Hi,
the title says it: Why is a video sequence that we can edit in the VSE assigned to a certain scene (the one selected at the top of the application window)? IMO, VSE is a superordinated job as we can add strips from different scenes into one video sequence.
Umm, sorry what? You can add other scenes as strips themselves to a VSE edit. The VSE can live in any scene but is most use residing in its own scene. The scene you use to edit in is generally not an animation scene unless you need to sync audio.
Why is a video sequence that we can edit in the VSE assigned to a certain scene? I don’t know how to say it in other words. Creating the video sequence in the VSE is a superordinated job as it can contain many scene strips, hence it does not make sense to assign the video sequence to a certain scene. In Blender, you can have as many video sequences as you have scenes. This does not make sense because there is no relation between the number of scenes and the number of video sequences. If you want to create more video sequences than you have scenes, you have to create additional, empty dummy scenes even if you only want to have additional video sequences.
As a programmer and database designer this kind of relationship would be called bad design (or just wrong), unless there is a good reason to do so, and that’s the reason I’m asking for.
Remember, the VSE is an old clunky part of Blender that was brought forward from the 2.4 series without much re-design. While it is tempting to think of the dependancy graph as a database, it is not. It is more of a pool of linked resources that must be evaluated in a the correct order for animation to work correctly. The Scene is an arbitrary division within Blender, just like a layer or a group. But all data resides in the pool.
DAG:
http://www.blender.org/development/release-logs/blender-240/dependency-graph/
I don’t think of it as a database. It should say I have a good understanding of relations. Reasonable relations should exist in any application, not only in a database. But I understand that Blender has grown and not everything is as expected as if designed from the start.
Imagine there is a family consisting of whatever number of members. They can have 0 or many photo albums. There can pictures of any member in any photo album, and a photo album can contain pictures of different members. Nobody has to hold the albums. They are just lying there in the family’s house. The number of photo albums has no relation to the number of family members.
What Blender enforces is that every album must be held by a family member, and every member can hold only one album. Consequently, if you want more foto albums than family members, you have a problem. You would have to “produce” a new family member that can hold an album. That additional member is not really wanted because it is just there to hold an album. And there’s the not-answerable question which member should hold which album. That’s why this enforcement does not make sense.
(oh, just in case: this is only an analogy!)
As I say I gather that it’s original intended use was for sliding in audio to sync to animations.
Just a quick question, maybe it’s off topic, maybe it’s related, I’m not sure, but do you HAVE to animate your scene in Blender, then light and render in Blender, to use the VSE for piecing together the image sequence you render out? Because I have been struggling to import image sequences into Blender for compositing and I still can’t get it right. Probably just making stupid mistakes, but that’s been bugging for a while now and can’t get it right.
Thanks in advance guys 
Ot but, you can place any image sequence in the vse for rendering. Blenders or not. Just make sure they are a numbered set with last characters numerical. Eg picture00100.png
You don’t have to render a scene from vse it’s really just for adding sound and trimming shots.
Awesome stuff, Thanks!