I think of recorder making in terms of gross and fine work. The gross aspects of recorder making include: turning; boring and reaming the wooden cylinders that contain the vibrating air column; cutting the wind-way; making and fitting the block; carving the ramp and window; drilling the finger holes; and cutting the beak. The fine aspects can be divided into two related areas, one of producing the sound and the other of getting the right pitch for each note. These are known respectively as voicing and tuning. Voicing involves the removal of very small amounts of wood, usually of the order of a few micro-grams, from certain areas where it makes a significant difference to the quality of the instrument's tone. One of my goals was to get a better understanding of where and how best to make these refinements. The recorder nomenclature used in this report is shown in Fig. a.
For turning and boring he uses a West-German wood lathe with 1000 mm between centers, about 300 mm center height and a lever operated tailstock spindle for boring long holes. The headstock spindle hole is 50 mm in diameter, allowing him to put a billet half way through the headstock for better centering when boring.
Philippe's method of tackling the challenge of boring long, straight holes is to use a combination of a Forstner drill to start the hole and a spoon drill of matching diameter to extend the hole through the billet. Spoon drills, which are probably not commercially available these days, are shaped like a spoon with long, straight sides on a shank instead of a handle. First he turns the end of the billet square using a parting tool, then he uses a spindle gouge to hollow out the centers so that only the circumference of the Forstner bit is acting in the initial cut. The Forstner bit is used to produce a hole with a depth of about four times the diameter.
The matching spoon drill is then gently guided into this hole, pushing it slightly away from the cutting edge so the initial cut is guided by the back side of the drill, not the point. Otherwise the first cut of the drill may push it off center.
One of the highlights of my trip was to see the Continental style parting chisels he uses. English style parting tools (used mainly to cut a narrow, deep groove perpendicular to a turning cylinder) are pointed with two bevelled sides as in Fig. b, whereas Continental style parting tools are pointed with one bevelled side, as in Fig. c, making them much sharper for less sharpening work. In ten years of turning wood and reading about it I have never seen or heard of the Continental style tool, yet it is such a simple and effective improvement.
The steel Philippe uses to make reamers is called 'Stub' steel but this appears to be a brand name that is not available in Australia and I don't know the composition of this type of steel. Ordinary bright mild steel seems to work well. His method of making a reamer is to turn a tapered cone corresponding to the desired bore shape, accurate to within about 0.1 mm for an alto, and leaving a suitable length of shank. He then drills a hole axially through the tapered section, then hack-saws of slightly less than half the metal from one side of the tapered cone. He then mills a cutting edge along the appropriate side of the remainder of the tapered cone and grinds a little relief following the cutting edge. This produces a reamer as in Fig. d showing the cross-section of the thin end of the taper.
The test of a reamer is whether it produces shavings or compacted wood fibres, the former leaving the most acoustically effective bore surface.
Joint sockets are bored as required while the head and foot joints are still in the lathe. Philippe reams his billets in a vice with three jaws that holds the work in a triangular grip, driving the reamer by a steel rod through a diametrical hole through the shank.
After reaming, the joint is put back in the lathe, held in the self centering chuck as it was for boring, and the bore surface is polished. This is done using a stick of dowel, about 12 mm diameter, with a kerf about 25 mm long cut in the end of it. A short strip of very fine abrasive paper, 400 grit, is passed through the kerf and rolled around the end of the rod, then the billet is rotated at high speed and the rod is passed along the bore surface to get the smoothest finish possible. This method ensures uniform removal of material from the bore.
At this stage the tenons are turned, ensuring the tenon fits loosely into its socket until the waxed cotton or silk thread is applied. If the tenon exerts pressure on the socket before the thread is applied, it may split the socket. When turning the head joint, the area that will eventually become the beak is left thick until after the block is fitted to reduce the risk of splitting the head in the process of fitting the block. The diameter of the head at this stage needs to correspond exactly with the wooden side packing used to hold the head in Philippe's wind-way cutting jig.
Before finishing the outside turnery, any fine defects are filled with superglue and powdered sawdust. The surface is then sanded, while rotating at high speed, with abrasive paper starting with 240 grade then progressively finer grades, up to 400 grade. Then the surface is polished with a cloth with a few drops of linseed oil and some pumice powder sprinkled on it. Next it is polished with tripoli powder in the same manner. Then it is further polished by a handful of wood shavings, clasped around the outside of the joint and moved along it, still rotating at high speed. Next it is taken out of the lathe and buffed on a cloth wheel loaded with a very fine jeweller's abrasive powder. At this stage the joints are soaked in a bucket of linseed oil for a few days.
The device used by Philippe and Fred Morgan, cuts the wind-way with side taper by manipulation of the jig and lateral curvature from the shape of the cutter edge, but with no longitudinal curvature. This is later done by hand, using a template and files and scrapers. Philippe's argument in favour of his device and method is that it provides the opportunity to observe the effects of the changes he is making to the longitudinal curvature of the wind-way. All makers agree that this longitudinal curvature is an important factor in producing a high quality instrument. In Philippe's device, the jig that manipulates the recorder head is of his own making. He says that in the end you use the tool that you're happy with, and that works for you.
The device used by Alec has a long pivot arm mounted on a sliding block. The short end of the pivot arm holds the cutting tool and the longer end rests in templates that control the side taper and longitudinal curvature of the cutting strokes. As in Philippe's machine, the lateral curvature comes from the shape of the cutting edge. The ratio of the length of the template arm to the length of the cutting arm provides a four to one reduction of any dimensional error in the templates and wind-way height adjustment. This level of accuracy is an attractive feature because dimensional changes in this area of less than 0.1 mm will have significant effects on the performance of the instrument. Both types of cutters cut on the pull stroke.
When turning the head joint, the grain direction is tested. Philippe cuts the wind-way roof against the grain for two reasons. Firstly, it means that when finishing the wind-way roof by hand, he is working with the grain, and secondly, when cutting the ramp he will be working with the grain. Alec cuts the wind-way roof with the grain to produce the smoothest finish off the cutting tool. Technically speaking, the wind-way is being scraped because the cutter is perpendicular to the wood, but in common parlance it is called cutting.
Both makers place the wind-way roof with the annual growth rings running vertically because wood tends to split perpendicularly to its growth rings. Cutting the wind-way weakens one side of the piece of wood. If the growth rings were horizontal, the natural place for the wood to split would be the weakened wind-way side, which would render the instrument useless.
Before clamping the head into the wind-way cutter, a slot about 3 × 8 mm that is the start of the window, is milled in the drill press, which is fitted with a compound table. (Philippe has a separate mini-mill for milling reamers.) A 3 mm hole is drilled at the north end of the head joint, exactly opposite where the wind-way will be cut, in wood that will eventually be removed to form the beak. Philippe's system of clamping the head joint in the jig necessitates the use of two identical wood blocks, previously mentioned, on either side of the head joint as packing. One of about a dozen cutting tools of varying widths and curvatures is then chosen. The lateral curvature for the baroque alto head we made was 29 mm radius, and for a renaissance instrument the cutter radius is half the radius of the bore. The cutter is then screwed into the slide block to a position where its sides are vertical, then clamped with a locking nut. The depth of travel of the slide is adjusted so the cutter protrudes about 5 mm south of the block-line, so it will be able to cut the "candle" shaped underside of the ramp, where the cutter radius converges with the bore radius.
The cutter slide is then adjusted laterally so that the corner tips of the cutter are both in contact with the bore at the block-line. The cutting edge is then slid to the north end of the head joint and tested for central alignment. If necessary, it can be adjusted by rotating the angle of the top-plate on the jig base. The central alignment of the cutter at the block-line then needs to be re-adjusted with the lateral slide.
When the cutter is centered along the length of the intended cut, the height of the cutter is gradually raised while levering the cutter back and forth. As soon as the two initial side cuts come together below where the labium is to be formed, the cutter depth stop, which controls cutter penetration into the head joint, is adjusted so the cutter will no longer go past the block-line. The distance between the tip of the 'candle' and the labium should be about 1 mm. When adjusting the depth stop the cutter location can be seen through the 3 × 8 mm slot that will become the window. Cutting is then continued until the wind-way is cut away to the desired height.
Philippe often takes the head joint out of its jig to measure the progress of the cut. This is done with a telescopic bore gauge, by comparing the horizontal bore dimension with the vertical bore dimension at any point along the wind-way. The locating pin and its corresponding hole in the head joint ensure it goes back into the jig in the same position.
The step (vertical distance between the underside of the ramp and the roof of the wind-way) that he aims for is 0.7 mm for a soprano, 0.8 mm for an alto, 0.9 mm for a tenor, and 1.0 mm for a bass. His intended wind-way heights at the blowing end are 1.2 mm for a soprano, 1.2 to 1.3 for an alto, and 1.4 to 1.5 for a tenor. His gauges for measuring wind-way dimensions when voicing are made from a set of harp strings with increments of about 0.03 mm. These enable him to choose the diameter that best suits the instrument. They are not used progressively from one string to the next.
After the wind-way is cut, some preliminary cutting of the ramp is done to provide better visibility of the south end of the wind-way, but this should not be nearly finished yet, so as to maintain adequate strength for fitting the block.
The grain on the block is chosen to be able to cut towards the window as this makes it easier to do fine work on the wind-way exit. The annual growth rings should run vertically to give maximum stability to the wind-way floor, by minimizing moisture absorption. Philippe uses a small engineer's square clamped on top of the block as a guide when planing the waste off either side of the wind-way floor. This is done with two shoulder planes, one dedicated for each side of the floor, so that the blades are always flush with the respective sides of the planes. The underside of the block can be rounded in a vice. All Philippe's vices have soft jaws of cork, covered by leather, bonded to the existing jaws. The block ends are clamped on the jaws and coarse abrasive paper is drawn across the block, while pulling both ends of the paper down. The width of the abrasive strip is slightly less than the length of the block.
Sometimes, depending on the quality of the block wood available, it can be difficult to read the shiny spots that need scraping off. If so they can be highlighted by brushing talcum powder on the block before it is pushed into the bore. A varnish made from shellac is applied to the ends of the block, when finished, to minimize moisture absorption. The name and address of a supplier of block cedar that has been treated for maximum dimensional stability is listed in CONTACTS.
After the block has been fitted, the preliminary lowering of the platform that will become the wind-way floor begins. This is done by scraping the platform evenly until some light from the inside of the bore can be seen through the wind-way. Finishing the wind-way floor is part of the voicing process.
Philippe says the wind-way floor (top of the block) needs to be lowered only far enough so that the view from the blowing end is almost a clear view through the bore, but the light is being refracted from inside the bore, in effect giving a small halo of even thickness. The wind-way floor has a slight longitudinal concavity, with the middle being about 0.1 mm lower than the ends.
The ramp is cleaned up at this stage and the sides of the window may be trimmed until they form a uniform surface with the sides of the wind-way. I asked Philippe about the final thickness of the labium and he said it was something he could not quantify. For him it was more an intuitive thing. Alec says this thickness should be about that of a sharp pencil line.
About this stage the window exit chamfers are cut with finely sharpened chisels, using a flat chisel for the block chamfer and the left and right handed bevelled knives for the wind-way roof chamfer. The chamfers are cut at an angle slightly more than 45 degrees, and a width of about 1 mm, though the optimum dimensions depend on the overall wind-way configuration. Philippe uses a small nasal surgical mirror, about 12 mm in diameter, to inspect the chamfer on the wind-way roof.
Philippe said the voicing process consists of ensuring that the curvatures of the wind-way roof and floor are regular and even, the ramp and wind-way roof are as smooth as possible, the chamfers are absolutely clean, and the vertical sides of the ramp have no ridges or irregularities. He says: "In finishing work you need to be looking at all these points at once. I do quite a bit of work with abrasive strips glued to wood. It's often trial and error... seeing something that doesn't look quite right and fixing it and seeing the change in the instrument... no ridges or wooden fibres, angles perfectly clean and things like that. I always find it very difficult to finish an instrument. It's easy to get it to where it's almost perfect, but to get it just right is very difficult. I like to voice an instrument a little bit at a time. Perhaps I get to know the instrument better between each time I work on it."
One of the basic principles of voicing is to only make one change at a time, so as to be able to see what effect it has. The chamfers can be cleaned with a strip of 1200 grade abrasive paper, about 4 mm wide and 100 mm long. To clean the top chamfer this is passed through the window and underneath the wind-way roof and out the blowing end, then gently drawn across the chamfer a few times. The bottom chamfer is cleaned in a similar manner with the block removed. When working on the block platform after blowing the instrument, it is best to remove the moisture with a hair dryer.
As part of the voicing work the finger holes are polished by a small die grinder using a felt wad loaded with beeswax. Under-cutting the finger holes is a good example of how voicing and tuning are related. The inverted funnel shape of the finger holes, produced by undercutting, greatly improves the volume of the instrument, while increasing the size of the holes in this manner will raise the pitch of the affected notes. (Philippe says he is obsessed about getting nice loud low notes on his instruments.)
Philippe introduced me to a number of very useful tuning aids. These were: a diagnostic chart of effects of bore changes; a computer program to calculate note pitch with any bore and finger hole configuration; a program to convert an instrument design from one pitch to another; and a blower with a pressure gauge to monitor air pressure when tuning. Adrian Brown (1989), in his excellent book "The Recorder - A Basic Workshop Manual", explains a very useful system of charting possible bore changes. One such chart on the instrument we re-tuned is shown in Fig. h. The bore diameter and finger hole positions are charted across the top of the page, then the effects of passing a piece of plasticine about 5 mm in diameter by 20 mm long on a thin wire through the bore are monitored. This effectively reduces the bore diameter at the location of the plasticine. It may either cause the note to go sharp or flat, or have no effect at all. It may also cause a burble, or a stronger tone. All of these effects are recorded throughout the length of the bore and charted. This chart can then be used for diagnosis of problems in a particular instrument, or to see where to alter the bore to improve tuning in a design. Philippe says: "Look for isolated points, that is, those that affect one note, to make changes. If a change affects a and c, then try to find another place where that change affects only a."
Two useful computer programs for tuning are one by Philippe, based on the work of the acoustician, V. Mahillon, and a French MSDOS program called Resonans (Bolton 1996). Philippe's program will convert bore dimensions and finger hole sizes for an instrument in one pitch to another, for example, a = 440 to a = 414 cycles per second, and is based on bore surface area. The Resonans program is only available in French, in MSDOS, but it allows the user to key in any bore dimensions in a number of conical segments, place different sized finger holes in any position along that bore, and it will indicate the resulting notes and their harmonic analysis. This program will be very useful to analyse a recorder plan before starting to make the instrument. Details of the supplier are listed in CONTACTS. Philippe sometimes uses a mechanical pressure gauge in conjunction with the electronic pitch meter that he often uses to verify what his ear tells him. The pressure gauge reads 0 to 100 mm of water and the air supply to the gauge comes from either the mouth of the maker blowing the instrument or a small organ blower which is almost silent. When using the blower, a valve made from a water tap is used to control the air pressure, and the gauge is fitted between the valve and the instrument. The recorder is connected via a rubberized membrane over a pipe end, with a hole slightly smaller than the diameter of the beak cut in it. Plasticine is used to cover the finger holes. Because the blower air is colder than a person's breath the instrument should be tuned so that on the blower it is about 15 cents below the correct pitch for the player's breath.
The steps previously outlined produce high quality recorders when applied with great care and experience. Philippe also pointed out that to succeed as a recorder maker, the aspiring artisan also needs a range of instruments to offer potential customers. His range includes about twenty different recorder models, from Medieval to Baroque in various pitches, mainly A = 415 or A = 440.
Alec Loreto also points out that when he and his contemporaries started making recorders, the market was more tolerant of variations in quality. These days, he says, new makers need to produce instruments comparable with highest standards, which have risen greatly in the past twenty years, before they can expect to sell them.
For me this amounts to a big challenge, and I'm glad I'm not dependent on selling my recorders for my income. It has so far proven to be a fascinating and rewarding hobby and my visit to Philippe has shown me many ways in which I can improve the quality of my instruments.
Recorder making is a specialised and uncommon skill. Many makers are precious with their knowledge, but Philippe and Alec have both been very generous with their knowledge and one of the most important things I have learnt on this trip has been that I also need to be as generous wherever I find the opportunity.
At some time I will be discussing Continental and English style parting tools with other wood turners to find out why we in Australia use only double bevelled parting tools and whether any of them share my views on the advantages of the single bevelled parting tools.
| PHILIPPE BOLTON [Recorder Maker] |
ALEC V LORETTO [Recorder maker] |
| DICK GmbH [Supplier of left and right handed chisels] |
ARTHUR GOHIN [Supplier of treated block cedar] |
| JOEL GILBERT [Supplier of RESONANS program] |
ORGELBAU AUG. LAUKHUFF GmbH & Co. [Supplier of organ blowers] |
© Copyright 2002, Winston Churchill Memorial Trust of Australia
Webmaster: Nicholas S.
Lander