Over the last century, the constant quest to improve the mechanical properties of ceramics has ultimately led to the appearance and production of transparent ceramics. Indeed, microstructural refinement and reduction of the residual porosity after sintering — required to improve the mechanical reliability and strength of ceramics — are the key parameters for the production of transparent polycrystalline alumina (PCA). Fabrication of transparent PCA has attracted increasing interest because of the possibility of producing near net shape sintered ceramic pieces. This is particularly interesting for non-planar shape final ceramic pieces, where machining from a single crystal bulk material is fastidious, results in significant material loss and is very costly. Furthermore, use of transparent PCA potentially allows for production of a much larger palette of colored ceramic pieces than is possible for their mono-crystalline homologues because of the existence of a large number of dopant accommodation sites within the grain boundary area. In this work, a complete fabrication cycle from the starting powder choice to the sintering profile is studied in detail to establish the key parameters for industrially viable transparent PCA production. It is shown that the use of a novel fast sintering technique — namely pulsed electric current sintering (PECS), also known as spark plasma sintering (SPS) — requires doping to avoid excessive grain growth despite extremely short sintering cycles (∼15 minutes instead of 10+ hours). The effects of dopants on powder processing and sintering were therefore investigated in detail. A full 3D pore reconstruction was obtained by FIB nano-tomography, to provide, for the first time, accurate pore size distributions and pore volumes in highly dense ceramics (> 99.9 %) produced by PECS. A detailed residual porosity analysis using this FIB data unambiguously disproves an optical model generally accepted in the field used to describe the optical performance of transparent birefringent ceramic materials. The characteristic size parameters used to describe the grain and pore scattering cross-sections in this model are shown to be incorrect. The uncritical acceptance of this model within the community has hindered progress by erroneously placing emphasis on reducing the remaining 0.05 vol.% residual porosity. In fact, it is demonstrated experimentally that pores with diameters below 50 nm only marginally affect the real in-line transmittance (RIT) — the metric characterizing the optical performance — of the final ceramic piece. Indeed, further RIT improvements rely on texture creation and/or microstructural refinement to reduce the effect of grain scattering, implicit for birefringent materials, with grain size diameters ideally below 300 nm. Investigating filter pressing and slip-casting in a strong magnetic field (8 Tesla) to create texture, it is shown that the most accessible means of improving the RIT is through microstructural refinement using smaller starting powders. After definition of specifically required powder characteristics and proper scaling of key processing parameters for use of finer starting powders, RIT values of 64.7 % (at 640 nm and a sample thickness of 0.8 mm) could be achieved. This is the best result ever reported combining simple powder processing routes with PECS under moderate pressure. Using stereological analysis of EBSD (electron backscatter diffraction) data to determine the 5D grain boundary characters, new ways have been opened to link experimental observations to atomistic models and thereby move from an empirical to a predictive use of dopants in the ceramic industry.