Something like that. I could base it on frame-to-frame changes, or if my app was doing the screen recording, I could look at keyboard/mouse input as another sig

Something like that. I could base it on frame-to-frame changes, or if my app was doing the screen recording, I could look at keyboard/mouse input as another signal of "non-silence". I'm pretty much looking at silence removal as a good starting point. My overall goal is to cut down on the manual editing required, so just looking for repetitive processes that I could add automations for.