A graphic output device for generating a mixed graphic-textual depiction having at least one text object consisting of multiple letters for identifying a geographic feature, the output device comprising a processor unit, which is set up to reshape the text object in such a manner that the reshaped text object comprises, in addition to a normal text area having letters in the original size, a compressed text area having letters shrunken and/or compressed in relation to the original size, the boundary between the normal text area and the compressed text area running along a semantic jump point of the text object. A method for preparing a text object consisting of multiple letters using a processor unit, for the purpose of depicting the text object using a graphic output device.