Extracting Stills using Resolve Python API

I'm writing a small program that will make metadata reports from video footage using Python.

I need to be able to extract metadata and images from nearly every kind of major video format. The Resolve Python API is very exciting, and it's doing a great job of making metadata extraction very simple.

The missing link for me right now is to easily get "preview" stills for each clip. I've figured out rendering clips with Python through the README documentation, including how to set in and out points for render timelines. I realize that I could do this over and over, adding new render jobs for every still, but Resolve performs very slowly doing this. I'd like to get a few dozen stills from each clip, and processing them with this method sometimes takes longer than just transcoding the whole clip.

I think extracting stills automatically could be a really useful application of Resolve's new scripting abilities. I'm wondering if anyone has any ideas, or if there'd be any backing for a feature request.

Thanks in advance!

The problem is that the stills don't exist as such until you explicitly create them by rendering which you have already tried.

There are clip thumbnails, but they are lowres and they also don't exist until generated.

There are still image galleries that the color page creates, but those stills don't exist until user generates them.

The only feasible method I can think of is to use FFMPEG to extract stills from the source media files. But you could run into codecs FFPMEG doesn't support.