Hi again all,

I am applying the routine "Voltra":

voltra( t0, h, g, kappa, t, f );

When computing the kernel "kappa", I make a call to the NR-routine "cisi". As the final answer "f" for the Voltra-routine is very small, I need more precision than "double" can offer. I guess I need "long double" or something similar.

Is it possible to increase the precision of the return value of "cisi" and "voltra" by applying "long double"?

Greateful for any help!

Best regards, perr