Ulugh Beg’s 15th c. Samarqand observatory and madrasa is one of the most famous Islamic scientific institutions, producing astronomical observations that were not equalled until Tycho Brahe. Less is known about the processes of research and education at Samarqand, but a number of commentaries produced by Samarqand scholars promise to shed light on the intellectual life of the classroom. Instruction in the sciences was primarily an aural process, in which students listened as a text was read aloud, with frequent breaks for the instructor to comment and elaborate. However, as anyone who has perused the Almagest can attest, it is quite difficult to visualize the complex and dynamic motions produced by the mathematical machinery of eccentrics and epicycles without the aid of images. Historians of the mathematical sciences overwhelmingly focus on the text (geometric proofs, numerical parameters) and have scarcely paid attention to the role of manuscript images in transmitting knowledge, typically considering geometric diagrams as ancillary to the proofs. But astronomical manuscripts contain other images, not related to proofs, that convey meaning to the reader as well, and in many cases are more helpful than geometric diagrams in illustrating how a model functions or a phenomenon occurs. In this paper, I explore the ways in which manuscript diagrams and illustrations worked in conjunction with the text in astronomical commentaries written by the director of the Samarqand observatory Qadizade al-Rumi, and one of his prominent students Fathallah al-Shirwani. Using examples of how these commentators portrayed the components and motions of complex models, as well as more basic celestial phenomena, I describe how students could begin to put together an impression of how they combined to produce the apparently irregular motions we observe in the heavens, as well as the outstanding problems of Ptolemaic planetary models that these astronomers were attempting to explain and resolve. I analyze the extent to which images may have helped the reader visualize different kinds of motion and extrapolate from two-dimensional static “mathematical” images to the three-dimensional dynamic “physical” motions of the heavens. I suggest that differences in the kinds of images each commentator chose to add to their texts, the manner in which they portrayed the models and particular motions, and the ways in which they highlighted and analyzed different problems reflect differences in the intended audiences for the two commentaries.