The present invention concerns a method of processing a coded digital signal containing on the one hand a set of samples of different types obtained by coding a set of original samples representing physical quantities and on the other hand a set of information representing original samples and parameters used during the coding, characterized in that it includes the following steps: determining the subset of samples corresponding to a part of the coded digital signal using the set of information, obtaining the number of samples of at least one predetermined type and which are contained in the given subset of samples, deciding with regard to a modification of the determined subset of samples according to the number of samples obtained.