Laser-based Powder Bed Fusion of Metals (PBF-LB/M) faces critical challenges such as spatter formation, porosity, and melt pool instability. Beam shaping, which directly alters the melt pool behavior, has emerged as a promising strategy to stabilize melt pool dynamics and suppress defects [1]. Physics-based simulations can elucidate these thermal impacts, yet their predictive accuracy depends on experimental validation. Currently validation approaches rely on melt pool geometry comparisons, but such methods lack thermal boundary constraints due to the lack of temperature data. Therefore, there is a need for accurate melt pool temperature measurement. However, conventional pyrometry cannot resolve absolute temperatures in molten metals because their emissivity evolves unpredictably during processing [2]. To address this gap, at the Professorship of Laser-based Additive Manufacturing, we established a coaxial multispectral-imaging (MSI) system that simultaneously resolves temperature and emissivity in situ [3, 4]. The MSI system is validated according to melting point and vapor formation, with a high accuracy over 98.4% [3]. Experiments on stainless steel 316L plates reveal four zones in the melt pool, namely (1) the laser–metal interaction zone, (2) the melt pool zone, (3) the vapor-smoke zone, and (4) the spatter trajectories. The absolute temperature maps provide physics-based inputs for multiphase CFD simulations. The high-fidelity data, with quantified measurement uncertainties, directly constrains melt pool models, vapor plume, and spatter trajectory. The high-accuracy dataset from the coaxial MSI system provides a quantitative foundation for simulations to understand melt pool mechanisms in PBF-LB/M.