AI Visual Storytelling

How do you combine numerous short video clips of 5-10 seconds into a whole movie? Technically we are not talking about whole full length movies but more like short movies of 3-10 minutes. That is basically the range I tend to work in.

If you Google how to make a movie you end up with something like this: The Film Making Process for Beginners.

It is rather superficial, just mentioning things like make a script, storyboard and hire a crew. To someone who wants to develop a good sense of how to combine numerous short AI clips into a movie it isn't very helpful.

In this story I will cover the following topics:

1. How to make longer movies from few clips. Essentially how do you effectively reuse clips?
2. Thinking in terms of A-roll and B-roll
3. Shot types, when and how to use them. Head shots, Medium shot (waist-up), wide shot (full body).
4. Transitions such as L-cuts and J-cuts and why they matter in AI film making
5. Special limitations with AI video clips and how to get around them

I will not cover these entirely in order or specific sections. It is just to give you an idea of the topics covered.

Making Longer scenes

Some video making services today allow you extend clips. But it is generally not working well and hard to control well. Hence that should not be your primary way of creating a longer video.

A classic approach is to animate a single image multiple times and string those clips together. The problem is that this looks very choppy. You get unnatural looking jumps in the footage of the same scene.