I added analogical 'if' statement to the one that is in the current version of plotmitrace.m. It now outputs a reasonable decorrelation plot on my machine with the 'Statistics...' toolbox
...the same problem occurs in estimate_gamma_correction and the fitted curve is the same in all cases, this looks like a good candidate for making the fitting a separate function
I added analogical 'if' statement to the one that is in the current version of plotmitrace.m. It now outputs a reasonable decorrelation plot on my machine with the 'Statistics...' toolbox