Loop points detected or automatically generated in gap streams. Since gaps areas might be allowed to grow or shrink (depending on the decoding settings preferred by the host application using the library) in the case of making a gap larger than it used to be, some data must be automatically created.
Such missing gap data is generated through looping either data elements that are marked as loop points through scripting (automatic or manual) or if no loop point is found, the last or first bit pattern defined in a gap stream gets looped according to the direction of gap generation.
When generating the encoded bit stream for a gap area, the decoder specifies that size of the area to be filled with a gap stream. There is a fixed size of a stream without looping data; if that’s enough to fill the buffer no looping gets performed, otherwise a scripted loop defined in the gap stream is used; if the loop point is not defined the automatically detected loop point will be used.