Back to Main
Contents Page
Variable Definitions
Threshold Event
Search Between Events
Threshold Algorithm
This dialogue is used to define one threshold event type. More than one such event may be detected within a trial.
To find events at different threshold values, you should create a threshold event type for each value.
These settings allow you to control how the Threshold detection algorithm is used to detect threshold events.
This defines the variable in the trial for which you want the find the thresholds.
You can define either a trajectory variable, or an analogue variable in this box. You can select a suggested name from the drop-down list (see the Preference Page to find out where these come from) or simply type the name of the variable you want.
For a trajectory variable, you must indicate the Component of the variable that should be considered. This setting has no effect for analogue variables.
Enter the threshold value at which you would like the event to be defined. Note that you should bear in mind the units of the variable when selecting this value.
The Window value allows you to define a window around the Threshold value, which can be used to take into account the effects of noisy data. The window defines two thresholds, one above and one below the required threshold value.
Note that you should bear in mind the units of the variable when selecting this value. Clean smooth data should have a low value (zero may be appropriate in many cases). For noisy data, a value close to the peak to peak value of the noise should be used.
See the Threshold Algorithm for more details about how this window is used.
The units used by the threshold detection system are fixed by the Vicon PlugIn interface, according to the variable type (POINT FORCE MOMENT etc). You can choose the appropriate units here, which will help you make the correct choice for the Threshold and Window values.
An event is only defined when the variable passes through the threshold in a positive sense.
An event is only defined when the variable passes through the threshold in a negative sense.
Defines an event whenever the variable passes through the threshold, in either sense.
Supply the name of the Event you want to be defined. You can access the event later in the Event Analyser using the same name.
This determines which context bar the event will appear on in the Workstation time bar.
These option determine how events are named when more than one event is found during a single search period (see Search between Events).
uses the same name for each event which is found during the search period
This option causes an incrementing number to be added to each event to found with any given search period. When using the Event Analyser to analyse these events, you must manually supply the event number for the actually event you want to analyse.
Where the Search Between Events option is used, and more than one search period is found, the incrementing number is reset to 1 again at the start of each search period.
This option only searches for the first event in any given search period.
These options control the time periods in the trial during which the threshold detection operates. The time periods can affect the way the events are named, depending on the option selected in For Multiple Events.
To define the whole trial as a single time period, leave the tick box empty.
To Search between pairs of events, tick the box, and define the Begin and End event names and their contexts. Periods will then be defined within the trial which span these two events. You can use the same name and context to define consecutive periods.
The threshold algorithm actually uses two thresholds as defined by the Threshold and Window values. The upper threshold is defind as the threshold plus half the window value; the lower threshold by the threshold minus half the window.
The threshold variable value must pass through both the upper and the lower thresholds, before a threshold event is generated. If you have selected the Up or Down crossing direction, these threshold must be passed in the correct order to qualify.
When this occurs, the data which is found between the upper and lower thresholds are fitted to a straight line, and the time at which this straight line passes through the required threshold value is used to define the event.
Note that since a straight line is fitted to the data, you will lose accuracy in the timing of events if you set the Window value too high, since the underlying curve of the variable data may not be linear. However for very noisy data, a straight line approximation is reasonable.