Working in 2D and 3D

In DNA there is no wall between flat design and 3D — everything shares one space, and your 2D work just lives in it.

Most apps make you choose up front: a 2D canvas or a 3D scene. DNA doesn't — flat design and 3D share one space. What decides where a thing starts is its type: make a mesh, scatter points, or load a gaussian splat and it's created in the 3D scene; draw a flat shape, type some text, or bring in an image and it's created as a 2D composite layer. Either way it's the same world — orbit the camera and your flat work is right there in space — and you can move anything between the scene and the layers whenever you like. No conversion, no "switch to 3D" button.

One space, no mode switch

There is no global 2D-or-3D mode to flip. The behaviour you get is just a result of what you make and where you put it.

• Draw a circle, rectangle, line, some text, or a pen path and it shows up flat and front-on — exactly what you'd expect from a 2D tool.

• Add a sphere, import a model, or scatter points and it lands in the 3D scene.

• Mix them freely. Your flat shapes and your 3D objects are in the same world the whole time.

New flat content quietly behaves like a stacked design layer (handy for posters, titles, UI mockups), while 3D content behaves like a scene object. You never have to declare which — DNA reads it from the kind of thing you made.

Two ways to look: flat-on and orbiting

The viewport has two camera styles, and you can switch between them without losing your place:

• Orthographic — a flat, head-on look with no perspective. This is the natural way to work on 2D layouts, where pan and zoom feel like a normal canvas.

• Perspective — the real-world look with depth and vanishing points, for working in the scene.

Switching keeps the view stable, so flipping to perspective to check depth and back to flat to keep designing is painless. You can also look through a camera you've placed in the scene — like cutting to the "active camera" — and navigating then drives that camera directly.

How a flat shape lives in 3D

Here's the part that makes the whole thing work: a flat shape stays a flat shape, but it can be placed and turned anywhere in 3D.

Each shape (or each letter, each row) carries its own placement — a position, an orientation, and a scale. So you can tilt a flat logo so it recedes into the distance, stand text up like a billboard, or lay a panel flat on the ground. The shape itself is still the same crisp flat artwork; it's just oriented in space.

This is why flat work in DNA never feels "stuck to the screen." Tilting a flat shape with its orientation does not turn it into heavy 3D geometry — it stays light and sharp, just angled. Only genuinely three-dimensional content (a shape pushed into real depth) becomes scene geometry.

Crisp vector work, in a real scene

Flat shapes in DNA are resolution-independent. Zoom in and edges stay clean; the picture is drawn fresh at whatever size and angle it ends up, rather than being baked into pixels early. That holds true even when the shape is tilted in perspective or seen through a scene camera.

Because of this, your 2D vector work and your 3D scene aren't two separate pipelines that meet at the end — they're rendered together into the same image. A flat title can sit in front of a 3D product shot, or behind it, or be lit by the same lights, all in one go.

Stacking flat work into layers

When you do want classic layer-style compositing — this on top of that, with blend modes and opacity — you promote content into ordered layers. A flat design naturally falls into this stacked-layer behaviour, and a layer can even contain a whole 3D scene of its own.

You don't set up a separate compositing document for this; layers are just an arrangement of the objects you already have. See Layers & compositing for how stacking, blend modes, and track mattes work.

See also

• Analytic shapes

• Layers & compositing

• Fields

• The viewport

• shape.circle