A summary of statistics relating to the multiple scales defined in the model (see Setting Scale) and/or Check points. The data shown here is a summary of multiple scale and check point fit data (see also the Check Distances Table and Check Points Table) where deltas are shown as a percentage, as an absolute value, and as signed value.
Multiple scales help spread out any measurement error and allow for cross-checking and verification (as opposed to a single scale where there is no beneficial redundancy). Check points can also be used to compare the known imported positions with the solved position in PhotoModeler.
However, as multiple scales are applied to the model or points imported, the resulting coordinate transformation will be the best fit to all of these scale distances or points. In a perfect world, perfectly measured scales, and a perfectly built 3D model, will fit together perfectly. In reality however, best fit will result in some ‘delta’, where one or more of the distances or point positions in PhotoModeler doesn’t match exactly with the associated measured scale distance or position.
So a delta can be due to scale or position measurement error and/or model weakness. When deltas are large, then both should be verified. If scale measurements are good and imported points are accurate, then there is likely some model weakness which can come from miscoded targets, misreferenced points, inaccurate point marking, weak camera angles, etc., and once the model is solving well, deltas will get smaller.