As per the discussion in #39, the original stimulus processing code introduced a spurious one-frame offset into the timing of stimulus presentation data. This PR removes that offset. It also adds a command line parameter allowing users to add a by-hand offset in timing to better bring stimulus timestamps in line with what the mouse sees while in the rig. Based on experimentation with real data, Corbett Bennett has determined that an 8 millisecond offset between the timestamp recorded in the sync file and the timestamp recorded in the stimulus table is appropriate.
As per the discussion in #39, the original stimulus processing code introduced a spurious one-frame offset into the timing of stimulus presentation data. This PR removes that offset. It also adds a command line parameter allowing users to add a by-hand offset in timing to better bring stimulus timestamps in line with what the mouse sees while in the rig. Based on experimentation with real data, Corbett Bennett has determined that an 8 millisecond offset between the timestamp recorded in the sync file and the timestamp recorded in the stimulus table is appropriate.