Reply to thread

Thanks, changed all of that. 1-3 actually seems like a good option, everything above that is prone to generate noise over the data.