Thank you for creating the CONN toolbox, comes in very handy.

I've used it to analyse a task-based event-related design using gPPI (2 conditions).
However, I would also like to analyse intrinsic FC results from this task-based design. I know this can be done, but don't know how it should be done in CONN. 

I made a condition for my design called 'rest' that is present during every entire session of every subject and performed low-pass filtering (0.008 - 0.09 Hz) and the normal denoising (with 'effects of rest'). I then analysed using a weighted GLM, looking at bivariate correlations.
What I'm wondering is, do I remove my 2 conditions from the design in setup, or do I leave them in?
I thought of leaving them in so that I could account for their presence and also so that they would be included in the denoising step ('effects of ...').
I also thought of deleting them: acting as if there are no conditions present and just perform low-pass filtering and denoising as if there were no conditions/tasks. 

I've been looking at articles to get an idea of what the correct approach is when trying to get intrinsic FC results out of a task-based design, but still am confused. Any advice on the matter?