A camera that shoots 40k FPS decided the 100-meter sprint final
petapixel.com
petapixel.com
The headline is clearly, in some sense, silly. It’s silly because the usual notion people have of fps is something like video where you need to move your roll of film or read out from your sensor at that rate. The ‘frame rate’ for a photo finish is much more incidental to the way the image is made.
If you have an image from a typical digital camera that is 4000px high formed by a typical two-curtain shutter that shutters in 10ms, you expose 400 rows per ms or, in some silly sense, 400k FPS. If you wanted a 1-pixel high slit to be exposed then that’s an exposure of 1/400000s, which is faster than any camera I’m aware of. But I think the analogy is useful – instead of a shutter moving a slit over the focal plane, exposing different parts of the sensor at slightly different times, imagine the sensor being moved behind a fixed slit and then reading out the values from it. You can reasonably easily imagine doing this with film too – just move it past the slit at a steady rate – not dissimilarly from taking an old-school panorama camera and moving the body instead of the lens.
I think the thing happening here is not a moving sensor but rather a 1 pixel wide sensor (or perhaps a few pixels for colour reasons). This makes it thousands of times smaller than the resolution of the final image so even a fairly typical cmos sensor at 2e9 pixels per second could read 50000 pixel ‘frames’ at ‘40k FPS’. (In practice the number would probably be lower for synchronisation reasons). When your frame is very skinny, that still gives you plenty of resolution.
I don’t like the article because they did some silly arithmetic that produces a big number instead of digging into interesting details, e.g.
- talking about how the system works as a whole (when does it decide to start/stop the image, maybe something about buffering)
- talking about how much light you need and how you get enough
- talking about the optics, how you keep everyone sharp while still getting enough light, how you even focus such a thing
- talking how you make sure the camera is setup fairly (eg perpendicular to the lanes, able to get a good view of all the lanes)
- maybe something about reliability and how you avoid the bad scenario of the system failing when it matters most
No need to imagine, this is how photo finish cameras at horse and greyhound races have worked for decades.
Which is a fairly big deal as sometimes it isn't set-up correctly.
Who won the Amstel Gold Race? Human error in photo-finishes (April 29, 2021 ) [1]
https://youtu.be/7Xnr805bm4E?feature=shared&t=417
The track is all white because the finish line is all white. And the Omega/Olympics flag "banner" behind the runners is not a banner, it is a led screen, one pixel wide.
The shadows are what's projected on the line from the stadium lights, and the shape is how they changed in time.
Again, not well explained in the article.
That’s also the reason for the distortions. It’s not a single frame taken at one time.
As the runners' shadows cross the finish line, they are recorded by the camera in the same way that the runners are.
This short video might be a better explanation of the "line camera" concept of multiple photos of a single line being stretched out over the x-axis :
You can see it here at 9m50s:
https://youtu.be/7Xnr805bm4E?feature=shared&t=590
It's just a single animated strip, one pixel wide. I assume the camera array is on the opposite side.
Writing a check for $12,000,000,000.00 will do that to you.
https://marketrealist.com/p/how-much-did-nbc-pay-for-the-oly...
https://www.youtube.com/watch?v=tcxyXnPIF4o#t=2m45s
(You can see it in normal speed too, but I can feel the formation of shapes better in the slow-mo, instead of it just feeling like blinking)
I guess pick one or the other depending on your country (or if neither works, search for one that does).
Clever.
I think they did a pretty good job, it's only slightly stretched for an event when there was only 0,12 seconds between the first and the last runners, which is unusual.
And of course when I say ‘speed’, I mean [width of strip] / [time between the capture of each strip].
Now that I look at it, the logos in the men's photo finish look stretched, so maybe it's calibrated for the women's pace.
Questions that came to mind and answers as far as I can tell:
Q: Are the scan lines parallel to the plane of completion?
A: There is only one scan line and it is parallel to the plane of completion. "In track, the cameras are only focused on the 5 mm near the finish line." [0]
Q: If yes, is the sequence of lines scanned in the same direction as the competitor movement?
A: There is only one scan line.
Also interesting: https://www.hodinkee.com/articles/olympics-2024-mens-100m-ph...
This effect is the origin of the "fast race car slants forward" trope:
https://about.usps.com/news/national-releases/2011/images/pr...
https://en.wikipedia.org/wiki/Strip_photography
Early film versions used a highspeed spinning slit aperture to film fast objects. This paper from 1931 shows some very impressive results for ballistics, including the shockwave: https://royalsocietypublishing.org/doi/pdf/10.1098/rspa.1931...
Outside sports, digital line scan cameras are used in various quality control applications (objects on conveyer belts, vehicle mounted road/rail scanners, etc). This unit can film an 8k px strip at a rate of 80khz.
It's actually hilarious what you really need to scroll it to see it on your Wiki page.
For a while, handheld scanners were a thing: hand-held strip cameras with roller wheels you would swipe across a page.
In drift-scan imaging you keep the telescope pointed at a fixed location and you continuously read out the image at a rate that matches the motion of the stars across the field of view. This allows you to continuously collect imaging from a strip across the sky.
To the question of the Omega logo and Olympic rings asked below, this article states "In fact, the entire photo is literally the finish line viewed one pixel at a time (the Omega and Olympic branding is added at the top)."
Another weird (until you think about it) thing that’s true: every pixel in the image is at the finish line.
This is clearer, I think, from considering how the banner is made: as others have pointed out, it is produce by having a single row of LEDs in line with the finish line, each one being turned on and off in a sequence which results in the banner shown in the picture.
But since the company Omega produced this photo it actually makes sense to artificially add their logo next to the Olympic rings.
> And the Omega/Olympics flag "banner" behind the runners is not a banner, it is a led screen, one pixel wide.
So the one pixel wide scan line camera outputs and image that has time as horizontal axis, and the one pixel screen that draws the omega banner is a 40kfps single line video that does marketing one the photo finish that likely serves as a clock sync check.
If fairness were the primary consideration, then they would use any old junk camera and be generous in calling ties. This kind of theatrical accuracy is there purely so that Omega can advertise.
But I don't quite see how using less accurate measuring tools makes sense. At the end of the day getting across the finish line first is the only thing that really matters. Seems to me that more accuracy can only be a good thing
The extreme level of precision, I suspect, is not the same at the finish and the start.
So race organizers try to make things as fair as possible, but everything beyond that is just “part of the game.” The athletes accept that, and spectators should as well.
We don’t need to run the race.
It is not. They do a great job of getting people to think that it is, however.
I join millions of other people who find it enjoyable to watch. By that measure it's meaningful. In fact, more meaningful than the _vast_ majority of Olympic events which have less parity and orders of magnitude fewer viewers.
Every competition has factors preceding and during the moments of measured performance that are like this. You could also say that a runner’s performance is about their ability to train consistently, apply self discipline to their daily lives, perform under pressure, etc.
It seems odd to me that in the sum of all moments leading up to the start of a race, somehow reaction time would no longer be relevant.
If a runner gets distracted by the crowd, that’s just as impactful to the race but has nothing to do with their potential speed. The race is a measurement of all factors - mental and physical - and all participants are measured the same way.
If a faster auditory reaction is intrinsic to the abilities of the runner, then that's an advantage they can exploit.
I don't think this method would be as entertaining to watch.
It's not possible to build swimming pools to tolerances such that timing to 1/1,000th of a second would be fair (that is, based on athletic performance rather than distance traveled), so swimming events are only judged to the nearest 1/100th of a second. Finishers within the same 1/100th are considered to have tied.
<https://news.ycombinator.com/item?id=41128437>
<https://olympstats.com/2014/02/12/timing-accuracy-at-the-oly...>
FINA standards: <https://resources.fina.org/fina/document/2022/02/08/77c3058d...> (PDF).
The docs I've linked aren't fully clear on this and I've not found official rules.
The team also makes really interesting stuff on other sports, such as Beach Volley. Worth checking how it's done.
[0] https://www.theguardian.com/sport/2012/jul/02/jeneba-tarmoh-...
[0] Zilog Z8 BASIC model controller.
The x axis in the image (left right) is the time domain.
Imagin instead of the people running from left to right, instead they were still and the pixels were moving right to left, or if that breaks your brain, think about if the track was a people mover and the runners were a static diorama moving left to right and the sensor is just it over time. But also none of that is real.
It looks plenty reliable as long as the camera was positioned and calibrated correctly (which I assume was carefully tested beforehand).
They don't.
The horizontal axis is time. The red lines mark the time each runner's torso was counted as crossing the finish line.
Here is the image of the camera:
https://www.swatchgroup.com/sites/default/files/inline-image...
There have been times in the past when photo finish cameras have been installed crooked relative to the finish line painted on the ground [1] and while I have no great love for yet-more-advertising, this does serve to confirm that the camera was pointed in the right direction :)
We can still use cameras for visual confirmation but transponders are much more accurate than any camera.
For the longer distances, the lap times and finish times show up almost instantly when there's a big gap between athletes, although I believe there's a laser at the finish line too.
The first "line photo" is the right-most column of pixels on that photo. The next photo is the second-to-the-right column, etc. This way the winner can be determined as the line first photo that has the contestant's torso in it.
On second viewing, you can see that some things are very distorted and this is obviously a composite as has been explained.
If you look at the still image in the article, Lyles' right shoulder is leaning forward and visible to the camera, positioned to his right.
But Thompson's left shoulder seem to be leaning forward, but is hidden in the camera by his head and neck. It's possible Thompson's left shoulder is ahead of Lyles' right shoulder, but the image doesn't seem sufficient evidence to conclude on that.
What do other think?
> According to UCI regulations, the camera is the finish line.
> The painted line on the road or track is placed there as a visual reference for riders and spectators, and obviously organizers try to align the painted line with the photo finish camera as best as possible. But that's all it is - a visual reference. The finish line is officially defined as the plane of the photo finish camera, not the painted finish line.
https://www.reddit.com/r/bicycling/comments/1emivk1/dont_if_...
So I can't imagine a human can take a conscious step to get a FIVE millisecond advantage over another.
In reality the gun is usually on the inside of the track, so it would be Lyles overcoming this (negligible) advantage.
Tbh, it's all just bullshit to me. Less than 20 years ago, this would have been an ex-aequo, and glorifying someone who happened to dive 5-milliseconds-better is just sad. The obsession with "having a winner", for the benefit of sponsors and image contracts, has taken sports in some ridiculous places.
The torso was chosen as the body part for determining when a runner crosses the finish line precisely because torsos were easy to identify/distinguish and there is no ambiguity as to a runner's torso.
Although the pelvic area is anatomically part of the torso, for consistency in photo finish judging, it is more practical to define the lower end of the torso as the horizontal cross section of the body through the hip line . . .
IAAF Photo Finish Guidelines (June 2015): https://worldathletics.org/download/download?filename=4423f7...
UCI is Union Cycliste Internationale, i.e. the biggest international governing body of cycling events. Not athletics (i.e. track and field). I checked out the C1.1 & C2.1 rulebook from the IAAF and I'm not seeing an equivalent rule. Just a lot of rules for how to place the camera and mark the lane lines and finish line in 19.13
index of all IAAF rulebooks https://worldathletics.org/about-iaaf/documents/book-of-rule...
direct PDF link to the English version of the 17 JAN 2024 C1.1 & C2.1 - Competition Rules & Technical Rules rules (2.5 MB PDF) https://worldathletics.org/download/download?filename=675a00...
> Book C 19.22: Times from the Photo Finish System shall be regarded as official unless for any reason the appropriate official decides that they obviously must be inaccurate.
The same problem could also be addressed by having two cameras, one on each side of the track, synchronized and recording the same plane.
One could align the cameras into the same plane by putting a vertical array of distinctive marks (e.g. rapidly blinking LEDs) on each camera, above and below the optics and in line with it, such that each camera would see the other’s marks when correctly aligned.
This doesn't work, as light beams diverge. Even taking a high-quality laser beam with a divergence of 0.1 milliradian (for comparison, a typical laser pointer is about 1-2 mrad), after crossing the 11 meters width of a 9-lane athletics track, you end up with a beam diameter of 2.2 millimeters. At the 10 m/s speed of the athletes and a 40000 fps framerate, they travel 250 micrometers between each frame.
More concretely, if the cameras are photographing along different but parallel planes, then they won’t see each other.
More concretely, you can have two cameras photographing along different but parallel planes that do see each other.
Not to mention there was no controversy afterwards.
Then I realized it's me thats stupid.
So much for supposed freedom of speech, Elon
https://www.phantomhighspeed.com/products/cameras/tmx/7510
The spec sheet lists up to 512GB of RAM depending on configuration. There's no way you're scaling cache-like SRAM up to those numbers.
Regular high-speed cameras that shoot a video are a whole other story, though.
At some point this seems about as fair as a coin toss to me. :)
Edit: I missed that this point had already been made https://news.ycombinator.com/item?id=41230449
It shows that you won.
Some potential purposes of a race:
- To compete in order to see who ran the fastest - To determine who the best runner is - To provide entertainment - To build a career for the runners - To show the result of the runners' talents, hard work, and dedication - Build a sense of national pride, ...
My guess is that many people would think of the first point as the "real" purpose of a race. In that case, measuring as accurately as possible makes sense, since it will give the best answer.
If you agree to some extent that races are also about determining who the better runner is - then you must also consider that there is a point after which the runner doesn't control the outcome. Wind seems to be a big factor; the maximum allowed wind is 2 m/s, and the average tailwind for all world records since 1977 is 1.18 m/s (none were made with a headwind). The wind can't be exactly the same in different lanes, so for two very similar runners - there is also a bit of luck involved in the outcome.
I don't agree with this to a meaningful extent. "Better" is too vague to be meaningful.
A purpose of the race is to determine who won the race. Sometimes that is to some extent the better runner, but not always. I don't care who the better runner is. I want to know who won the race.
Also from the article:
> an image taken by an Omega camera
Photos no matter how fast still require human objectivity in evaluation.
Instead there should be some digital signal or laser reflection that is precise?
I don't like the rule where it's any part of the athlete's body first to cross the line, it should be the first athlete to ENTIRELY cross the line, no part of them still remains before the line, not even their trailing foot.
What's the difference?
a) Why entirely across, and not just the first to touch the line?
b) I get that basing it on the whole body feels a bit more objective, but it would probably also lead to people diving over the line to gain a few milliseconds. As well as looking a bit silly it could be dangerous. The lean through a (now virtual) tape at chest height is traditional.
I think it's any part of the athlete's torso, which corresponds to the red lines on the image.
Though I wonder if a runner could gain an advantage using very long breast implants?
Photo finish is one of the best methods.
Where is the difference between first body part crosses the line vs the whole person?
It's still hard to know who has won in a photo finish.
No, because then you'd be changing the distance of the race.
Runners start behind the starting line. So for consistency, they finish when the first part of their torso passes the finish line.
Your suggestion would only make sense if they started with e.g. their trailing foot touching the starting line and the rest of their body in front of it. But that's not how they start.
Also, if you based it on foot position rather than torso position you'd be introducing a rather random element, since it would change based on what part of their stride they were in. Torso position is much more continuous, and therefore is a far more accurate assessment of athletic ability.
These photo finishes are really precise. 40k FPS, that's 25 microseconds, or about a quarter of a millimeter in distance (at 10m/s running speed). If you look at it, with the lines drawn, it is very obvious who's won.
The organizers took some time to double check everything since it wasn't obvious from the naked eye, but with the photo finish there is absolutely no ambiguity. In fact, the photo finish is 400x more precise than what is required here.
> I don't like the rule where it's any part of the athlete's body first to cross the line.
It is not the case, the article says the torso matters, not "any part". Also, the athletes stand completely behind the starting line, it makes sense that the race ends just as soon as they touch the finish line, otherwise it wouldn't be 100m but 100m + one body length.
The top two runners in a marathon keep pace with each other, both know that they're holding some energy in reserve for the final sprint. Perhaps one sprinter could try to push themself a little faster throughout the entire race, but I suspect the consequences of running a little over your optimal pace are worse than running a little under your optimal pace, so they err on the side of running a little under their optimal pace and save energy for the final sprint.
They are usually attempted at smaller events, coordinating with other runners ("hares") to achieve the perfect pace, without having to worry about the result (hares will typically drop out throughout the race).
If you want to see what happens when the favourites run a little too hard a little too early, check out the 1500m final: Ingebrigtsen and Kerr were supposed to slug it out between themselves, but they both pushed too early (Ingebrigtsen unwisely pulling from the front most of the time), so they got to the sprint too tired to match the surging Hocker.