Hacker News new | past | comments | ask | show | jobs | submit login

This is a very interesting article, and I like how it's written in the perspective of a functional watch user instead of being about fashion or watch collecting.

I couldn't help but laugh at this part:

> Also, though it sounds trivial, I enjoy the perfect 60 FPS smoothness of Apple Watch’s second hand — a smoothness no mechanical watch could ever match.

Isn't a mechanical watch hand ∞ FPS by definition? Real life has got to be at least better than 144hz :)




I'm going to take your ∞ FPS comment on go on a wild tangent because physics is fun :-D:

Nothing can be infinity FPS because it would be limited by the frequency of the light that we use to see the watch hand. In the case of the visual spectrum, it caps out around 668 THz with blue light (red light is lowest at 400 THz). This would make it "only" ~4.7E12 times better than 144 Hz ;-).


Brilliant!

With that said, though, I believe that the movement of the actual atoms making up the mechanical hands is still happening in an "infinitely smooth" sense, although the way we measure and observe that movement (via reflections of light or perhaps some other type of electromagnetic radiation) may be limited.


Assuming movement is continuous and not in discrete units.

http://en.wikipedia.org/wiki/Planck_length


You don't even need to appeal to Planck length here, because it's clear from more basic quantum mechanics that particles simply do not have classical paths.


Please don't misinterpret the Planck length. Please don't interpret it in any way, it's right in the wiki article:

>There is currently no proven physical significance of the Planck length;...

It's just dimensional analytical playing with some fundamental physical constants. Physicist believe that this scale length may have some connection with the yet to be discovered laws of quantum gravity. It has nothing to with the "finite resolution of the universe". There is no evidence that the universe has finite resolution.


If time or space are quantized, then "infinitely smooth" would be impossible. I guess no one knows.


Really you're probably limited by the speed of sound in the watch hand


Huh. Now I remember why I read hackernews.


To further complicate the issue, the human eye cannot see 668 quadrillion 'frames per second'. Under the best possible conditions, for parts of the eye, 1000 'fps' is the best we can do, though generally it's much, much lower.


Mechanical watches, think Rolexes, have smooth second hand movements but they are acutally still ticking, just about 6 times a second. The Apple Watch, on the other hand, has a second hand with 60 ticks per second so that the second hand is about 10 times smoother.


Some "hi-beat" watches tick 10 times per second (still well below 60). Seiko's Spring Drive (http://en.wikipedia.org/wiki/Spring_Drive) offers continuous movement, but it's not completely mechanical (a quartz oscillator controls the escapement). I would rather have one of these than an Apple Watch: https://www.youtube.com/watch?v=8m-5YpNUwgc

(Edit: only now have I noticed it was already mentioned in another comment)


Correction: Rolexes tick 8 times a second (or 28.800 beats per hour, as is used by watch people). 6 are the vintage ones.


You are wrong. The movement moves at 8 Hz per second, but it actually takes two cycles to actually tick once, so it's actually 4 ticks per second.

The old watches tick at 3 times per second on 6 Hz movements.


No, I'm not. It's 8 ticks per second. I have one and I have checked. Look at this video of the hand up close (it's a 28800 bph movement):

https://www.youtube.com/watch?v=FxgahvNH3q0


I don't know many mechanical watches (especially Rolex) that ticks at 6 times per second. That requires the balance wheel to be beating at 12Hz.

Old Rolex tick 3 times per second on 6 beats/second balance wheel movements, and modern Rolex do 4 ticks per second on 8 beats per second movement.


That's incorrect. The seconds hand moves 6 times in one second on my 7S26 movement (21,600 bph). On a modern Rolex, it would move 8 times in one second. You can check this with a high speed camera (such as the one in the iPhone 6). Nonetheless, there are vastly superior mechanical drives out there that are continuous (Seiko Spring Drive for example).


True as long as there are sufficient number of pixel density to movement of 60 ticks per seconds.


Antialiasing.


And Bulova has the Precisionist line that moves much more smoothly:

> In 2010, Bulova introduced the Precisionist, a new type of quartz watch with ultra-high frequency (262.144 kHz) which is claimed to be accurate to +/- 10 seconds a year and has a smooth sweeping second hand rather than one that jumps each second.


Huh. Cool. I always wondered why they didn't make quartz second hands jump in smaller amounts. Something difficult or not efficient to achieve, I guess.

By the way, Seiko has had a movement for some time that is rated to +/- 5 seconds a year. I think it was on a special edition of the Grand Seiko line (already rated to around 10 seconds a year, so you'd get 5 seconds off less. Yay!)


The guy has probably never heard of the Seiko Spring Drive, where the hands are moved by a spring and regulated by a brake. Therefore they're completely smooth.

https://www.youtube.com/watch?v=ZsBTXgksr4s


Spring Drives are pretty rare outside of Japan. Most mechanical watches today tick 8 times a second. On a small watch it's pretty smooth to me.


The balance wheel oscillates at 8 beats per second, but it requires 2 beats to complete a full second "tick", so most modern mechanical watches actually tick at 4 times per second.

http://en.wikipedia.org/wiki/Balance_wheel#Period_of_oscilla...


It is 8 ticks per second. Look at this video of the hand up close (it's a 28800 bph movement):

https://www.youtube.com/watch?v=FxgahvNH3q0


It's a bummer that video plays at 30fps. This is a case where 60 (does YouTube do higher framerates?) would pay off.



> The guy has probably never heard of the Seiko Spring Drive

Gruber is the unpaid wing of Apple's PR team. I doubt he cares.


I'm not so sure about that, 60fps means that every single rotation of a second hand has 3600 individual steps; does a mechanical watch with a sweep second hand have that much precision? I'm inclined to think not, even from the perspective of backlash the digital version is going to appear significantly smoother to the eye.


https://www.theverge.com/a/apple-watch-review

I found this more informative because he used it a bit more like I expected someone to.


Welcome to Daring Fireball :)


> Isn't a mechanical watch hand ∞ FPS by definition?

You guys are so silly sometimes.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: