Hi all, I'm working with a team on what is essentially a JRPG and we've decided to rebuild our cutscene engine using the sequence editor because of the great visual feedback while developing instructions to all of your NPCs and environment objects - we have people working on it that work much better in the GUI than in lines of cutscene instruction scripts. HOWEVER we've come across some snags, one of which seems impossible to solve in the sequence editor's current state. Some quick background on our current design:
- NPCs each have specific objects/sprite indexes
- Since you can't update an object's sprite index in the sequence editor, we currently have a system where we convert any objects used in a cutscene into sprites. We put one invisible frame of the associated object in at the beginning and end of the cutscene to tag their starting and ending locations and when the cutscene starts a manager object grabs all the NPC objects being used in a cutscene and overrides the objects in the cutscene with the actual game objects.
- ...which means for the length of the cutscene the NPCs lose any smart handling of depth sorting/game states. Sprites can't walk around each other and be depth sorted properly because the entire sequence is drawn with sprites at a single depth.
- We considered using broadcast messages to update sprite indexes so we could just use objects rather than sprites but then you lose the whole point of the sequence editor - you can no longer see what your cutscene actually looks like and you end up with a lot of difficult-to-manage code inside of broadcast messages. Also sprite indexes change constantly within a cutscene.
- work around the inability for sprites to have dynamic depth sorting in cutscenes
- do something fancy where we use the sequence editor as a sort of instruction map by parsing the sequence objects for keyframe data and manually applying it to NPC objects in the cutscene rather than having the sequence draw the sprites