A method of decoding video data, the method comprising: uncapsulating a fragment of a random access point image, RAP, of a bit stream from a network abstraction layer unit, NAL, in which the NAL unit includes a value of the NAL unit type that indicates whether the RAP image is an instant decoder update image, IDR, and also indicates whether the RAP image is a type of RAP image that may have associated images main ones that are decodable main images, DLP, in which each main image follows a random access image in decoding order and precedes the random access image in output order, and in which to be the unit type value of NAL a first value of the NAL unit type indicates that the RAP image is an IDR image that can have associated DLPs and in which the value of the NAL unit type is a second value of the unit type d e NAL indicates that the RAP image is an IDR image that has no associated DLP; determine that the RAP image is an IDR image that DLP may have associated on the basis of the value of the NAL unit type being the first value of the NAL unit type, or that the RAP image is an IDR image which has no associated DLP, based on being the value of the NAL unit type being the second value of the NAL unit type; decode the RAP image and bitstream video data following the RAP image based on the value of the NAL unit type; determine, for a subsequent NAL unit, that a value of the NAL unit type indicates that the subsequent NAL unit comprises a supplementary improvement information NAL unit, SEI, debuff that includes a suffix SEI message, wherein the suffix SEI NAL unit succeeds, in decoding order, to the last NAL unit of the video coding layer, VCL, in an access unit, AU.