(No.4 of 6 articles published in Choir & Organ in 1998-9 under the heading 'Spit and Polish')

4. Scaly Monsters

One of the implications of my last article - mainly concerned with the role of mixtures - is that certain pipe scales may have different effects from others.

The scaling of organ pipes is at heart the business of determining the diameter of each pipe relative to its length, but in practice encompasses all those details of pipe construction that can be set at the manufacturing stage and before the voicer begins work: material, thickness, shape, mouth style and other factors for flue pipes; resonator and shallot details for reed pipes. The possibilities are endless and bewildering; mastery of the subject is exceptionally rare even amongst organ builders, and those who have become well versed in one way of doing things are usually happily oblivious of other methods.

The importance of pipe scaling - as a part of an artistic recipe in successful organ building - is a matter for considerable debate, because the final results, regardless of scale, are in the hands of the voicer. A pipe of given scale may be voiced loud or soft, keen or dull, quick or slow - and well or badly! Further modifications come in on-site finishing, where individual pipes pre-voiced to a reasonable expectation are finely balanced to their neighbours and to their situation in the building. After all this vital work, one has to wonder whether pipe scales are, in fact, all that important.

To say anything about this subject is to walk into a lion's den of argument. Many organ builders will tell those who take a special interest in pipe scales that their subject is unimportant compared to the art of voicing, but they will at the same time dismiss with considerable feeling any method other than their own!

The most truthful assessment is perhaps to say that pipe scaling is not un-important. It has a role to play, alongside the many other factors that make organ building such a complex and fascinating craft.

To the organ historian the measurement of pipe scales can be an important tool, revealing, for example, how techniques have changed over time or from one contract to another, giving clues about workshop method, and often showing the influence of master on pupil. Pipe-scales give an important insight into what an organ-builder was trying to do: among the English builders of the nineteenth century the large diapason basses of Walker, the bold choruses of Hill, and the exceptionally small-scaled upperwork of Willis are all highly indicative.

Perhaps scaling is better judged as a measure of intent than of actual results; this qualification allows us better to assess some of the techniques in use today, and to attempt - as in previous articles in this series - to sort the sheep from the goats - and perhaps to spot the occasional scaly monster.

In 1833 the theorist J. G. Toepfer proposed the first scientific 'solution' to the problem of organ pipe scaling, relating the cross-sectional area of the pipe to its length and determining a standard progression where the diameter halved on the seventeenth pipe of the chromatic scale - and every seventeenth pipe thereafter. (1) Such mathematical progressions had been in use before Toepferºs day alongside other methods, but the publication of his method made papre calculation relatively easy for others to follow. The old masters used the magic of geometry to lay out the scales directly on the flat sheets of metal and happily enjoyed an atmosphere of secrecy similar to that surrounding the calculations of stonemasons; now the numbers were available to all.

The simplicity of this method had considerable impact in the age of factory organ building and at a time when anything that appeared 'scientific' had a particular appeal. German organ builders were almost unanimous in their acceptance. Toepfer's progressions came to England with Schulze and were adopted by Lewis and one or two others - but the grand London firms of Hill, Willis and Walker kept faithfully to their older methods. In France Cavaille-Coll was in full support of the mathematical plan, but with characteristic facility and elan expanded the idea to encompass many different progressions suited to varied circumstances. In North America the early establishment of an independent pipe-making industry by builders of German descent seems to have led to the widespread adoption of a single tabulated form of the Toepfer progression - and by around 1920 everyone knew what a ÂNumber 44º was (it was the exact equivalent of German Normalmensur, a scale starting at 155.5mm diameter at 8º C and halving on the seventeenth note).

This last example brings us to the first of our scaly monsters, one in which the tail is wagging the dog. In the twentieth century the big American organ factories sometimes relied too heavily on their pipe suppliers and their standardised scales: the variety of methods used in earlier times was one of the things that so elegantly distinguished the house style of one firm from another. Still more alarming is the thought that this unimaginative system is still in use, for though G. Donald Harrison demonstrated the value of subtle variation from the standard scale sheet and used different progressions, many of his less lively contemporaries and successors stuck rigidly to the industry norm. Speak it not in Gath, but some builders still adhere to this method even now.

If 'Industry Norm' is the Charybdis of pipe-scaling monsters, then the Scylla is the method known as 'free-variable' scaling. At the Freiburg organ conference of 1926, where so many assumptions of the romantic period were disputed or challenged, and where the Organ Reform Movement can be said to have begun in earnest, a lively debate developed on the subject of pipe scaling. One camp believed that the work of the old masters was regulated by geometry and the magic of numbers; the other side held that the choice of scale was purely a matter for artistic intuition and free empirical choice (2).

To cut a long story mercilessly and mercifully short, the empiricists won, and 'free-variable' scaling became the standard method of European neo-classical builders. At its best, this method was as well capable of fine results as any other and the basic premise - that matters to do with pipe scaling should be led by the ear and by the artistic process, not by mathematics or artificial 'norms' - is eminently sensible. However, in the hands of the less skilful the method became an ugly beast with far too many tentacles.

From a inadequate handful of measurements of a few famous old organs, it became clear that ancient pipe scales did not conform to the Toepfer norm (an indication of surprise at this news is entirely optional). Indeed, plotting the old scales on a Toepfer graph showed that the old masters preferred not a straight line, but a complex series of interlacing curves. 'Aha!' saith the empiricist; 'we now know that the secret of the old masters lies in constructing a series of pretty curves on the graph paper'.

A most elegant version of this argument is to be found in Ralph Downesºs book 'Baroque Tricks' (3). Now, I do not deny that Downes was a most important influence on British organ building; indeed I would go so far as to say that he was probably the best organ builder in Britain in the 1950s (despite not being a member of the profession), and that his great works at the Royal Festival Hall (Harrison & Harrison 1954) and the Brompton Oratory (J. W. Walker 1953) are masterpieces of their kind. However, the argument about pipe scaling is not entirely valuable. Downes offers us the scales of Dom Bedos and those of the magnificent van Hagebeer and Schnitger organ at Alkmaar as examples of the desired elegant curves. No matter that Toepferºs progression would itself appear as a curve if plotted on one of Dom Bedos's graphs! No matter that the scales at Alkmaar are greatly confused by the fact that Franz Caspar Schnitger re-used so much older material made according to a different tradition!

The result - and it is an obvious criticism of the work of Downes and of some who unblinkingly followed a similar path - is that if the pipe scales studiously avoid any regular progression or any obvious relationship to each other, then the various ranks of the organ will not blend. Good voicing and finishing may well make the result tolerable, even excellent, but in so many less successful neo-classical organs free-variable scaling acts as a disruptive influence. Individual ranks may well be beautiful in their own right, despite changes in tone through the compass; add two together and the effect may still be charming, albeit more obviously uneven; try a combination of three, and the result is extremely unlikely to blend in an entirely satisfactory way. In larger combinations it can get chaotic.

Some elegant variety from one rank to another can, of course, be musical and desirable, and is perfectly common in old organs of all periods, but the old organ building masters who created the finest musical results had access neither to graphs, nor to pocket calculators, nor to the marvellously elaborate pipe-scaling slide rule invented by the German organ builder Richard Rensch and marketed by the organ supply house of Laukhuff. They were obliged to use simple and practical methods in the workshop, and in setting out pipe scales worked from a handful of templates directly on to the sheet of pipe metal. The result was simplicity and repetition: in the pre-industrial period it was impractical to use more than a very few recipes, and once a good one had been found it would be used again and again.

Where does this leave us in our attempts to avoid many-headed Scylla on the one hand and the whirlpool of Charybdis on the other? The answer is perhaps more simple than it may seem: to secure a good musical result and maximum blend, the example of the old masters actually indicates that we should use simple scaling, for forced variety is ultimately disruptive - and this underlines the message of the last article, that choruswork must be of similar scale and power to the foundation if it is to blend well. However, to avoid banality across the industry and to preserve the artistic integrity of the instrument, each builder might well use a different system according to taste. All may agree that, as far as pipe scaling is concerned, no one set of rules will lead to success, even if there are one or two methods that lead one to expect the worst!

(1) J. G. Toepfer: Die Orgelbaukunst - Weimar 1833
(2) C. Mahrenholz: Die Berechnung der Orgelpfeifenmensuren - Kassel 1968
(3) R. Downes: Baroque Tricks - Oxford 1983

next article in this series

back to: Other Publications

back to: Essays