| A short disclaimer |
I have nothing against Generative Fill. It’s a really helpful tool, and we use it a lot in our editing workflow. However, sometimes it’s just not enough, especially when it produces poorly detailed or unreliable results.
So I’d like to share a small on-set trick I’ve been using for years, which may come in handy during your post-production stage.
You’ve found a perfect scene, but there’s a distracting element that may be difficult to remove in Photoshop (for example, a traffic light).
Slightly change the camera position while keeping roughly the same viewpoint, and capture an extra shot where the unwanted element has shifted out of the way.
This extra frame can be used for compositing during post-production, or at least as a reference when you’re not exactly sure what details are hidden behind overlapping objects.
A comparison between Generative Fill and manual editing using the extra shot.Yes,
I understand that in most cases you wouldn’t zoom in that much and such GenAI results are acceptable too.
However, it’s always good to have extra material to work with during editing, isn’t it?
However, it’s always good to have extra material to work with during editing, isn’t it?
