I added a function to calculate the semblance of a stream, but the problem is semblance is calculated on a multi-channel timeseries, not a "stack" of individual samples like the sum and product stacking we currently have. This makes me think we should add in the "window-based" stacking in addition to the "sample-by-sample" stacking currently. I think the ability to evaluate the stack based on some measure of waveform similarity is important. Thoughts?
I added a function to calculate the semblance of a stream, but the problem is semblance is calculated on a multi-channel timeseries, not a "stack" of individual samples like the sum and product stacking we currently have. This makes me think we should add in the "window-based" stacking in addition to the "sample-by-sample" stacking currently. I think the ability to evaluate the stack based on some measure of waveform similarity is important. Thoughts?