A novel variation of the internal tin method, which improves upon the conventional approach is described. It offers promise as a way of producing reliable and economical Nb3Sn for high energy physics, fusion and other applications. Preliminary results show several important advantages over the conventional internal tin and bronze approaches. These include improved filament arrays and low distortion which allow the introduction of high percentages of niobium alloy filaments without the occurrence of "bridging". As a result the material is expected to exhibit high J(c)'s with low losses. In addition the new process is less complicated than the conventional approach and produces a reliable and economical product in relatively long piece lengths.