Automatic conversion
Wire whatever you have into whatever you need — DNA quietly converts it for you. Only the baking and meshing conversions cost real time.
One of the nicest things about DNA: you rarely have to think about whether two nodes "fit together." Plug a shape into an image input, a number into a colour, a gradient into a fill — if there's a sensible way to convert, DNA does it behind the scenes. No special "convert" node to hunt for.
It still keeps every socket honestly typed (so socket colours and previews stay meaningful), and it tells you in the console whenever a conversion actually happens. But mostly, you just connect things and they work.
The one sentence to remember
Wire whatever you have into whatever you need; only the baking and meshing ones cost real time.
Most conversions are instant re-labelling. A few do genuine work — turning a recipe into pixels, or shapes into a solid mesh. Those are the only ones worth being deliberate about.
Free conversions (instant)
These cost essentially nothing — DNA is just re-reading the same data a different way:
Number into a colour — becomes a shade of grey (the number sets the brightness).
Number into a position or vector — fills the first axis.
A Gradient into a colour input — the gradient becomes a left-to-right colour field, ready to sample or composite.
A shape into a field or fill input — your Analytic shapes flow straight in as a usable field.
A light into a colour input — a light is a colour-over-space, so it drops into colour graphs directly.
An image into a colour input — an image already counts as a colour field, no conversion needed.
A specific kind of collection into a general one — e.g. your circles flowing into a node that accepts "any collection."
If a wire connects without a pause or a console note, it was a free conversion. Connect freely and don't overthink it.
Conversions that cost real time
These do actual computation — often on the GPU. They're still automatic, but they're the ones to be aware of when a graph feels heavy:
Shapes or collections into pixels (rasterising) — when shapes meet an image input, DNA renders them to a pixel image first. This is the most common "real work" conversion.
Baking a field into an image — a procedural field (noise, a gradient recipe, a colour formula) gets rendered to actual pixels because the next node needs real pixels to sample, not a recipe. Baked at a default resolution you can adjust.
Building a Distance Field from a mesh — measures distance to the surface so the shape can be blended, offset, or combined smoothly.
Points into a solid shape — a cloud of points becomes a smooth blobby surface (think metaballs), ready to mesh.
Meshing a field into geometry — turns a Distance Field or Volume into real geometry with points and surfaces.
Baking a Distance Field into a Volume — fills a 3D grid so volume-only nodes can read it.
A baked image or volume uses a default resolution. If the result looks soft or chunky, the automatically-inserted conversion exposes a resolution setting you can raise — no need to rewire anything.
Not everything converts. Some pairs genuinely don't fit (for example, a 3D volume into a flat-image-only input). When that happens the wire is refused and the console explains why, rather than silently producing something wrong.
Watching it happen
DNA reports conversions in the in-app console as it builds your graph. You'll see when an automatic conversion is inserted, and a special note whenever data crosses between CPU and GPU — those crossings are often the slowest link in a chain, so the console flags them on purpose. If a graph feels sluggish, this is the first place to look.
The "cook" badge on a node is related: it tells you how often a node re-runs its work. A heavy conversion that re-cooks every frame is worth more attention than one that runs once.