It is important to get calibration loads are different flow rates which are representative of all harvesting conditions.  It is important to keep these different flow rates segregated into seperate loads, so that the monitor has something to work with throughout the spectrum of flow rates.  It is important not to dilute the representative cal loads by inputting a large number of loads at high flow rates.

I have received considerable fire here for stating that a better job of calibration can be done through use of proper calibration proceedures with an inaccurate method of determining weight (ie, cut exactly 'till the buzzer goes off every time and assign the same weight) then can be done by weighing semi loads and entering everything in the monitor without paying any attention to flow rates when cutting the calibration loads.     I don't care about drawing fire for stating it, because it is true.  Anyone who claims otherwise is not considering how the yield monitor works. 

Later changes to calibration will apply to all data still in the monitor.  They will not change date which has already been transfered to the computer, unless you make special efforts to apply the new calibration to the old data through file modifications or by reloading the files to the monitor and allowing the new calibration to be applied to them.  

My strategy is to attempt to get the combines calibrated accurately, and more importantly, get all of them to agree among themselves, at the beginning of the season and then leave it alone.  I am using the monitor to harvest representative data across the field, absolute accuracy of the totals from the yield monitor is not that big of a deal....if I want to know the exact yield of field ZZ I will simply consult the scale ledger and calculate it.  The yield monitor is primarily intended to quantify yield differnces across the field and I operate them accordingly.