A system for providing auditory feedback to visually impaired users when defining visual models. As the visually impaired user creates a visual model, the system detects when an object is placed in such a way that it overlaps with at least one previously placed object. In response to detection of an overlap, the system causes an auditory warning signal to be produced, indicating to the visually impaired user that object is overlapping a previously placed object. The disclosed system operates to create an invisible object within the user interface to cause a warning to be output, or modifies the object whose placement caused the object overlap to include the overlap warning. The user interface focus is then changed or reset to be on the invisible object or the object that caused the overlap to occur, enabling the user to conveniently fix the overlap condition.