Deconstructing the Ideal

by Carlos Perez

August 20, 2001

One of the ever present problems we face in attempting to identify the highest quality manufactures, and especially in ranking a “first tier” of contemporary manufactures, is the lack of consistency within their product collections. A given manufacture may produce two or three different grades (so to speak) of commercial wristwatches, not even counting their works of high horology. It seems to me then, rather than attempting to classify all of the products of a given manufacture within the ostensible first tier of watchmaking as better or below that of another, we should personally try to develop the ability to identify whether individual pieces merit the denomination “high-grade,” and to determine where they shine or fall short of expectation against an objective standard — if such a standard is possible. After all, even the most affluent collector does not seek to acquire all of the pieces offered by a brand, but rather certain select pieces which he or she may find appealing or collectible for any number of reasons. I suggest then that it is not just “the best” which we seek, but the “best of the best.”

But why this obsession with “the best?” The victory of quartz a quarter century ago lies in its almost total superiority over mechanical watches in practical usage. Thus as an impractical pursuit of connoissuership in the quartz/atomic age, it is only the higher art of “fine” watchmaking which can justify the continued propagation of mechanical watches — at least for those in the school of craftsmanship. For others, anything mechanical, no matter how crude, is to be preferred to anything quartz, no matter how fine. Sadly for the lovers of “craft,” the bulk of today’s mechanical revival would scarcely rate even a classification of “good” relative to pre-quartz mechanical wristwatches, and much of what we consider “the best” only reaches the standard of average quality established historically. I should note that by “high-grade” and “fine” I do not refer only to the inaccessible world of haute horlogerie with its tourbillons and striking watches, but more to the distinguished heritage of best-quality simple watches, “gentlemens watches.”

Of course the form of the fine watch has changed over time with changes in technology and mode, from the single-cased cylinder and chronometer pocket watches, to the high-grade lever pocket watch, then to the fine lever wristwatch which is what primarily concerns us today. I am inclined to think that a high-grade wristwatch is one wherein the finest workmanship and the features of optimal timekeeping performance have been irrevocably intertwined, and it is this which is the basis of “craft” as a watchmaking and collecting philosophy. Herein I have attempted to deconstruct the platonic ideal of a high-grade wristwatch into some of its essential identifying factors — a basic guideline rather than an inflexible and unreachable standard.

Section 1: Distinguishing factors of the high-grade wristwatch

Free-sprung Adjustable-Mass Balance Wheel

A regulator is a device which allows for adjustment, or more properly, for the regulation of timing by manipulation of the balance spring, usually within a narrow range of +/- 20 seconds a day — so it is really a fine tuning device. Regulators take many different forms and are an indispensable element of most mass-produced mechanical watches. Yet they are often called “de-regulators” by experienced watchmakers due to the problems which they bring: they work by altering the effective length of the balance spring (or “hairspring”), thereby distorting the shape of the spring which ideally should be perfectly concentric. This will negatively affect overall accuracy when measured in various positions. Additionally, regulators are susceptible to shocks — a simple knock can drastically alter timing, and regulators will often “drift from their original setting. Finally, they increase the difficulty of positional adjustment. They have the great advantage of being quick and easy to use, but their many faults naturally leads us to a superior solution that discards with their unnecessary complexity.

The solution? Eliminate the regulator entirely and regulate timing by means of an adjustable mass-balance. The screwed bimetallic compensation balances which evolved to adjust for the temperature problems of steel balance springs were also the first adjustable-mass balances. The screws which were used to poise the balance and to adjust for temperature could also be used to regulate timing. In fact only two screws were needed to do so — simply by equal adjustment of two screws that were diametrically opposed. In time a special version of the compensation balance appeared with two dedicated timing screws in-line with the balance spokes. The advantages were clear: the free hairspring was not manipulated by a regulator and could “breathe” concentrically without distortion, resulting in superior positional performance, and the free-sprung balance was much less susceptible to timing changes due to shocks or drift. However, regulating an escapement via the timing screws on a balance wheel was a rather time consuming process, and so the technique was used primarily for chronometers and other high-grade watches. Thus the use or omission of a regulator came to be a defining division between the good and the best.

When improved spring metallurgy eliminated the need for temperature compensation from the balance, the split-rim bimetallic balance was replaced by the solid rim monometallic balance wheel of Glucydur. This also featured screws which were still used to poise the balance, and which could also be used for regulation. This kind of balance is still in use today, mostly for aesthetic reasons. In watches using regulators the screwed monometallic balance was eventually replaced by a smooth balance which could be poised perfectly by computer, and which would have a larger diameter in the available space. For free-sprung watches there evolved a number of purpose built adjustable-mass balances which were easier to use for regulation than the timing screws on a screwed balance, and due to their better distribution of mass also offered superior performance. Purpose built adjustable-mass balances take many forms, such as the Gyromax balances with eight or four “masselotte” rim weights originally created by Patek Philippe, and also used by Audemars Piguet (shown above), Vacheron & Constantin, and independents like Philippe Dufour and F.P. Journe; the Rolex Microstella; a new adjustable-mass balance from Omega; and others. In sum, the only type of balance which may not be free-sprung is the smooth type — the most common in use today.

Recommended reading from Walt Odets:

The Balance Wheel of a Watch

The Anchor Escapement

Adjustments for Temperature, Isochronism, and Positions

One of the more intangible and least understood areas of fine watchmaking, and yet one of the most important and expensive stages of movement preparation, is that of “adjustments.” Adjusting for temperature and positions is a problem brought to watchmaking by the invention of the hairspring around 1675, and so all techniques for compensating for temperature and gravity-induced errors naturally center primarily around the hairspring. And while the hairspring vastly improved “isochronism” (consistent rate of timekeeping over the entire slope of the mainspring’s torque curve), achieving the best isochronic performance requires tedious fine tuning. Most movements produced today are “unadjusted,” which once again places them on par with the “economy grade,” and even the disposable mechanical watches of the past. As a general rule movements are engraved with the number of adjustments, but as tariffs on imported watches is or was scaled to the number of adjustments, medium quality watches with only a few adjustments sometimes omit this information. “Adjusted to heat, cold, isochronism, and five (5) positions, also sometimes seen as, “eight (8) adjustments,” has long been the established standard for high-grade wristwatches — though more adjustments are sometimes used for special pieces.

The hairspring of the pocket watch age and the early wristwatch era was made of steel, which is highly sensitive to temperature fluctuations. Changes in temperature could result in wild swings in timekeeping performance, so as long as hairsprings were made of steel, a special temperature compensation balance wheel was used to adjust for performance in “…heat, and cold…” In the 1920s and ’30s new breakthroughs in metallurgy resulted in self-compensating hairsprings (from alloys like elinvar, conoruma, and nivarox) which were largely immune to temperature effects in the range which watches are used, and which are also amagnetic and corrosion resistant. This allowed manufacturers to dispense with the compensation balance, and with the time-consuming physical adjustments for temperature. I should note that the compensation balance is not obsolete, and can be used in combination with a self-compensating hairspring to eliminate the last traces of middle temperature error — taking the pursuit of perfection further than most will go. Today the production of self-compensating hairsprings is monopolized by Nivarox-Far, which makes hairsprings available in a few grades of quality. Fine watches will always be equipped with a Nivarox 1 hairspring, which is the best available.

The most tangle device that was traditionally used for best isochronism and positional performance was the overcoil hairspring. The hairspring, as mentioned above, was invented to improve isochronism, but the addition of an overcoil to the hairspring by Abraham Breguet further improved it, and also improved performance in positions, and made positional adjustment easier. Overcoil hairsprings were used almost universally on high quality watches throughout the 19th century and most of the 20th century. By and large the disappearance of the overcoil hairspring is tied to the fashion for extreme thinness which came into vogue in the late 1960s. By switching from an overcoil hairspring to a volute, or flat hairspring, a movement designer eliminates well over 1mm in thickness from the total height needed for the escapement, and the depth of the escapement is a primary determinant of the overall thickness of the movement. Almost all movements developed since that time have been extra-thin or ultra-thin in design, including the standard mass-produced calibres of our day. I should note that all movements of the last 150 to 250 years at least merit the classification of “thin,” denoting the change from full-plate movements to full-bridge and 3/4 plate movements.

If I may use Patek Philippe as an example of this industry-wide sea change in design philosophy: Its pinnacle simple automatic calibre of the early 1960s was 5.35mm thick, and its simple handwinds were about 4mm thick. Today, its central-rotor extra-thin automatic is 3.22mm thick, its ultra-thin micro-rotor automatic is 2.4mm thick, and its 10 ligne and 7 ligne hand-winds are both 2.5mm thick. Thus it is that overcoils have been largely phased out other than by Rolex, which still makes automatics in the 6mm range, in non-thin complications like high-grade traditional chronographs, tourbillons, and repeaters, and in pocket watches. This is one of the reasons why extreme thinness is a debatable part of “craft.” It is a great challenge of watchmaking to push the practical limit of thinness, but generally the thinner one goes the worse timekeeping gets, due to the loss of important performance features and ultra-narrow tolerances which defy human handling.

The actual adjustments made for isochronism and positions is more art than science, and are usually made by innumerable small tweaks to the hairspring and the rest of the escapement, and thus is something best described by a watchmaker — please see the recommended reading below. For reference I will list the positions used for adjustment here:

“…and five (5) positions.” = Dial-up, dial-down, crown-down, crown-up, and crown-left (or crown-right).

“…and six (6) positions.” = Dial-up, dial-down, crown-down, crown-up, crown-left, and crown-right.

“…and eight (8) positions.” = Dial-up, dial-down, crown-down, crown-up, crown-left, crown-right, crown-up 45, crown-down 45.

Recommended reading from Walt Odets:

The Concept of Timing

The Basic Principles of Watch Adjustment, Part 1

The Basic Principles of Watch Adjustment, Part 2

Timing results on five high-grade watches

Balance Wheel & Mainspring Size, and Beat Rate

Another factor regarding balance wheels which is intrinsic to high-grade movement design is balance wheel size. Indeed it is this and the size of the mainspring barrel which are traditionally the fundamental problems of calibre design, for it is foremost in the mind of a designer of fine movements to maximize the diameter of both. It was this technical need which forced the development of form movements for form wristwatches — not an aesthetic or conceptual purism. It is an even more pressing problem in automatic calibre design, where the winding train is also competing for space. As much as there is “craft” in manufacturing and finishing, there is craft in designing, and both of these elements are performance related. For example, the size of the mainspring barrel directly correlates to the movement’s potential isochronism.

A balance wheel should theoretically be as close to 1/2 the diameter of the movement as possible, and 40% appears to be the average (median, from a range of 37% to 45%) actually achieved in high-grade movements. In practical terms this means that full-size 10 to 12 ligne movements (hand-winds and autos) should have balances around 9mm or 10mm in diameter, small ultra-thin movements have 7mm balances, and up to 12mm balances for large traditional chronographs. Unfortunately most modern movements have balances 30% of movement diameter or smaller, typically 7mm balances in most full-size calibres. Some might argue that this is due to the high beat rates used in most movements today, but Rolex’s 313x series of fast-beat automatics use a large Microstella balance 35% of the calibre diameter (~10mm). The diameter of the balance wheel will not only have real-world performance benefits, but it will tell you where a manufacture’s watchmaking priorities are. An excellent example of this is was Rolex’s modification of the Zenith chronograph, where in addition to reducing the beat rate, they discarded the regulator and installed a vastly larger Microstella balance wheel.

On the subject of beat rate it appears that there can be no objective consensus, but I will try to point out the pros and cons of the two philosophies, and the reasons behind manufacturer choice:

The advantages of fast-beat (28,800 v/h and 36,000 v/h) are obvious: better isochronism, and better performance in both vertical and horizontal positions even with minimal adjustment or no adjustment at all — one of the reasons fast-beat has been almost universally adopted by mass-producers. In fully-adjusted fast-beat watches positional adjustments are naturally less laborious. The disadvantages are the insidious problems of greater friction, stress, and wear — not just on the escapement where it is obvious, but throughout the wheel train and in the winding systems of automatic movements. For best isochronism it is vital that a watch only run during the flattest part of the mainspring’s torque curve. This is why a movement that is expected to be fully re-wound every 24 hours is typically given a 48+ hour power reserve; only the first half of it is meant to be used. A high beat rate uses power more rapidly and thus requires a stronger mainspring to get adequate power-reserve and thus best isochronism. This increased torque output puts greater stress on all of the pivots throughout the gear train as well as increased friction between wheels (wear). Anyone who has hand-wound a fast-beat Zenith chronograph knows that the winding is relatively “stiff” due to the strength of the mainspring. For automatic winding this requires extra high-mass rotors, and the winding train suffers greater wear as it is caught between the resistance of a strong spring and the high winding force of the rotor. Taken all together, this means that fast-beat movements have shorter service intervals, and the greater rate of wear necessitates more frequent parts replacements. Prominent proponents of fast-beat rates include the manufactures: ETA (and Valjoux), Seiko, Rolex, Zenith, Jaeger-LeCoultre, Glashutte Original, Chopard, and Parmigiani Fleurier.

The advantages of slower beat rates (18,000 v/h, 19,800 v/h, and 21,600 v/h) are less immediately tangible. Lower power needs allow for soft mainsprings, limiting stress and friction throughout the wheel train, winding train, and the escapement. Service intervals are longer and more flexible, and parts replacements are rare — extending practical and theoretical longevity of the movement. The disadvantages are far more tangible: As was noted above, a slow beat movement will generally not perform as well as a fast-beat one, and while slow beat movements can perform very well it requires more skill and effort from the watchmaker to achieve, primarily through laborious positional adjustment. Thus slow-beat is embraced primarily by manufactures of high-craft movements, most of whom consider 21,600 v/h to the best compromise — with measurably better performance than 18,000 v/h and with much less wear than 28,800 v/h. Prominent proponents of slow-beat include master watchmakers like George Daniels and Philippe Dufour, and manufactures like Patek Philippe, Audemars Piguet, Frederic Piguet, IWC, Roger Dubuis, and A. Lange & Sohne, and slow-beat accounts for most of the movements produced by Nouvelle Lemania.

As we have seen above, it is the quality of the escapement which is first and foremost in defining a premier quality movement, but even if the escapement is the heart and soul, it is not the whole. Yet for the sake of brevity I will not delve into the more subjective exoteric issues regarding cases and dial-work, which are just as important as movement quality.

A note on the Geneva hallmark

The Republic of Geneva has provided us with a unique symbol and standard of their fine watchmaking tradition, which applies only to watches from that canton. However, the Geneva hallmark should not be considered a sole “proof” of quality, but rather as the minimum standard to be expected from high-grade Genevese watches, which they should always have in addition to the factors already mentioned above.

Further Reading:

Paragon and Prodigal: Aspects of Genevan Horology

Section 2: Piercing the Veil of Illusions

Next, I would like to go over some things which may be found or featured on a fundamentally high-grade watch, but which in themselves do not necessarily constitute identifying marks of fine watchmaking, and which are probably found most often on less-than-fine watches as a superficial mimicry of quality watchmaking:

Cosmetic Finishing

The “finishing” of a watch movement is something often babbled about, but generally little understood. Finishing is comprised of two general areas: functional finishing and cosmetic finishing. The former is vital to the basic quality of all mechanical watches, and the latter is a final touch of refinement and polish, but in truth finishing is not a clear divide but a continuum, and for most students of the school of craft decoration is an important and integral part of fine craftsmanship. Unfortunately it takes an experienced watchmaker with a microscope to judge functional finish accurately. It is the more flamboyant elements of cosmetic finish which is easy to see – and meant to be seen – and often is erroneously used to judge quality between watches. The problem created by this in the age of the display back is that there is a new and growing trend of “display back finishing.” Here a movement is given an luxury finish on all of the visible elements of the top-plate, while the functional finishing within the movement is neglected, or finished to a lower, if barely adequate, standard. One example which comes to mind are the Chopard LUC calibres 1.96 and 3.96 — the same calibre with different finishing. Chopard makes it clear that the first is a higher grade movement, due to its higher price and features like an overcoil hairspring and Geneva hallmark, but a look at the movements through their display backs would seem to otherwise indicate equivalent quality, and many collectors have treated the 3.96 and the center-seconds 4.96 as such. An experienced watchmaker who has disassembled and evaluated both versions of the LUC calibre rated the finish (decorative and functional) of the 1.96 a “1” on a scale of 10, on par with the best, and the finish of the 3.96 a “5,” or merely average quality.

Cotes de Geneve, perlage, and anglage and other forms of decoration all evolved on first quality movements of the late 19th century, so it is only natural that we instinctively equate them with quality — and it is this instinct which is often used against us. A “luxury finish” is a standard option now available from the industry’s largest supplier of movements, ETA — just one more thing which can be easily mass-produced , or at least as a facsimile. That most cosmetic finishing falls short of what it ought to be is easily seen when comparing a six-figure haute du game piece to a regular production commercial piece. Anglage for example was traditionally hand-cut and polished, but today it is mostly faked with “pressed anglage.” The money that any manufacture can spend upon a movement is limited, and the more that is spent on the superfluous appearance of quality, the less that is available for actual quality watchmaking. It perhaps should be recalled that these often worshipped elements of cosmetic finish are a relatively recent product of the industrial age and powered lathes. The chronometers and other fine watches from the peak of the handicraft era dispensed with the elaborate movement decoration of the centuries prior, using just a simple matte gilt finish (example shown right). Personally, I have long found this simpler style to be more appealing than the Genevois style affected by most, and a matte gilt finish also appears to be the prime preference of modern horology’s few remaining bespoke watchmakers. By no means is there anything wrong with a luxury finish per se, but it should be the icing on the cake, representative of all of the work that has gone into the movement, not a sleight-of-hand illusion.

Recommended reading from Walt Odets:

How to judge the finish of a watch movement

The A-B-C’s of Watch Finish

A comparison of the keyless works in a Patek and a Lange

COSC Certification

In proper English watchmaking terminology, a “chronometer” is a watch with a chronometer escapement (spring detent, pivot detent, or other detent types). Period. End of story. Chronometer escapements are very sensitive to disturbance and are quite fragile, and are not suitable for use in wristwatches — in truth they were only marginally suitable for pocket watches. To the best of my knowledge there has never been a chronometer wristwatch.

The practice of observatory certification which developed in the 19th century was instituted to ensure than marine chronometers were fit for their critical role in nautical navigation. An astronomical observatory would test a watch over several weeks (45 days or more) and grade them on a performance scale, awarding a certificate with the final score and rating — observatory certification was originally not a simple pass/fail test. I should note that in those days it was astronomical observatories which set the time standard, and thus it was their ability to objectively determine minuscule errors over prolonged periods which made them ideal for testing chronometers. Today the time standard is kept atomically, and certification by an observatory would have no special technical significance.

Wristwatches were submitted for observatory certification as early as 1910, and submitting special observatory-grade watches eventually became a form of competition between manufactures, as to who could get the most top scores. As observatory competition became a marketing tool, so too did certification. Separate commercial testing centers (under the official aegis of the state) sprang up to certify wristwatches in volume for sale. This was combined with the efforts of Swiss manufacturers to redefine the term “chronometer” away from its well established technical heritage. In 1951 for example, the Federation of Swiss Watchmakers came up with a wonderfully vague and broad definition, wherein “a Chronometer is a precision watch, which is regulated in various positions and at different temperatures and has received a certificate to that effect,” representing a general trend which transformed the term “chronometer” into brilliant marketing vehicle, permitting watch manufacturers to certify their lever escapement wristwatches (not chronometers) as “chronometers” with official sanction. It has given us watches with extensive marketing copy on their dials, like “Astronomical Observatory Chronometer Officially Certified” (Seiko), “Superlative Chronometer Officially Certified” (Rolex), and stranger things.

The so-called “chronometer certification” in vogue amongst mass-producers today is the final fruit of this long transformation. Today, uncased and often partially assembled movements are tested for a mere two weeks (15 days) in a narrow temperature range to a very generous standard of accuracy, and given a passing or failing result by the Controle Officiel Suisse des Chronometres (COSC, founded 1973). All in all a far cry from real chronometers and the rigor and authenticity of the observatory certification of marine chronometers and deck watches. “Chronometer certification” is not used by any independent watchmakers, and sees little if any use from the top manufactures who were avid participants in observatory competitions. At most these brands now dedicate a single model to “chronometer certification,” seemingly to serve the market for those who insist that their fine watch come with a piece of paper from the COSC.

“Do better if possible, which is always possible.”

— Francois Constanin

I do not expect that this process of discriminating high-grade watches from common quality ones is something that will appeal to the broad majority of those fueling the mechanical revival. Nor do I expect that the cabinotier values of fine watchmaking will dominate and guide the present-day upper echelon of wristwatch manufacture, much less the industry as a whole, where brand image is more important that real watchmaking. But for those in the school of craft there can be no other reasons to study this ancient art. It is not just an appreciation of tradition, and far more than a static horological classicism. It is a quest of principle, of areté, reaching for the simple, perfect form.

Image Credits

Simplicit courtesy of Philippe Dufour

Jules Audemars Reserve de March, and calibre 3090/3900 detail; courtesy of Audemars Piguet & Cie, S.A.

Patek Philippe calibre CH 27-70Q detail by Mark Kolitz

Geneva hallmark (Vacheron & Constantin calibre 1002) by John Davis

Tourbillon calibre detail courtesy of F.P. Journe

Copyright © Carlos A. Perez 2001

All Rights Reserved