Currently at 1khz contact checks, both biosemi and actichamp have a gain very close to 1 ~ 0.99 so we have ignored this, but this is not always the case.

This meganerd https://github.com/Teuniz/EDFlib who made EDFBrowser has functions to write bdfs, so it should be possible to make some fake ones we know the exact values for to check the processing.

Memory estimate is based on approximate size of channel, but with files with long gaps at start and end this is overestimate. Change to only amount which would be loaded in reality

User should have to get HDR, Trigger and TT everytime, so make a function to choose it with uiget

We want to test Load_data using either more simulated data or real data e.g. Action Potentials/EEG.

Stuff is on Anna's PC at the moment...

You get difficult to decipher error messages when you load an empty file, or give sopen the wrong parameters. so should find a way of validation the HDR before doing things

in case trigger codes at start are missing, such as from a cropped file

Currently wont work with new system, so need to update this!

Also, wont work with ActiChamp, code can be refactored to be agnostic to which system is used

I thought it did this already

Biosig has a tendency to overwrite some default matlab stuff. Do not want

It would be nice to have, and it happened before with Tom that you loose somehow one trigger and the entire pattern becomes wrong from that frame.

Please use forward slash (/) to separate directories. Matlab can nicely handle ocnverting / -> \ on windows, but the contrary does not work.