Discussion of imalsogreg's comment: Spike rate method smoothed some events that looked (from rasters) like they should have been sharp and flat - how much "ramping" is really just, trial-to-trial jitter, smoothing, or cell-to-cell jitter (in summary plots)?