sean9keenan wrote:even a normal video camera would allow for MUCH more consistent measurements
Past experience with using video camera in an event named “Cow-A-Bungee” led to less than satisfactory results.
As I recall, Cow-A-Bubgee was a Physics Committee event, so I was not involved in many of the discussions. Not sure how accurate my memories are, but here's what comes to mind.
Video cameras were used for measuring how close the egg came to the floor. There were several problems inherent in their use.
Many people greatly overestimated event supervisor’s abilities and resources. Many supervisors didn't have knowledge and/or equipment suitable for the task.
Video frame rates allow the egg to drop quite some distance between frames. This led to mistrust of the measurement’s accuracy.
The images of the egg in motion were quite blurred and it was suggested in the rules that a camera with "high speed" function be used. High speed function only decreases "exposure" time, producing sharper images, but does nothing to increase the frame rate. This means the egg can still travel quite some distance between frames.
A major problem was locating the frame where the egg was at its lowest point. Only top-end VCRs had "Jog & Shuttle" or single frame capability. Most supervisors didn’t have a VCR with these functions and the time it took to locate the correct frame was so great, it was nearly impossible to score the event within the tournament time span.
The perceived distance of the egg from the floor is dependent on the geometry of the set-up and many questions were raised as to its accuracy. Eventually someone published a method for making accurate measurements, but its explanation was technical and many key people didn't understand it. By then fairly strong opposition to the event had become entrenched within Science Olympiad's ranks and the event's demise was unavoidable.
I realize most of these problems aren't applicable to the current discussion as today’s objective is to measure time instead of distance. This simplifies the measurement process immensely and brings the time base of the camera and player into question.
Video equipment has progressed from difficult to control/inaccurate mechanical components to totally digital devices whose time accuracy depends solely on its master clock This has increased accuracy of the recording/reproduction process to where time error is negligible for all intents and purposes.
As with Cow-A-Bungee, it could take considerable time to locate the frames necessary to determine the elapsed time from the video. While any given event supervisor/tournament may have the ability and equipment needed, many event supervisors across the country will not. Requiring them to use video could cause serious problems with the event.
Imprinting a visible time stamp on each frame of the video would simplify the measurement process. I'm not aware of any reasonably priced equipment with this capability beyond 1 second resolution. Most video editing software shows individual frame numbers, but it would be awkward and time-consuming to use for many venues.
One suggested solution might be to include a digital timer in the field of view. I'm not sure how well this would work, but it's worth a try. If the frame with the start of the device and the frame with task completion can be located, it should be a simple matter to subtract the start time from the finish time to get the elapsed time.
This leaves the human interface between the camera and actions being recorded. Without including elaborate interface requirements in the rules, how can the camera be sure to capture the actions necessary for an accurate determination of elapsed time.
If the camera’s position is fixed, some actions may be out of the frame (there is no gaurantee the vehicle will remain in a fixed camera’s field of view. If the view is too wide, it may not be possible to determine exactly when actions occur. If the field of view is too narrow, visual cues to approaching actions will not be available. Panning and zooming may also cause some events to be missed altogether. How would parallax error be compensated for?
Much of the feasibility depends on the current event rules, but much more depend on each individual device. It would be necessary for judges to determine a different best course of action on an individual team basis. I see using video cameras as a possibility for individual tournaments, but it’s probably not an option that should be required.
It’s unlikely Science Olympiad would adopt a requirement to use video cameras until/unless data were available to prove their feasability. This means someone would have to use them in a tournament, identify problems with their use and supply solutions to include procedures on their set-up, positioning, lighting, and how to extract elapsed time information from the recorded video. The need is to be able to counter any/all objections that might be raised.
