I am using Origin 2015 (64bit) Peak Analyzer FitPeaks(Pro) to fit a spectroscopic curve. See attached image.
The second peak from the left (spanning ca. 285 - 290 on the x-axis) is my own modified rectangular function with the start- and end-tails from erf and erfc functions respectively. Though the cumulative peak (red curve) fits well, the end-tail of the sub-peak (green curve) shown is linear and drops. This is evident from the number of dots plotted there - the x-interval at the end-tail is much lesser than at the start-tail. This is not just a graphical problem but the output x- and y-data for this sub-peak also shows this unequal x-interval.
When you are in the Fit Peaks(Pro) page of Peak Analyzer, could you Expand the Results branch, then expand the Configure Graph branch, then in the Fitted Curve Plot branch, please try changing the settings in the image below and see whether this makes any difference.
If all these does not work for you. It may because of your user-defined function. Could you send us your project file with source data and peak fitting results, and the FDF file of the user-defined fitting function so that we can further identify the problem.
You can send it to tech@originlab.com and quote the link of this post in the email.