(2023-10-02) Can a hacker's watch be... mechanical?
---------------------------------------------------
In case you didn't notice, I have added an instruction on DIY micro notepad
to the LuxDocs section. This is just something I wanna preserve from my 
chronovir.us blog, as well as some other useful things from there (stay 
tuned for updates). Anyway, yes, I had touched some fully analogue topics 
there, as well as the topic of what can really be considered a "hacker's 
watch" versus what the mass culture tries to position as such. There, I also 
performed some calculations I really don't want to describe now, but the 
bottom line was that any watch can store some additional bits of data if 
treated properly. But here, I want to touch another aspect of the "hacker's 
watch" concept, that is, what if we just use it as a watch and not try to 
squeeze anything else out of it? Can even an automatic watch qualify in this 
very case?

From the philosophical point of view, having a mechanical watch does indeed
go well along with the LPC ideology — it is fully autonomous, never needing 
to charge or change batteries, at the cost of requiring more attention 
throughout the year due to reduced accuracy and having to adjust the date 
every two months. Also, it can be vulnerable to strong magnetic fields but, 
on the other hand, fully immune to EMP attacks. Obviously, a hacker's watch 
should be rugged, functional and accurate enough for its intended usage. 
This is why, if we're talking mechs, it should have a decent amount of water 
resistance (screw-down caseback and screw-down crown), scratch resistance 
(sapphire glass and no domed crystals), strap versatility (even lug width), 
good amount of lume on hands _and_ markers, a chronograph or at least a 
rotating bezel and — most importantly — a reliable automatic movement that 
can be regulated at home, obviously with the second hand hackability and 
handwinding option. Who does all this for the price that doesn't make you 
cry if anything goes wrong? Orient does, for one.

Not to say the Orient RA-AA0001B (aka Kamasu Black) isn't heavily overpriced
where I live — well, it is, but I bought it without a slightest hesitation 
when I saw it selling here. Yes, replacing the bracelet to a 22mm steel mesh 
strap was the first thing I did, because the stock one made it so heavy for 
me I didn't even bother downsizing it, but other than that, it's a step up 
compared to Invicta 8926OB in every aspect: the crystal (sapphire vs. 
mineral with that stupid date cyclops), the bezel action (much smoother than 
in the Invicta), the lume (really shines all night through), the bilingual 
weekday display (as someone who's learning Spanish right now, I really 
appreciated it) and... the movement. Yes, it is something better than the 
NH35/4R35 (or NH36/4R36 if we're talking the day-date version) it's directly 
competing with. It is from the same Epson's family as the small-second F6222 
in RA-AP0003S, the new in-house Orient's movement generation, the one and 
only F6922. By the way, there also is an unbranded version of it called 
Epson YN56, but it's not so popular in OEM/ODM mechs as NH36 as of now. 
Anyway, adding to the whole package, this movement really is the star of the 
show that truly justifies 3x the price of 8926OB.

Out of the box, the Kamasu had been displaying not so great of an accuracy:
in my 24/7 wear mode, the first day showed the +12 spd, then it was +11, +10 
and finally settled on about +8.5 seconds per day. And the keyword here is 
"settled". The movement has shown extremely good positional stability of its 
timekeeping. This is why I didn't worry at all about the +8.5 spd deviation: 
I knew that I could regulate the movement and be fine with a very stable 
_and_ accurate timepiece. Well, guess what? I did. After all, I do have some 
watch related tools — not top-notch but they get things done. Of course, I 
don't have a hardware timegrapher yet and not sure whether or when I can get 
one at all. I also haven't created a software timegrapher myself, so I had 
to rely on the "Watch Acccuracy Meter" application I found for my 
smartphone. Sure enough, it's not very precise but definitely helps you 
orient (no pun intended) your regulation efforts in the correct direction.

In F6922, the speed regulator is the smaller, topmost lever above the balance
wheel (the lower and the thicker one is the beat error stud — don't ever 
touch that unless you have a proper hardware timegrapher!), and the + and - 
signs opposite to the speed regulator show you where to turn it to speed up 
or slow down the movement. Now, I'm not exaggregating anything when I say it 
really takes micrometric precision to move this lever to the required 
position, and it took me a good half an hour in total of two large efforts 
to finally get it close to what I wanted. Also, as the operation was done in 
the "dial down" position, what I didn't get at the first effort (but did at 
the second) is that I needed to subtract about 5 seconds from what I was 
seeing on my software grapher. After finally getting the closest to what I 
could get with no special equipment, I screwed the caseback, set the watch 
to the correct time and started a new accuracy measurement.

Speaking of which, as much as I despise Rolex and its subbrand Tudor, I have
to give them one credit for testing their movements and certifying every 
watch to not deviate beyond -2/+2 spd. Not saying this could justify their 
exorbitant pricing (and no less expensive service) but it definitely shows 
that the manufacturer at least tries to give them credibility as, well, 
watches, not just jewellery. A minute per month is something of a deviation 
that most rich people can live with, I guess. Well, you know what, maybe I 
already wrote this but I have a wonderful Casio W-800H in my collection that 
is 24 seconds slow per month, which disappointed the hell outta me: only 6 
seconds over the manufacturer's negative deviation allowance (most 
mass-market quartz watches are guaranteed to have +/-30 s/month). Do you 
think there is a random Rolex or (mechanical) Tudor in the world that's 
currently more accurate than this particular Casio? Maybe, why not?

But back to my newly regulated Kamasu. So, can you guess how much deviation I
could register in 48 hours after this regulation?

A bit under two seconds fast. Yes, this means under +1 second per day in the
24/7 wearing mode. I'm going to wear the watch all the way throughout this 
October and then tell you the final monthly deviation, but you can already 
see that it's going to be well within the quartz-grade monthly allowance, 
let alone various mechanical movement certification ranges. And that's after 
a single regulation session that even an amateur like myself can easily do 
at home.

Now, the question is: do you still want a Rolex or a Tudor?

I know I don't. I want a bit smaller, titanium version of the very same
Orient.

--- Luxferre ---