[Ishikawa, Yuzurihara]

We found it's possible to comtaminate the artificial signal on Q-transform. We updated the code for Pastavi.

Details

While working on the Q-transform plots, we could see some signals on the data at the time of  (starting time + ~1) and (ending time - ~1). Although we changed the data length, they appeared around both edges. When we turned off the whiten option on Q-transform, they disappeared from the plot. Accordingly, they are artificial signals. 
To take care of the artificial signals, we added the margin (4s) when we read the data. After performing the Q-transform, we croped the margin from both edges. (This process is similar to the process of the application of a time-domain filter.) The bottom and top of Figure shows the result without and with adding margin, respectively.