“The subject of accuracy in mechanical watches is a funny world,” laughs Bart Grönefeld, “because the fact is that, no matter what you do, they’re already inaccurate compared to Quartz watches.” That, he concedes, doesn’t stop watchmakers like him from trying, “taking tiny steps forward to improve the precision of something that’s already backward,” as he puts it, primarily by tackling the mechanical watch’s key nemeses: shock, magnetism, gravity and constant force. 

Indeed, at the high end, some watchmakers periodically return to finessing historic mechanical methods of giving a movement that fraction more accuracy. These are often complex - if, on paper at least, typically based on somewhat basic physics - expensive and come with their own problems. So why bother, for all the practical difference any one of them will make?

Zenith’s product strategy director Romain Marietta argues that it’s a nod to the roots of watchmaking, its pioneers seeking precision because of their devices’ then life or death relevance to the likes of navigation or military manoeuvres. “So at one point precision suggested the best in class,” he says.

But also, because, argues Clemence Dubois, chief product officer for Girard Perregaux, while not all collectors are fascinated by precision - others may be more excited by the craft, decoration or history of watchmaking - and while ever greater precision may provide little real world benefit, it nonetheless remains a benchmark in the advance of mechanical watchmaking. And, with the deployment of new materials and manufacturing technologies, such advances will only keep coming.

“Technology will allow watchmaking to design and make movements that would have been impossible 150 years ago,” she says, “so we may well yet see more accurate watches. Of course, there are those collectors whose idea of precision is the goal of zero seconds lost per day. For others, just two seconds is more than fine...”

“The remontoire is actually really simple to understand if you imagine a small, very short gear train and a very light barrel,” says Bart Grönefeld, of the device used in the company’s 1941 Remontoire Constant Force, inspired by the remontoire mechanism used in the local church, which the Grönefeld’s have been maintaining for three generations. “Wind it and it runs perfectly - but for a very short time, and then it’s dead. So, you add a mechanism, the remontoire, a constant force mechanism, that provides new energy every few seconds and keeps that smaller movement running. Is adding this worth it in terms of precision? Yes. In terms of effort? Ha, no...”

A remontoire, he explains, is delicate - and it’s hard to make a delicate system reliable. Given its complexity, one might argue, as Grönefeld suggests, that the gains in precision are too small. And then it typically adversely affects the fragility of the system to position it so as to see it working through a caseback window, which limits its appeal for some.  

“And yet watches shouldn’t be designed by book-keepers, and the fact is we could have sold twice as many of our Remontoire watches as we did,” he adds, “which shows there’s a fascination with precision at this level, when pursued through such a traditional and very niche idea.”

This niche idea is that of giving constant force to the escapement, long a challenge to mechanical watchmaking seeing as the energy from the main spring, inevitably, gradually decreases as it relaxes from its fully-wound, maximum force state. So, John Harrison, he of the world-changing marine chronometer, came up with the remontoire back in the 1760s.

It’s clever because when its own spring - or, in larger mechanisms, maybe a weight - reaches the end of its power, its re-winding is automatically triggered - at an interval of any time between one second and one hour - and it’s this frequent re-winding that adds to accuracy because it averages out any changes in the force of the remontoire itself.

“It is a great idea, even if it costs a lot of money to develop,” says Grönefeld. “I think watchmaking is very close to the maximum precision for mechanical watches already anyway, just through using the best materials. But sometimes the traditional ways are worth revisiting.”

It was back in the mid 17th century when Christiaan Huygens, co-creator of the spiral balance spring and pioneer of pendulum clocks, first recorded the effect of resonance. This is when two oscillators, close by to each other and with the same frequency, start to beat in synchrony. He first noted it in pendulums - two pendulums in resonance soon end up swinging in precisely opposite directions, no matter how or at what amplitude they started out moving. 

Resonance is not easy to see at the small scale demanded by wristwatches - not least because getting the oscillators at a very similar frequency is no mean feat, requiring daily adjustments - but it’s a natural phenomenon that Abraham-Louis Breguet later explored in similarly arranged balance springs too. If the rate of either oscillator drops off, the other, as if by magic, corrects it. Result: greater accuracy.

That, in fact, is crucial to the working of any mechanical watch, since there’s resonance between balance wheel and hairspring - but it’s sufficiently complex that Breguet didn’t pursue the idea for watches with any great conviction. Only a few makers, among them Armin Strom, Haldimann and F.P. Journe, have done so since. 

Journe has been pondering this rather exotic form of movement since the early 80s. Come the early 2000s he launched his Chronomètre à Résonance and last year, to mark its 20th anniversary, introduced a technically more complex version, albeit one which still, fantastically, transmits energy between the two balance springs by waves across a backplate rather than by direct linkage.

“In theory,” says Romain Marietta, “the faster the rate at which a watch movement operates, the more accurate it is. And in principle, you can make an escapement discharge its energy at a faster rate, but then all the other components in the movement have to be able to move faster too. And not just the movement. You have to indicate by the use of hands, of course - they have to move, and they have their own weight as well...”

Marietta is only just getting started on the trials of high frequency movements too - the likes of the Grand Seiko SBGH205, Audemars Piguet’s Jules Audemars Chronometer, or Zenith’s own new caliber 3600 have put this solution to accuracy through its paces. This, thanks to the application of seriously high-powered CAD and CAM, is an overhaul of its legendary El Primero, such that its seconds hand now moves across the dial six times faster than a ‘regular’ chronograph hand, allowing for measurements of 1/10th of a second. 

There are the manufacturing constraints: it’s one thing to make a one-off watch operate at a high frequency - like its limited edition Defy Inventor, which replaces the balance, balance spring and lever with a single piece made of silicon - and it’s another being able to serialise that in production, and for it to be reliable. It’s also no good having a movement that, as he puts it, “works at speed”, if it can only do so for a few minutes. The watch requires a functional power reserve, with most collectors now expecting between 40 and 60 hours.

“So you can always have more power reserve with a lower frequency mechanism and less accuracy. But it’s the demand too for greater accuracy that means in five decades we’ve only extended power reserve by around 10 hours. It’s a balancing act,” he explains. “Yet certainly there’s a sense that higher frequency is a way to more precision. And we can’t know if there’s a limit to how we might achieve that because it’s probably a matter of materials yet to come [to watchmaking].”

But will it ever compete with Quartz? Is that the seemingly impossible goal? Compare Quartz, which runs at 32,768 HZ, with most mechanical watches, which operate around 2.5 to 4HZ, or 18,000 to 28,800 vibrations per hour (with a swing of the balance wheel in either direction equal to one vibration). Yet the latter will already be fast enough to give dependable, accurate time. 

Zenith’s answer: it’s now working on finessing a mechanical movement with the target of being able to show 1/1000th of a second. “Well,” says Marietta, “it’s a competition really. It’s about proving we can.”

“The fact is that the moment you start to use a watch, gravity begins to deform its hairspring, and that will negatively affect the time-keeping,” laments watchmaker Romain Gauthier. “So even the regular hairspring - which is the heart of any mechanical watch - is a problem for accuracy.”

Of course, it’s not just the universe-defining force of gravity that’s at play in a mechanical watch. There’s conservation of energy too: that hairspring produces ever less and less energy the moment it’s released from its fully wound state. It’s these fundamental truths of watchmaking that turned Gauthier to exploring to potential of the fusee et chain, or rocket and chain, an idea also explored by the likes of A. Lange & Söhne and Breguet. Or, more specifically, how can isochronism - making the rate, and hence accuracy, of a movement independent of available energy - be achieved? 

The fusee - or rocket - is a cone-shaped component connected to the mainspring barrel by a chain. As the spring is wound, a rotating barrel wraps a chain around the fusee. Then, as the power in the mainspring decreases, the tension from the chain around the barrel, now rotating the opposite way, ensures that a consistent amount of energy is released through the movement - rather than an ever-decreasing one. It’s an idea introduced by one Gruet of Geneva in 1664 and which was popular - at least in clocks and pocket watches, which had the space for what is a positively out-sized set of components for a watch - until advances in stronger, more flexible alloys made it more or less redundant.

“It’s a simple principle really,” reckons Gauthier. “If you give constant energy at the centre pinion that will affect the entire gear train up to the balance, but the constant input has to be at that pinion point. My thinking was to take another look at the chain, which seemed the tricky aspect of this complication - and I’ve seen so many pocket watches broken just because the chain had rusted and there was no-one who could repair it.”

For his Logical One piece Gauthier devised a shorter, thicker, re-formatted (and patented) ruby-link chain with sufficient simplicity to make servicing feasible, and setting the cam and the barrel on the same plane to make the transfer of energy more efficient. “The basic idea is not a new one,” concedes Gauthier, who’s now working on a new, already patented constant energy mechanism due for launch late 2022, “and the watch industry does tend to attempt reinterpretations of mechanisms used in the past. Really, if we want much more consistency and precision in mechanical movements, we’re going to have to start thinking about the question of constant energy in new ways.”

“Silicon isn’t magnetic either,” Girard Perregaux’s Clemence Dubois notes, almost as an after-thought. But what really counts, she says, is that the piece the company devised in 2008 and, five years later, finally built right into a more traditional escapement, is six times thinner than a human hair, which only silicon allows. 

Why is this important? “The crazy idea here is to introduce a new component that can capture the energy from the barrel and release exactly the same amount throughout its cycle and then start the process again, making the energy constant,” she says, “and that’s only possible because new materials, in this case silicon, allow for parts that are super reliable and don’t break over time with all that flexing.”

A comparison helps here. Those more historic means of improving the accuracy of a mechanical movement - the likes of a fusee and chain, or a remontoire, both sitting outside of the escapement - give more constant force than a traditional escapement alone. But Girard Perregaux’s arrangement gives an escapement that is powered continuously at the same level throughout its cycle. This very, very thin bit of silicon stores energy up to a certain point, when it’s fit to burst, so to speak, then transmits it in one instantaneous hit - not gradually - before starting the cycle again. 

The idea is said to have been that of watchmaker Nicolas Dehon, who noticed that if you bent a train ticket into a C shape between thumb and forefinger, then pressed on the curve, there’s resistance until it finally pops into its mirror image, supplying force as it does. Of course, the distance between that insight and the working movement - with its two escapement wheels and twin barrels, each with two springs - is far, far more complex. But it’s a sign too of things to come in the on-going pursuit of accuracy.