0.005 seconds: How high-tech cameras decide gold at the Olympics (2024)

On Sunday, August 4, the spectators were amazed: in the final of the supreme discipline of sprinting in the Stade de France, Noah Lyles and Kishane Thompson crossed the finish line at the same time after 9.79 seconds. Nevertheless, only the US American Lyles was able to celebrate Olympic gold. At first glance, it seems unfair that, unlike in other sports, not both athletes were awarded the gold medal. If you look at the finish photo, you might even think that Fred Kerley, who ultimately came third, had won the sprint because his foot crossed the finish line first. But the head and limbs don't count in running competitions, only the torso. And there Lyles was 0.005 seconds ahead of Thompson (9.789 s) and Fred Kerley (9.81 s) after analyzing the finish photo with 9.784 seconds.

But how far do the sprinters actually move in 5 milliseconds? According to the running analysis, Noah Lyles maintained his top speed of 43.6 km/h (equivalent to 12.1 m/s) after 65 meters right to the finish. That means he covered 6 centimeters in that time – a difference that should be clearly visible at the finish line. But when does the camera have to trigger for the finish photo? How does it recognize the torso instead of the head, arms or legs? Since the 1932 Olympic Games, the Swiss company Omega has been responsible for the appropriate timekeeping and recording technology, although its 100-year (!) contract is about to expire.

Where the measurement accuracy comes from

The new Scan'O'Vision Ultimate finish line camera from Omega, which was introduced for the Olympic Games in Paris, not only takes a photo, but also up to 40,000 images/s (time resolution 25 microseconds) "in 4K". It is the successor to the Scan'O'Vision Myria, which 'only' achieves 10,000 images/s (100 microseconds). However, the "4K" in the Scan'O'Vision Ultimate is to be understood differently than expected, as it is a slit camera without a shutter that captures images with a width of just one pixel.

Although we have not yet been able to find any detailed technical data for the new camera, based on the previous model, which records 1 × 2048 pixels, the Ultimate should have a resolution of 1 × 4096 pixels. Incidentally, the Ultimate camera would not have been needed for the 5 milliseconds (5000 microseconds); the Myria would have sufficed. To fully exploit the possibilities of the Scan'O'Vision Ultimate, however, the lead would have had to be in the order of one millimeter. But then the discussion would certainly have been even bigger ...

Three cameras are used for the finish line shots at the Olympic Games – a main camera, a backup camera and an infield camera –, which look at the track from different angles. The different perspectives are needed to clearly capture all the runners in case they obscure each other at the finish line. According to the International Olympic Committee, the footage from the main camera is analyzed by the main timekeeper, main referee and team leader, while another official evaluates the perspectives from the infield and backup cameras.

Deformed athletes

If you look closely at the official finish photo, you will notice that the runners on it look strangely distorted. This is because the vertical strip that makes up the image is one pixel wide and only shows the finish line. This means that the horizontal axis of the "photo" does not represent the spatial position of the runners, but the chronological order in which the athletes cross the finish line. This recording technique makes the moving body parts appear distorted. Arms and legs are stretched or compressed depending on the speed of movement. Instead of hanging a finish line tape between two posts as in the past, the finish line posts are equipped with photocells that trigger the cameras as a virtual finish line tape. Omega illustrated the entire functional principle for the 2016 Olympic Games in the following video:

Athletics fans will wonder what the small "sparkling" column next to the finish post on the inside track is all about when watching slow-motion footage of the finish line. The flickering is due to the fact that LEDs in columns display a pattern that creates the "perimeter advertising" in the composite finish photo, which in reality does not exist and is made up of the changing Olympic and Omega logos at the top of the picture.

Finishing photos since 1948

Although the resolution of the finish line cameras has steadily increased over the years, the technology itself is not new. It was first used for finish line photos during the 1948 Summer Olympics: "Magic Eye" photocells from Omega on the one hand and a slit camera developed by the British Race Finish Recording Company on the other. While the photocell recorded the exact time at which the finish line was crossed, thus replacing the finish line tapes, the slit camera enabled the referees to determine the order of finish without any doubt. Then as now, the finish photo had to be consulted in the men's 100-meter final to determine the winner.

Discrimination at the starting gun?

Some critics consider the measurement effort to be excessive, as the sprinters would already be at a disadvantage at the start. They argue that the measured difference of 0.005 s between first and second place is irrelevant, as the sprinters could perceive the starting shot on their track at different times. After all, the sound only travels at 330 m/s. Thus, there would actually be a total time difference of 0.025 s (7 × 1.22 m / 330 m/s) between the – lanes 2 to 9, each with a width of 1.22 m, occupied in the 100 m final.

Between Thompson on lane 3 and Leyles on lane 7 there would therefore be a time difference of 3 × 1.22 m / 330 m/s = 0.011 seconds –, i.e. almost exactly twice the 0.005 second difference at the finish. However, these considerations are pointless, as since the introduction of the electronic "starting pistol" no shot has been fired at all. Instead, the timing is triggered electronically and the start signal sounds simultaneously from several loudspeakers behind the starting blocks, so that athletes on different lanes are not disadvantaged. Much more important is the reaction time of the athletes: Noah Lyles did not leave his starting block until 0.178 seconds after the starting gun.

Incidentally, when Crowdstrike paralyzed a number of Windows systems shortly before the Olympics, Omega's Olympic timepieces were fortunately not affected, otherwise the competitions might have been delayed.

(vza)