Built in holdtime in SinglaTarget/WaitThenHold?

Hi Jaewon,

As you know from our previous discussion, I have been reconstructing joystick reaction time from joystick trace, instead of relying to one recorded by the adapters, because of failure if the joystick was moved across the threshold right between two scenes.

But whenever it did not fail, I am comparing the adapter-recorded RT to trace-reconstructed RT, and I noticed occasional discrepancies. It is as if tiny momentary movement of the joystick beyond the threshold was not recorded by the adapter even though WaitThenHold's hold time was 0. As if there was a built-in hold time.

For example, the attached file shows a zoomed-in joystick trace, the threshold was 4, the direct reconstruction from the trace was 205 while the adapter returned 531.8. Of course the adapter's figure makes more sense, I am just asking what is the mechanism that the adapter uses; I would like to replicate it in my code.

It is briefly mentioned in the remarks of SingleTarget.
https://monkeylogic.nimh.nih.gov/docs_RuntimeFunctions.html#SingleTarget

When the threshold window is crossed, SingleTarget marks when the crossing occurred but does not change its Success state immediately. If the cursor stays on the same side stably at least for one frame, then it changes the state and set the marked time as AcquiredTime.

This is to avoid the case that fixation is declared incorrectly due to noise in the signal. It does not apply to the touch input and digital buttons (when used with SingleButton).

OK, so one frame - thank you. I'll add a similar condition to my reconstruction code and then the results should match better!