This is obviously a high frame rate photo of a very close cycling race finish. What gets me is not the closeness of the race (although the closest I've ever seen). Rather, it's how the spokes appear to bend, including "displacement" of the drilled holes in the wheel. I find it extremely unlikely this is actually happening in real life (quantum physics anyone?). Can anyone explain what's going on here photographically? It reminds me of how sometimes spinning airplane propellers are distorted in some photos I've seen.
Search for Rolling Shutter. I'm sure this is a duplicate but I'm on my iphone and can't find it right now.– Jim GarrisonAug 27, 2017 at 18:28
4See: What is a rolling shutter? When do I have to be aware of it?– mattdmAug 28, 2017 at 1:44
Photo-finish photographs are made using slit-scan cameras, not conventional cameras. The left to right relationships in the image are temporal rather than spatial...the distance in the image between the front and the rear of each bicycle reflects the amount of time between the front and the rear of the bicycle crossing the finish line (and passing in front of the slit).
The bending of the spokes reflects the different times at which the spoke crossed the finish line and was captured through the slit. The entirety of the hub of the wheel crosses the line in a shorter amount of time than the circumference of the wheel.
Of course the spokes are not bending. It is an effect of (mostly) focal plane shutters. These take perhaps roughly 1/3 of a second for the shutter opening to move across the frame (speaking 35 mm film). Faster shutter durations are implemented by a second curtain following closely behind, closing the shutter opening after the measured delay. So a 1/4000 second shutter is implemented by a 1/4000 wide open slit moving across the frame in about 1/3 second.
Most focal plane shutters today move across the short frame dimension (faster). My guess (just my guess) is that this one was an older shutter (old camera, not the picture, 30 years or more) that still moved horizontally, left to right as we see it. Speaking of bottom half of the wheel, the outer end of the spoke was captured earlier in time than the inner end, so it appears curved after the wheel rotated more (shutter slit moved right).
The early classic example of this effect is 1912 picture by Jacques Henri Lartigue.
It is why cartoons show race car wheels as elongated ovals to indicate speed.
1I'm no expert on this, but it seems to me your answer should be treated as no more than (incorrect) guesswork, if @ben's information is correct (which it presumably is)– osullicAug 27, 2017 at 18:50
The effect in both is the same, the fast shutter speed in essence creates a slit scan shutter.– thsAug 27, 2017 at 20:22
1Right, this "slit" is exactly the same thing as a focal plane shutter (used in most SLR and DSLR today). Here's an article explaining it's the same thing, which even includes the same classic Jacques Henri Lartigue photo. people.rit.edu/andpph/text-slit-scan.html Concept is nothing new, the photo is 1912.– WayneFAug 27, 2017 at 22:58
2I'm not sure where you get 1/3 of a second from. Most modern cameras have x-sync speeds faster than 1/200th of a second, so the leading curtain has to move at least as fast as that.– BillDOeAug 28, 2017 at 0:02
11/3 second is several orders of magnitude too slow for any modern focal plane shutter. Curtain transit time for most FF cameras are around 1/300 second and for most APS-C cameras around 1/400-500 second. The transit time must be shorter, not longer than the flash sync speed. Aug 28, 2017 at 3:57