Thank you so much for your detailed reply. Just a follow-up:

So, for a typical task based experiment (block or event design with multiple conditions/tasks, not involving resting state), you would recommend using a HPF (0.01 Inf)?
I have seen some Conn papers use (0.008 Inf) and also use (0.008 0.09 Hz) for task based experiments while quite a few of them have not overtly mentioned the values used for denoising. Could you point me towards any sources that I could refer to gain a better understanding of the issue at hand (and potentially cite)?

Further, there are some studies that have used a frequency decomposition of time series and have found meaningful results in particular frequency bands. I am assuming that the time-frequency decomposition in the Setup would do something similar? In your opinion, is it worthwhile to investigate particular bands instead of the entire time series?

Finally, for the part about using different HPFs, based on your reply, I guess the best option would be to implement the 0.01 to Inf for all the subjects?


Thank you once again.



Warm Regards
Pravesh