When the discontinuous structure of matter was yet an intriguing hypothesis, Jean Perrin performed a set of elegant and pioneering experiments that marked the birth of what today we consider quantitative optical microscopy. Picking up the baton from Perrin, today microscopists face incredible challenges, aiming to extract quantitative information from the increasingly content-rich and complex images made available by modern microscopy techniques. Here, I provide an overview of these challenges and describe the solutions adopted to succeed in this complex task when investigating colloidal systems or systems in which colloidal particles are embedded as microrheological probes.