Mirror-image inversion in commonly used compound microscopes.

Optical microscopes are essential tools for magnifying and analysing a wide range of samples and are generally expected to faithfully preserve the original object in all its dimensions. However, depending on the optical design and software pipeline, many commonly used systems silently generate mirror-reversed images. This transformation alters handedness and therefore the three-dimensional (3D) structure of the object, unlike the simple 180° image rotation produced by lenses. We show that mirror inversion arises whenever the emission or detection path contains an odd number of reflective elements, and that ocular and detector paths might not be concordant. Although essentially undetectable on most specimens, this effect crucially affects the interpretation of left-right (LR) asymmetric, polarised or chiral structures. Using LR-asymmetric H-neurons in Drosophila, we illustrate how the same specimen imaged on different systems can appear anatomically inverted. Because post-acquisition operations (e.g. flips in Fiji software) or camera readout processes can introduce additional mirror transformations, the risk of error is cumulative. We provide practical guidance for users, imaging facilities and manufacturers, and advocate for explicit orientation reporting of published images. Orientation should be treated as a genuine metadata element that is essential for reproducibility, analysis and data reuse.