Mind-Blowing Ways AI Is Transforming Our Video Edits - from an editor at BearJam

 

So, I’ve been chatting with our editors to gather everything you need to know - so all you have to do is sit back and read. (And who doesn’t love that?)I’m even thinking of making a reel about this… so watch this space.

Alright, let’s crack on.

 AI VISUALS WITH RABIA

“AI tools have already become integrated in the editing software we use, but there are a few extra platforms I like to use, each with their own unique purpose.

For Upscaling and Sharpening Video

We use the same software, regardless of where I am editing, which is Topaz Video AI.

For sharpening, we had a situation where the lens we were using to have a continuous hazy effect in certain shots was too strong for one of the portrait shots and wasn’t discovered until post. I took the raw clip into Topaz Video Ai and fixed it in there and then graded it and brought it back to the edit.

 

Image 1Image 2

 

Topaz is also great for upscaling. We use this when we have a low resolution footage, so it looks better in the edit we would upscale it.

BearJam AI Video:

Image 3 Image 4

Content Aware Fill

Sometimes for a shot, we can use AIi to generate a frame to cover up objects.

An example of this is when we recorded a 5-person chat for Company of Cooks. Due to the location, we needed to put a light above the table - which you can see in the left image.

In post, we cropped the light out of shot as much as possible, then brought a still into Photoshop and used generative fill to mask the light out. That mask was then brought back into the edit on top of the footage - a seamless patch!

This technique really only works for a static camera shot; for a moving camera, we’d deploy a more intricate but similar technique in After Effects.

 

Image 5Image 6

Multi-Cam - AutoPod or AI Smart Switch

I tend to only use Multi-Cam tools when doing podcasts.

To quickly explain, Multi-Cam is referring to when you are switching between multiple cameras. An example in a podcast setting: you would typically have 3 cameras, 1 wide and 2 mid shots. The Multi-Cam process switches between these three angles depending on who is talking and the reactions of each person.

I would generally use the Vision Mix as the base and make any changes I want from that. However, if I don’t have a usable vision mix, I would use a Multi-Cam tool. I then treat what it gives me as a base and make changes from there.

In Premiere Pro, I would use the plug-in Auto Pod, but in DaVinci Resolve, I would use the new built-in tool AI Smart Switch. These two tools save me a lot of time and give me a great base I can work from.

AI is a great tool to help with mundane and time-consuming tasks, but it can’t just be used on its own. The work still needs the human spark that makes the work shine.”

 

I don’t know about you, but my mind is blown. If you’d like to dive deeper, feel free to reach out - and keep an eye on the blog for more... 

Read On