Spatial smoothing of time-frequency plots

Hello,

I have been trying to apply spatial smoothing on source reconstructed MEG whole-brain time-frequency plots and it worked as intended. However, I would like to first average some time blocks before applying the smoothing (e.g. -0.5-0s; 0-2s, 2-5s, 5-8s, 8-11s, 11-13.5s), so I applied the "group in time or frequency bands" process and selected the desired time blocks. However, when I apply spatial smoothing to this, it seems the time domain collapses entirely and I don't understand why.

Before smoothing:
Time: [-500.0, 13,500.0]ms

After smoothing:
Time: [-500.0, -479.2]ms

I was wondering if you know why this happens and/or if you have a better solution to perform this averaging across time blocks without using "group in time and frequency bands" process. I don't like this process very much since I still have to navigate through each sampled time point (e.g. 0.000 > 0.042 > 0.083...) instead of just having one time sample per averaged block (one for -0.5-0s, one for 0-2s, etc.).

However, when I apply spatial smoothing to this, it seems the time domain collapses entirely and I don't understand why.

The time-frequency files with time bands are not supported by many processes, with no clear error messages... They don't seem to be used much, therefore we don't put a lot of efforts in maintaining them.

I added some extra tests so that they work with some of the "Filter" processes, such as the spatial smoothing: https://github.com/brainstorm-tools/brainstorm3/commit/51cdbb9fd441b52017ea8fb6dcae45b548d72763
Update Brainstorm to get this fix.

I don't like this process very much since I still have to navigate through each sampled time point (e.g. 0.000 > 0.042 > 0.083...) instead of just having one time sample per averaged block (one for -0.5-0s, one for 0-2s, etc.).

This is to keep them compatible with the display of the file from which they are computed.
If you want one step per value, you can edit the structure manually: delete the field TimeBands and replace the field Time with a time vector of your choice (increase between 2 time points must be constant).

Thank you! I updated brainstorm and it works as intended. Also, thanks for the tip to change the structure manually. However, it would be great if this was possible to be done automatically from the interface (for example by adding a checkbox when you compute the grouping in which you can choose to either have it as it is now, or to make only one sample point for each group).