Pre rendering files on timeline has no effect

john.kuehne wrote: ...slowing of export/render times when using 4K media on a 1080 timeline....
6. Size the media to 150-175%...we continue to deal with having to re-render these timelines when re-opening them on a new FCP start.

Can you try using a 4k timeline? Render file generation might take a bit longer but 4k render files have a greater chance of being re-used when doing a resize operation.

For export, consider trying File>Share>Master File>Settings, Format: Computer, Video Codec: H.264 Faster Encode, Resolution: 1920 x 1080. In most cases I cannot see any difference in quality using the "fast" export. That will only address the encode phase, not the render phase, but every little bit helps.

The invalidated render files upon re-launch IMO is a bug. I'm aware of nothing you can do to change that.

I believe some of these "lost render" issues might involve effects (both Apple's and 3rd-party). 
Due to Apple Silicon, all 3rd-party plugins that previously used the FxPlug 3 framework are being re-written for FxPlug 4. I don't know if built-in Fx use FxPlug or not. Many are actually Motion templates,

FxPlug 4 (which x86 will likely also get) will have several advantages:

- Ability to run plugins out of process so they won't crash or destabilize FCP.
- Control over rendering. Currently that is controlled mostly by the FCP, so when Neat Video, etc. bogs down the system and you didn't order a render, that's because FCP did for some unknown reason. With FxPlug 4 the plugin developer has control over this.
- Improved multi-threaded behavior. Currently when you simply add a compute-intensive plugin, it may grab the FCP UI for several seconds where you can't click on anything. FxPlug  4 will give developers more control over threading so plugin UI behavior should be smoother.
- FxPlug 4 plugins will use "tileable rendering" which should be more efficient, esp. on Apple Silicon unified memory designs.

It's conceivable that some of these internal improvements are already implemented for native effects on Apple Silicon. Since there is an interaction between Fx and the render problem, it would be interesting to test this on an M1 machine using the simple scenario I posted at  #110419

The current render behavior is frustrating, but there is hope it will be improved in the future. I should probably get an M1 for testing.

That's great news about fxfactory, yes I do use quite a few, mostly AddMotion