I’m taking a look at the effects of stimulus delay settings in one of my datasets so I was surprised to notice that the default setting of mkanalysis-sess is a delay of -TR/2 seconds. What is the reasoning for this? If, as I expect, it was just empirically found that it worked better than without, my question is whether this decision was made on the basis of all the event-related options or just some subset of them? I’m worried that this might be an optimization that isn’t appropriate for my particular parameter choices (FS 5.3, spmhrf0). Any comments would be appreciated.