Hi Fidan,

The .01-.08Hz filter covers approximately the range between 10s and 100s and it is relatively standard for resting-state connectivity analyses. Of course with the growing availability of faster acquisitions there is also an interest in exploring higher frequencies (above .10Hz, below 10s) but that is basically limited by your TR (you can not go higher than half your sampling rate, i.e. can not go lower than twice the TR). The low-pass portion of the filter (.01Hz, 100s) is mostly governed by the increasing contribution of physiological and 1/f noise components to the very-low frequency segment of the BOLD signal, so I have not seen much interest in pushing beyond 100s in the literature. 

Regarding the raw signal, the CONN gui allows you to visualize the effect of the individual Denoising steps in real-time. In the Denoising step you may simply change any of the parameters there (e.g. change the band-pass filter, add/remove noise regressors, etc.) and observe their effect on the histogram of voxel-to-voxel connectivity values. For additional analyses, you may use the ROI BOLD signals in conn*/data/ROI_Subject#_Session#.mat (before Denoising) and conn*/results/preprocessing/ROI_Subject#_Condition000.mat (after Denoising), or similarly the voxel-level BOLD signals in conn*/data/DATA_Subject#_Session#.matc (before Denoising) and conn*/results/preprocessing/Data_Subject#_Condition000.matc (after Denoising).

Hope this helps
Alfonso

Originally posted by Fidan OZCAN: