Hi,

using the sensorgram comparison functionality in most cases standard deviation will be used since the standard sensorgrams experimental variation will be evenly distributed. The deviation plot is then constructed by subtracting the average curve from all data and divide by one SD (the average curve itself becomes a straight line at zero) and the limit senorgrams become straight lines at the set SD value. The data handling is symmetrical for upper an lower limit.

But if the standard sensorgrams are not randomly distributed (different batches/lots of the standard for instance) Min/Max can be used instead. And then suddenly only one (the lower) of the limit sensorgrams will be straight in the deviation plot, whereas the upper will vary. As the data is no longer symmetrical (a median is used), developing this functionality it was decided to use the lower limit distances to the median for transfer to a deviation plot. So this is used for all data, also data on the "up side". The lower limit will then be straight whereas the distance from the upper limit sample data points to the median can still differ because of the asymmetry.

I have no idea if this makes any sence at all, but you can always look in the Software handbook to get a more stringent explanation.

Best wishes,

Asa