Ricardo Bruña

Hello, Jan-Mathijs, Indeed, as you point out the chol implementation uses the (or, better yet, one) square root of the R matrix, so it would make sense that the value...

In Eq. (6) in Lin et al. lambda is inversely related to SNR2, so that part should be right, according to this paper. However, in the MNE documentation (https://mne.tools/0.17/manual/source_localization/inverse.html#whitening-and-scaling), lambda2...

It makes sense to use sigma (standard deviation) squared for the sensor covariance: it is equivalent to a diagonal matrix of variances. However, for the regularization parameter the problem is...

It is, indeed, a good approach. The issue is that in beamformer the regularization parameter (for example in Gross & Ioannides 1999 and Gross et al. 2001, as van Veen...

That would require a good amount of documentation (and reading by the user), and maybe a warning. I think this is one of those cases where there is no "good...

Hello, Robert, Regarding bit24, it is at least as old as Matlab 2016a, as I just checked it. Regarding the block-read, indeed, it makes more sense reading a segment than...

In my case, the sampling rate is 2000 Hz, so the record length would be "0.0005 " (with two spaces). But I can see the issue with power-of-two sampling rates......