PURPOSE: To heighten the regulation precision of film-thickness by a method in which the correcting displacement-amount of each lip piece corresponding to film thickness-correcting amount is computed by accumulating the deviations between the mean value of whole width and the mean value of partial width, and the position of each lip piece is controlled, based on said correcting displacement amount, and then this process is repeated.

CONSTITUTION: A film thickness-measurer 6 is caused to scan the film in the width direction of a film. While mean value of the whole width which is the mean value of the value from the film thickness value measured in each one scanning by the measurer 6, is computed, the mean value of partial width which is the mean value of the film thickness-value in every partial width corresponding to each lip piece 1b1, is computed. The deviation between the mean value of whole width and the mean value of partial width is calculated, and the correcting amount of film thickness is computed by accumulating the deviation at each partial width in every scanning by the film thickness-measurer 6. The correcting displacement amount of each lip piece 1b1 corresponding to said film thickness-correcting amount is computed, and the position of each lip piece 1b1 is controlled, based on said correcting displacement-amount.