Age-related macular degeneration (AMD) is the most common cause for acquired blindness in industrialized countries. It mainly affects sensory receptors of the central visual field. Individuals with AMD have difficulty recognizing faces and reading or distinguishing the keys on a keyboard. However, their peripheral vision is not affected, so patients can still recognize colors and shapes in the periphery. With the rise of head-mounted displays (HMD), new possibilities have emerged to create novel visual aids. We present a visual assistant that magnifies a captured video stream and presents the magnified image to the user within a section of the screen. In this work, we examine how different sizes of this magnification section as well as magnification factors may affect the precision, accuracy and speed of inputting words on a keyboard for patients with AMD, which we simulated with eye-tracking. For this, we conducted a quantitative experiment with 16 healthy participants and a qualitative case study with a patient. Finally, we conclude with results and future work.