Context. The first direct detection of gravitational waves in 2015 marked the beginning of a new era for the study of compact objects. Upcoming detectors, such as the Einstein Telescope, are expected to add thousands of binary coalescences to the list. However, from a theoretical perspective, our understanding of compact objects is hindered by many uncertainties, and a comprehensive study of the nature of stellar remnants from core-collapse supernovae is still lacking. Aims. In this work, we investigate the properties of stellar remnants using a homogeneous grid of rotating and non-rotating massive stars at various metallicities. Methods. We simulated the supernova explosion of the evolved progenitors using the HYdrodynamic Ppm Explosion with Radiation diffusION (HYPERION) code, assuming a thermal bomb model calibrated to match the main properties of SN1987A. Results. We find that the heaviest black hole that can form depends on the initial stellar rotation, metallicity, and the assumed criterion for the onset of pulsational pair-instability supernovae. Non-rotating progenitors at [Fe/H] = -3 can form black holes up to ∼87, M⊙, thus falling within the theorized pair-instability mass gap. Conversely, enhanced wind mass loss prevents the formation of BHs more massive than ∼41.6 M⊙ from rotating progenitors. We used our results to study the black hole mass distribution from a population of 106 isolated massive stars following a Kroupa initial mass function. Finally, we provide fitting formulas to compute the mass of compact remnants as a function of stellar progenitor properties. Our up-to-date prescriptions can be easily implemented in rapid population synthesis codes.
