instructions for Making Double Reeds,' In GSJ XV.) Garsault. Notionaire ou Memorial Raisone, 1761, pp. 627-628. Hotteterre, J. Principles de la Flute . . et du Haut-bois, 1707. Talbot, J. Oxford Christ Church MS. 1187, cfl 1698. (See Anthony Baines's article on this MS in GSJI). [return]
Illustrations of early reeds include: Diderot D. and J. le Rond d Alembert, Receuil de Planches in the Encyclopedie, 1765, V, plates Vl, VIII, and X. Garsault, Op. cit. Reproduced here Anon. (German), Portrait of an Oboist, early 18th century. In the Staatl. Institut für Musikforschung, Berlin. Reproduced here. [return]
[4] The average width of modern English horn reeds is 8 mm. Many baroque oboists are now using a tip width of 9 to 10 mm. There are indications that such a width was common In the baroque period, both from the sources listed In footnote 3 and from surviving late 18th-century reeds. [return]
[5] See Borjon, Ch.-E. Traite de la Musette, 1672, p. 19. [return]
[6] Garnier, op. cit., p. 7, fig A and Diderot, op. cit., pl. X, fig. 15. [return]
[7] Cf. Garnier's advice on gouge size, pl. 6 ". . . Its curvature is proportionate to that of a piece of cane spilt lengthwise in two. [return]
[8] Garnier, op. cit. p. 7, fig. 8. [return]
[9] Diderot, op. cit., pl. VIII. [return]
[10] Garsault, op. cit. [return]
[11] Several players have recently been experimenting with 'bulgebore' staples i.e. staples with a belly rather than the straight conical bore described here. (Fig. 11) The Idea comes from baroque bassoon crooks which are apparently always 'bottle-shaped'. What works for the bassoon might work In miniature for the oboe staple. Experiments indicate that this system can help the response of the upper g" and the low f+ g and a. It has the advantage that it allows the bottom of the staple bore to be narrower while maintaining a wider conicity which could help the response of the high notes. The bellied staple can also be used In connection With a very large top diam. (about 3.2 mm.) if the opening Is quite flat. It is certainly more complicated to make. More experiments in this direction could be Productive although there are no known early staples which are made in quite this way. [return]
[12] Although I am told that Viennese oboists still actively experiment with staple dimensions for the modern Vienna oboe. [return]
[13] This Is apparently also true for modern reeds. See Sprenkle, op. cit., p. 91. [return]
[14] For an Illustration of a flat brass cutout before and after wrapping, see Garnier's tool plate, op. cit., p. 7. [return]
[15] Hotteterre, op. cit., p. 44. [return]
[16] Garnier, op. cit., p. 8. [return]