I've lost exactly one action camera to a bad mount, a suction cup on a car door that gave way on rough road at speed, and watching a five-hundred-euro camera bounce down a highway teaches you more about mounting discipline than any tutorial ever will. A few of the lessons from that and other, less dramatic near-misses.
Test every mount at the actual speed it'll experience
The suction mount that failed had been tested at a standstill and on slow-speed footage the week before, and it held fine. It never got tested at highway speed before the shoot that actually needed it, and that gap between "tested" and "tested under real conditions" is exactly where failures happen. I now test any new mounting method at the fastest, most extreme version of the actual use case before trusting it on a paid job.
Redundancy costs almost nothing
A safety tether, a cheap length of cord attaching the camera to something that isn't going anywhere, costs a few euros and takes ten seconds to attach. I skip it constantly on low-risk static mounts and never skip it anymore on anything involving speed, vibration, or a mount I haven't used a dozen times before. The cost of the tether is nothing. The cost of not having one, once, was real money and a missed shot.
- Test mounts at the actual speed and conditions they'll face
- Use a safety tether on any mount involving speed or vibration
- Check mount integrity again after any bump, don't assume it's still fine
The angle lesson, separate from the safety lesson
Beyond safety, I've also learned that a mounting angle that looks clever in my head rarely survives contact with actual footage review. A helmet mount tilted for a "dynamic" angle usually just looks like a mistake once you're watching it back rather than imagining it. I test angles now by recording ten seconds and actually reviewing on a screen before committing a full shoot to an unconventional mount position, which has saved more footage than it's cost in setup time.