An apparatus for processing an audio signal and method thereof are disclosed. The preset invention includes receiving object information and a downmix signal including at least one object, the object information comprising data type information and at least one of object gain and object gain ratio; determining whether preset information to render the object is included in the extension region of the object information, based on the data type information; generating initial preset information to render the object by using at least one of the object gain and the object gain ration, if the preset information is not included in the object information; and rendering the object being included the downmix signal by using the initial preset information.Accordingly, even if preset information is not received, it is able to reconstruct an audio signal by generating initial preset information using received object information.