> And it won't show with double precision floats, right? Does Opcode do double precision now? Last I looked (which was a long while back), it only did float, and "double" support was just a shim that converted the data to float. If it does real doubles, then you have significantly more precsision before running out. Cheers, / h+