The goal is to build a small recording rig that holds my iPhone above the tablet, giving it a bird’s-eye view while I write or sketch.

This is the first concept, not a finished print. That is part of why I want to share it now: it is much more interesting to capture the thinking and iteration than to reveal a mysteriously polished object at the end.

The starting prompt

I gave Codex this description:

We are going to print a custom stand for video recording via a mobile.

I have a remarkable 2 tablet.

I want to create a tower, stand, to place my iPhone (mini 12) so it can record the screen.

The phone would be placed in portrait mode.

The tablet would sit naturally in portrait mode.

I am left handed, so perhaps we need to factor in where the supports are, to allow freedom of writing with the pen, for right or left handed people.

That last constraint matters. A stand can hold a camera perfectly and still be useless if its supports occupy the same space as the person writing.

What the first design is trying to solve

The concept has a rectangular base, two vertical supports, and a phone cradle above the work area. The reMarkable sits naturally in portrait orientation beneath it, while the iPhone 12 mini records in portrait from above.

The open-sided structure is intended to keep hands and the Marker free. I am left-handed, but the useful version should not force anyone into one side of the desk.

It also needs to be a practical print. The model is split into parts that can fit on a printer bed and connect into a taller structure, rather than pretending the entire stand can appear as one enormous piece.

Why nurb

nurb lets me describe a 3D-printable part in Python. That makes the stand feel closer to the custom software work that prompted it: dimensions, clearances, and structural decisions can live as code and change as the design changes.

Instead of nudging shapes around until something looks plausible, I can work from the constraints and regenerate the model. At least, that is the experiment.

Next

The render is a useful checkpoint, but the physical print gets the final vote. I still need to learn whether the base is stable, the joints are strong enough, the phone framing works, and the supports genuinely stay out of the way for both left- and right-handed writing.

Then I get to change the model, print another version, and find a new problem.

That seems like a good thing to write about.