Comment by mrrrcs

One run per cell is more noise than the gap you are measuring