Despite significant progress, many of the physical processes that govern how galaxies form and evolve remain poorly constrained. Discriminating between competing models therefore requires direct comparison between galaxy formation simulations—often representing entire synthetic Universes—and multi-wavelength observations.Yet simulations do not ab initio predict what we actually observe on the sky: galaxy positions, luminosities, morphologies, or spectral energy distributions.

Bridging this gap between theory and observation is a central challenge in modern astrophysics. One powerful approach is the creation of synthetic observations from simulations, enabling like-for-like comparisons with real data.Over the past three years, we have developed synthesizer, a fast, flexible, open-source Python package—accelerated with C and extensively documented—designed to allow theorists to rapidly and reproducibly generate synthetic observations from simulations across a wide range of wavelengths and facilities.In this talk, I will outline the problem synthesizer is designed to address, describe its guiding principles and capabilities, and conclude with a live demonstration of its use in practice.