Summary
Which “direction” is the StartTime metadata intended to go in BIDS psychophys files, when there is no corresponding neural dataset?
Clarifying diagram
Hopefully this clarifies slightly what I’m asking.
We have a task that gets run both in fMRI with psychophys, and in a plain old out-of-scanner psychophys environment.
When we have fMRI data, it seems clear that the StartTime will be negative to capture the fact that the recording of psychophys started before the scanner.
But when we don’t have MRI, it feels slightly odd that the StartTime will still be negative, and I’m not sure the spec answers this question. Should it be 0 in this case, because the spec says “the first data sample in the corresponding (neural) dataset” and the psychophys data is the dataset in this case?
Positive feels wrong, but I’ve seen datasets that do have positive offsets for this value, and I can put myself in a headspace where I think “the dataset starts recording at 0 but the task starts at 5s in, therefore StartTime should be 5.0”.
Additional question
I’m having a hard time even finding examples of BIDS with psychophys sidecars where StartTime is negative. Including psychophys at all seems rare. All the sample datasets in the BIDS repo appear to just set StartTime to 0. Is it more “standard” to just trim the data to that 0 starting point and say StartTime is 0? Because that’s simple enough…
Appreciate any insight into my confusion on this!