A technique for combining a measurement that has a high accuracy but low precision with another measurement of the same quantity that has a high precision but low accuracy in order to produce a result that is better than either alone. Alpha processing is used to improve the vertical resolution of neutron porosity and other dual-detector nuclear logs. The detector near the source has better precision than the far detector in the sense that it responds more precisely to vertical changes. However, the near detector is less accurate because it is more affected by the borehole environment. Alpha processing mathematically superimposes the rapid changes of the near detector on the slowly changing but accurate far detector to produce an accurate log with high vertical resolution.

The technique is also used to improve results from the carbon-oxygen log and other pulsed neutron spectroscopy measurements. Two methods are used to determine the carbon/oxygen ratio. The windows method counts the number of gamma rays within energy windows placed at the main peaks for carbon and oxygen. This method has good statistical precision but poor accuracy, as gamma rays from other elements contaminate these windows. The other method, spectral stripping, compares the total spectrum against standards for many elements, inverting the spectrum to obtain the yield for each element. This method is more accurate but has less statistical precision. Averaging over a number of measurements, alpha processing adjusts the windows result with the more accurate spectral stripping in order to obtain a precise and accurate result.