The viewport
The viewport is your live window onto the scene — fly the camera around it, look through your own cameras, and switch on the overlays that help you compose.
Everything in DNA lives in one 3D-native space, even your flat 2D work. The viewport is where you see it, move around it, and pick things directly with the mouse. It always shows the result of your graph, so as you tweak nodes the picture updates live.
Perspective and orthographic
The viewport has two ways of looking at the world, and you can flip between them at any time:
Perspective — the natural "lens" view, where things get smaller with distance. Use this for 3D scenes you want to feel like a real space.
Orthographic — a flat, parallel view with no perspective falloff. Ideal for laying out 2D work, lining things up, and precise placement.
Switching between the two keeps your view stable — it doesn't jump or re-frame. You can still orbit in orthographic (hold Alt and drag), so it stays a fully 3D view, just without the lens distortion.
Flat 2D compositions feel most natural in orthographic. 3D scenes feel most natural in perspective. There's no separate "2D mode" to switch into — it's just a projection choice.
Moving the camera
Navigation uses the moves you'd expect from any 3D app:
Orbit — Alt + drag to rotate around the point you're looking at.
Pan — middle-mouse drag (or Space + drag) to slide the view sideways.
Dolly / zoom — scroll the wheel to move in and out; in orthographic this zooms around your cursor.
When you frame up something you want to keep working on, focus the selection to recentre the view on it so orbiting pivots around that object instead of the world origin.
Looking through a camera
By default you fly a free interactive camera that's just for you. But you can also look through any Camera node in your scene — the same idea as an "active camera" view in other apps. This shows exactly what that camera sees, framing and lens included.
The clever part: while you're looking through a camera node, orbiting, panning, dollying and zooming all drive that node directly. You're not flying a disconnected camera — you're animating the real one, and your moves are saved back onto it. Selection handles and picking stay locked to the rendered view, so what you grab is always what you see.
If the camera is in FPS mode, looking through it captures the mouse for free-look; press Escape to release it.
A Camera node is just the viewpoint and lens. To turn a scene into final pixels you still use the Render node — see Cameras.
Sharp vs Pixel Preview
The viewport can draw your composition two ways, controlled by the Pixel Preview toggle:
Sharp (the default) — draws everything at your screen's pixel density, so vector shapes, text and Analytic shapes stay crisp no matter how far you zoom in.
Pixel Preview — renders the comp once at your project's real export resolution and magnifies those pixels, so what you see is exactly what exports. A 1:1 pixel grid appears once you're zoomed in far enough to make individual pixels out.
Use Sharp for clean authoring; flip on Pixel Preview to check aliasing, fine detail, and how things will actually look in your output.
If the grey border around the canvas "disappears", you're simply in Sharp mode — that grey frame is Pixel Preview's letterbox. Turn Pixel Preview on to bring it back.
Overlays
Overlays are display-only helpers — they never change your scene, and there's a master switch to hide them all at once. The main ones:
Grid — a ground/working grid with adjustable size and subdivisions, drawn as lines, dots or dashes, with coloured origin axes. Doubles as a snapping target.
Composition guides — the canvas boundary, safe areas, centre marks, and rulers for framing your shot.
Guidelines — your own draggable horizontal and vertical guides to line work up against; objects can snap to them.
Wireframe and normals — see the underlying mesh edges, or draw little hairs showing which way each surface faces (handy for debugging flipped geometry). There's also bounding boxes, point IDs, object names, X-ray, and a "random colours" mode to tell shards and pieces apart.
Onion skinning — ghost the frames before and after the current one (with separate before/after tints) to judge motion. Motion paths show an object's trajectory over time.
Background — solid colour, checkerboard, or gradient, plus an optional reference image you can fade in behind your work.
Overlays like the grid and onion skin are viewport display settings, not part of your saved scene's look — they won't appear in your exports.
Snapping
Turn on Snap to lock placement to useful targets: grid intersections, your guidelines, the edges and centres of other objects, equal spacing between objects, and other shapes' vertices. You can also set rotation and scale increments. If snap is off, hold Ctrl to snap just for the current drag.