Abstract: Method of automating at least a portion of an output gradation adjustment job for an image output apparatus. Test image data are corrected according to an initial gradation correction curve, and a test image is formed according to the image data thus corrected. The test image is read by a scanner to obtain read data. Based on the read data thus obtained, test image data are newly formed. Based on the newly formed test image data, a test image is again formed. The second-time test image thus formed is read by the scanner to obtain read data. Based on the read data obtained by reading the second-time test image, the initial gradation correction curve and a predetermined reference output curve, candidate point data are obtained. Based on the candidate point data, a gradation correction curve is formed and set.