Uncertainty in absolute times often makes it necessary to be able to use the difference in time between two arrivals such as P and S or P and L. If no absolute times are available, the calculated origin time will be close to that at the first arrival station and is of course meaningless. However, a perfectly good epicenter and depth can still be obtained from P-S or P-L differences alone. To enable this feature, set the weight for the P phase input record to 9. This P is then assigned a weight of 0, effectively disabling its use. However, a time residual and azimuth, etc., will still be calculated for it, enabling an assessment to be made of its absolute time. A search will then be made of the entire input phase set for an S or L phase at the same station. If such a phase is found, its variables are used to store the observed and calculated difference times and their derivatives, and it's weight (0-4) is used for the difference phase. DON'T SET IT TO 9!! If two or more such phases (e.g., SN, SG, LG, etc.) are found, all their differences with the P time will be used instead of their absolute times. Blanks will appear beneath 'hrmn' in the residual summary for all such phases, while the observed and calculated difference times with the first P will appear beneath 't-obs' and 't-cal'.
NB. There must be at least one phase with absolute time to get a location.