If length contraction is a physical phenomenon, as many (not including myself) believe, then at a trillion frames per second, Raskar photography should be able to pick up the length contraction of an ordinary high speed bullet just by counting the number of frames it occupies when moving. Just like an atomic clock, the Raskar method should be able to act like a high resolution time microscope to see relativistic effects for low relativistic speeds.

So if a bullet might appear length contracted, why isn't the light pulse? We would have the same perspective to the bullet as to the light pulse. The light pulse is created by a clock in our frame but the clock of the light pulse's frame should be at a standstill relative to ours which means we shouldn't even be able to see the light pulse's frequency which should also be at a standstill.


Sensing length contraction has nothing to do with the framerate of the camera. The only thing that matters is the camera's resolution.

Let's be extremely charitable and say you're talking about a bullet from a high-power rifle, which has a muzzle velocity of around $1200$ m/s (source: https://en.wikipedia.org/wiki/Muzzle_velocity). The Lorentz factor at that speed is $1.000002$, so the length of the bullet will decrease to $99.9998\%$ of its at-rest observed size. Given that a typical bullet for one of these rifles (say, the .220 Swift: https://en.wikipedia.org/wiki/.220_Swift) is around 12 mm long, the magnitude of the length contraction will be around $24$ nm. Since the wavelength of the light used to observe the bullet is between $400$ and $700$ nm (which sets the smallest length scale that a visible-light camera can reliably observe), this isn't really observable by any camera that uses visible light.

  • $\begingroup$ I think the OP is taking about youtube.com/watch?v=7Ys_yKGNFRQ&t=54s in which a light pulse (at 8:43) is moving through air with a gamma around 41 (given n-1 x 10**6 = 273). $\endgroup$ – JEB Sep 22 '19 at 16:26
  • $\begingroup$ Ok but I meant count the number of frames the bullet takes up. If length contraction is real, a moving bullet should take up less frames. $\endgroup$ – ralfcis Sep 22 '19 at 16:28
  • $\begingroup$ @JEB Sure, but the underlying assumption of the question (that "Raskar photography should be better than normal photography at sensing length contraction") is incorrect. $\endgroup$ – probably_someone Sep 22 '19 at 16:28
  • $\begingroup$ I'm specifically talking about youtube.com/watch?time_continue=7&v=Y_9vd4HWlVA $\endgroup$ – ralfcis Sep 22 '19 at 16:40
  • $\begingroup$ I calculate 1.2e3m/s x 1e-12s/frame = 1.2e-9 m/frame resolution. So if the bullet is 24e-9m shorter the camera would pick that up as 20 frames shorter. But I get a different gamma than you. My calc is the bullet is only 9.6e-14 shorter which is less than a frame so the camera would not be able to pick it up. sqrt(1-(1.2/3e-5)^2) = .999999999992 not .999998. Is my calc wrong? $\endgroup$ – ralfcis Sep 24 '19 at 13:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.