Skip to main content

I love my Lynx system, and every time I go to another meet where they use something like FlashTiming or EagleEye I feel sad like I can expect some inaccurate places and times.

It seems to me that with a camera that's only capable of capturing an image every 1/60 of a second, there should be some times that are impossible to achieve. For instance, in a 100m race, a time of 11.60 is 696 frames at 1/60 second ((1/60)*696=11.60). However, if you add 1/60 to 11.60 that becomes 11.6166667 which should round up to 11.62, correct?

If I'm incorrect and times aren't rounded, then the example above where 697 frames is 11.6166667 is truncated to 11.61, adding another 1/60 makes that time 698 frames at 1/60 or (1/60)*698=11.6333333 which is truncated to 11.63. What happened to 11.62? Or, if we're rounding and that becomes 11.64, what happened to 11.63?

I don't have a 60fps full-frame system at my disposal to test this. I would be interested to see someone give it a try and see if I'm right. I know that 120fps is a different story and it should be able to give precise enough times. However, in close races I'll take my line scan camera any day.

I thought this knowledge could be useful when look at posted times and if I saw someone with an 11.63 for example I would know that the times were gathered using a 120fps or better system. I'm not sure how it would work with multiple races though. Like if the start signal was received halfway through a frame in one heat, but right at the start of a frame for another heat, does the software take that into account and add 0.0085 to all the frames in the first heat? I'm just curious how those systems work as I've only ever used a Lynx system.
Original Post

Replies sorted oldest to newest

The point that you're getting at is valid -- a system that samples every 1/60 second is accurate to only 1/60 second, not 1/100 sec. It's convenient to publish times in decimal format, which gives the impression that it's accurate to 1/100 sec. Yes, some decimal representations will not occur, but that doesn't really change anything.

Perhaps it would be better to publish times as 11 37/60 sec, 11 38/60 sec, 11 40/60 sec, etc.
We used an Imhof video timing system starting in the early 90th.
This gave us an accuracy of 1/50 seconds.
But only with a vcr, which was able to show half-frames.
On this machines, switching from half-frame to half frame changes the displayed time by 0.02 seconds.

The time for a runner was kept, if he was over the inserted finish line - so it could be 11,62 or 11,64, but never 11,63.

Ciao

Dirk

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×