I have a project that I need some direction on. The camera view will initially start out from above, far away from the landscape, showing the terrain with roads and railroad tracks. Then I need to zoom into a cell tower that has a shed with a radio in it.
My thought is that the zoomed out overview of the terrain would be one project. Then as the camera zooms in, it becomes another project with the two resulting movies placed sequentially in Premiere with a transition in between the two.
I want the terrain to look fairly realistic and then the shed and radio to look very realistic. Am I wrong in assuming that doing it all in one project would create an overly large file as well as a rendering nightmare?
My own experience says that it is very easy to have troubles if there is big differences in scale. I suggest that you split project to at least two scenes, example where you have one unit -> 1m where you have details and one unit -> 1km where you have scenery, clouds etc. If you zoom from outerspace, one unit -> 1000km where you get earth as sphere.
When you have assets in same .blend file project, it is easy to use them in different scale levels. You can even render smooth transition when you use identical camera track in different scenes.
It may be good idea to fly trough cloud in scene transition or you need to do some morphing to details that appear. Like using new scene objects as transparent and animate them to opaque, so they appear smoothly.
Basically, I suggest that you think of it like this: "the renderer is your camera, and the product thereof is a single strip of film.
“The movie- making process,” then, is actually what comes next: what Alfred Hitchcock aptly called “assembly.” This is where selected snippets of your various film-strips are joined to make the movie. There may be many “cuts” – sometimes in very rapid succession – between “strips of film” that were produced at entirely different times.
The CG workflow has sometimes been described as “edit, then shoot,” because of the computational expense of actually producing any finished shot. (Thankfully, EEVEE has made this very considerably easier!) In the earliest stages of your project, I suggest that you use stand-in objects, simple geometric things that are of the correct size and scale, and your human powers of imagination to try to advance your edit as far as you possibly can … before you even begin to undertake to add detail.
For instance, your present project concept seems to contain at least three entirely-different shots in more or less this sequence:
Establishing shot (“Long Shot”): Fly-over view of the terrain.
Cell tower.
Shed.
Radio in shed.
Each one of these shots would be rendered separately. Then, “the editing magic begins,” where you strive to get the rhythm of the entire sequence to “play out just so.”
Okay then: defer the business of “modeling the cell tower,” or the shed or the radio, until you’ve worked out that all-important rhythm. The cell-tower is a pyramid; the shed is box; the radio also; but all of them are to scale, and they’re set in similarly-plausible but “stand-in” surroundings. Strive to edit the movie before you begin to render it, so that these decisions will dictate what you do – and, don’t – have to render.
It reminds me of the director on the movie set holding his hands up and framing shots before the actual shooting of the film.
This is definitely how I will proceed with my project, use simple stand-in shapes, then animate the camera to achieve the most drama. Go back later and do the modeling and detail.