[Refine] Add "File -> Export -> Export to video" option under the `File` menu

[obrienciaran]

Doesn't necessarily rely on #14 but it makes sense to do that first.

Introduce a File -> Export -> Export to video option that allows users to render and export the current visual output as a .mov file. This feature should capture the right/output pane at its configured resolution and aspect ratio (e.g., 9:16 when enabled) and render visuals over time into a video file.

This is a placeholder issue that needs further refinement. We need to define (non exhaustive!):

• Scope

  • Simple real-time screen recording of the output pane vs. an offscreen/render-to-buffer approach, where frames are rendered directly into a video buffer without relying on the visible window
  • Real-time rendering that runs at playback speed vs. non-real-time rendering that can run faster or slower to guarantee deterministic frame output

• UX

  • Whether there will be a modal or pop-up to define export duration, e.g. like in TouchDesigner
  • Whether export can be controlled via a manual start/stop toggle, also e.g. like in TouchDesigner

• Technical approach

  • Render pipeline
  • Codec and container format
  • Frame rate
  • Audio handling

[aagentah]

Hey mate hope all is well. Thanks for sharing! This is definitely a cool feature to have; and appreciate you laying out the scope and approach. This all seems completely fine.

A few small notes from the top of my head we may consider:

• When does recording start: Once the user triggers the recording, does it instantly record or wait until first incoming hit/signal?
• Long recordings: How well does the recording hold up with long term recordings, 30-90 minutes? If something fails it'll be a sad day.
• What is recorded: Just visuals, audio on device too? There could be misconceptions especially as the inbuilt sequencer improves.
• File size: I haven't looked, but making some file-size estimates and checking if the user has sufficient storage before recording would be good UX.

[obrienciaran]

All very good points. Just to respond to those points.

• This point seems like a product decision that just has to be made I reckon. Waiting for a signal seems sensible, and immediately if there is already a signal.

• Indeed failures would be a disaster. I worked as an engineer in a video editing company for a few years and failing/lost recordings were a super common complaint. It's not a route you want to go down. IMO, I really think it's wise to focus on stability now. You have a wonderful product capable of making great visuals right now. That'll keep people happy which while work in the background can be done. Just my 2 cents!

• In my head, because there is a sequencer built in, my current mental model as a user is that the visuals seem to be coupled to the audio, and therefore would be coupled during the recording.

• Definitely. If there was a slider/input to select recording time such as in TD, this should be doable.

Sorry for the slow responses, I'm packing and sorting out some life admin ATM, and travelling to Japan for a few months. Ill be able to give this some better attention soon, probably from mid week next week!