Boundary conditions for Numerical Integration

[vskoupy]

When the precision of boundary conditions from MST is low, the numerical integration method fails or returns wrong value. This is true especially for higher l, m, omega (l,m=25, omega=6.).

[znasipak]

If this is only occurring for the R["Up"] in the s = -2 case and R["In"] in the s = +2 case, then this may have to do with the way the Toolkit is integrating the Teukolsky equation. Numerical error in the boundary conditions excites the other independent homogeneous solution that becomes dominate as you move away from the horizon or infinity