# DaVinci Resolve Fusion Light Node Tutorial: The Full Guide

> **Quick answer:** DaVinci Resolve's Fusion Light node adds Ambient, Directional, Point, or Spot lights to a 3D scene, but a light only works once it's wired into a Merge 3D node alongside your geometry and camera, and Renderer3D's Enable Lighting and Enable Shadows checkboxes are turned on. Only Spot Light casts shadows. All of it ships free, no Studio license required.

*Published by [TryUncle](https://tryuncle.com) — the on-screen assistant that teaches DaVinci Resolve on your own screen.*
*Updated 2026-07-25 · DaVinci Resolve 21.0.3 (July 2026) · Canonical: https://tryuncle.com/learn/davinci-resolve/davinci-resolve-fusion-light-node-tutorial*

You built a 3D scene in Fusion. A logo, a piece of text, maybe an object you're compositing into live footage. It looks flat, like a sticker pasted onto the frame, because nothing in the scene is actually lighting it. So you add a Light node, the obvious next step, and nothing changes. The scene still looks exactly as flat as before, or worse, it goes completely black.

That gap, between adding a light and the light actually doing anything, is where almost everyone gets stuck the first time they touch Fusion's 3D lighting. After running commercial edits for seven-plus years and building a professional video-editing community that's grown past 100,000 members, we've watched the same handful of Light node mistakes trip up editors who already understand keying, tracking, and 2D compositing perfectly well, because Fusion's 3D lighting model has one rule nothing else in the software prepares you for: a light isn't automatically part of your scene just because it exists in your node graph. This guide covers every light type, why your scene might be black, how shadows actually work, and where materials fit in, as of July 2026 on DaVinci Resolve 21.0.3.

## What is the Light node in DaVinci Resolve Fusion, and what light types does it include?

The Light node is Fusion's tool for adding a virtual light source to a 3D scene, and it comes in four distinct types that behave nothing like each other: Ambient, Directional, Point, and Spot. Each one exists to solve a different lighting problem, and picking the wrong one for the job is one of the quieter reasons a scene looks wrong even after a light is technically present and working.

Per [Blackmagic's Fusion manual on lighting types](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part737.htm), the four types break down like this. Ambient light "produces a general uniform illumination of the scene" that "exists everywhere without appearing to come from any particular source," and it "cannot cast shadows," instead softening a scene's shadowed regions rather than eliminating them. Directional light "is composed of parallel rays that light up the entire scene from one direction, creating a wall of light," with the sun given as the textbook example. Point light "is a light source with a clear position in space that emits light in all directions," the same way a light bulb does. Spot light is described as "an advanced point light that produces a well defined cone of light with falloff," and critically, it's "the only light that produces shadows."

That last fact is worth sitting with before you build anything, because it decides your entire lighting approach if a shot needs a shadow anywhere in it. Ambient, Directional, and Point lights can make a scene look believably lit, with highlights, falloff, and color, but none of them will ever cast a shadow onto another object or the ground beneath it. If your comp needs a subject to cast a shadow, a Spot Light isn't one option among several. It's the only one Fusion's 3D renderer offers.

Each light type also treats position, rotation, and scale differently, and this trips people up almost as often as the shadow limitation does. Ambient light ignores position entirely, since it fills the whole scene uniformly regardless of where the node sits in 3D space. Directional light ignores position and scale, responding only to rotation, since a directional source is defined purely by which way it points, like the angle of the sun in the sky rather than where the sun physically sits relative to your scene. Point light ignores rotation, since a bulb radiating in every direction has no meaningful "facing," but position and scale do matter. Spot light is the one type where both position and rotation genuinely apply, since it needs a location to shine from and a direction to aim its cone.

A DaVinci Resolve Fusion light isn't one tool with settings, it's four fundamentally different behaviors wearing the same node icon, and picking the wrong one for the job is a separate mistake from wiring it incorrectly. Knowing which type ignores which transform property before you start dragging Q, W, and E gizmos around the viewer saves you from wondering why moving a Directional light does nothing, or why rotating a Point light changes nothing either.

| Light type | Casts shadows | Position matters | Rotation matters | Real-world analogy |
| --- | --- | --- | --- | --- |
| Ambient | No | No | No | Uniform sky fill, no visible source |
| Directional | No | No | Yes | The sun |
| Point | No | Yes | No | A bare light bulb |
| Spot | Yes | Yes | Yes | A stage spotlight or flashlight |

## How do you add a light so it actually affects your scene?

A Light node produces nothing by itself. It has to be wired into a Merge 3D node alongside the geometry and camera it's supposed to illuminate, and that Merge 3D node's output has to reach a Renderer 3D node before anything visible happens at all.

This single fact explains most of the "I added a light and nothing changed" reports, and it comes directly from how Fusion's 3D system is structured. Per [Blackmagic's Fusion manual on the Merge 3D node](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part835.htm), "The Merge 3D node is the primary node in Fusion that you use to combine separate 3D elements into the same 3D environment." Without it, separate 3D objects exist in isolated bubbles that can't see or affect each other at all. The manual states this outright: "the camera would not be able to see the image plane and the light would not affect the image plane until all three objects are introduced into the same environment using the Merge 3D node." Camera, light, and geometry are three separate nodes that know nothing about each other until Merge 3D forces them into one shared scene. It's the same 3D-scene-assembly node behind the swapped background and foreground mistakes covered in our [guide to Fusion's 2D Merge node](https://tryuncle.com/learn/davinci-resolve/davinci-resolve-fusion-merge-node-background-foreground-backwards), just working in three dimensions instead of two.

The build sequence, every time, looks like this:

1. **Create your 3D geometry or image plane** (text, a shape, an imported model, or a 2D image projected onto a plane in 3D space).
2. **Add a camera**, either the default perspective camera Renderer3D provides or an explicit Camera3D node if you need control over lens settings or animated moves.
3. **Add one or more Light nodes** and set each one's type.
4. **Wire geometry, camera, and every light into a Merge 3D node.** Merge 3D accepts an expandable number of SceneInput connections, so a scene with three lights and two objects still only needs one Merge 3D to combine everything.
5. **Connect the Merge 3D's output into a Renderer 3D node.** This is the node that actually converts your assembled 3D environment into the flat 2D image you see in the viewer and eventually in your timeline.
6. **Confirm Renderer3D's Enable Lighting checkbox is on.** Without it, every light in your scene is present but functionally ignored.

A detail worth knowing before your first attempt: selecting a Light node on its own and loading it directly into the viewer shows you nothing at all, and that's expected behavior, not a broken node. Per [Justin Robinson's breakdown of Fusion's Spot Light node for JayAreTV](https://jayaretv.com/fusion/spot-light-node/), "Selecting a light node and loading it into the viewer does not show anything." A light has no image of its own to display. It only produces a visible result once it's part of an assembled scene that reaches a Renderer3D node, which is a different mental model from every 2D Fusion tool where loading a node into the viewer always shows you that node's output directly.

A Light node is invisible by design until it's connected into a Merge 3D node with the objects it's meant to illuminate and that Merge 3D feeds a Renderer 3D, which means the single most common reason a new light "does nothing" is that it was never actually introduced to the rest of the scene in the first place. That one wiring step is the entire difference between a light that exists and a light that works.

## Why is your Fusion 3D scene still completely black after you add a light?

This is one of the recurring questions in our 100,000+ member editing community, and it has two distinct causes that produce the identical black-frame symptom, which is exactly why it's confusing to troubleshoot from the viewer alone.

**Cause one: the light was never merged into the scene.** Covered in the section above, this is the most common root cause. If your light sits in the node graph but was never wired into the same Merge 3D node as your geometry and camera, it contributes nothing, and your object renders using whatever base material color it has, or fully black if the material itself expects lighting it never receives.

**Cause two: Enable Lighting is on with no lights actually reaching the scene.** This is the sneakier one, because it produces total blackness rather than a merely dim result. Per [JayAreTV's breakdown of the Renderer 3D node](https://jayaretv.com/fusion/renderer-3d-node/), "When the Enable Lighting checkbox is selected, objects are lit by any lights in the scene. If no lights are present, all objects are black." Read that carefully: it isn't that dim lighting produces a dark image, it's that Enable Lighting fundamentally changes how Renderer3D calculates every object's color, from "show the material's base color" to "show only what light actually illuminates," and if that second mode finds zero lights in the merged scene, there's nothing left to show at all.

That second cause explains a specific, confusing sequence a lot of people hit by accident: a scene looks correctly colored with Enable Lighting off, you turn it on to add a light for realism, and the whole frame goes black before you've even connected the new light node into the Merge 3D yet. That's expected. Enable Lighting switched the render mode first; your actual light hasn't joined the scene yet to fill the gap that switch just created.

| Symptom | Likely cause | Fix |
| --- | --- | --- |
| Scene renders at flat, unlit color, ignoring the light entirely | Light node never wired into the shared Merge 3D | Connect the light into the same Merge 3D as your geometry and camera |
| Scene goes completely black the moment you enable lighting | Enable Lighting is on, but no light has reached the merged scene yet | Add and merge a light before or immediately after enabling lighting |
| Light appears in the node graph but shows nothing when loaded alone in the viewer | Normal behavior, not a bug | Load the Renderer3D node's output instead, since lights have no standalone image |
| Scene is lit correctly but shows no shadow anywhere | Using Ambient, Directional, or Point light, none of which cast shadows | Switch to, or add, a Spot Light with Enable Shadows on |

A pitch-black Fusion 3D scene after enabling lighting doesn't mean your light is broken, it means Renderer3D switched into a mode where objects only show what light actually reaches them, and your light hasn't been merged into that scene yet to provide any. Diagnosing which of the two causes you're looking at takes one glance at the node graph: is the light connected to the same Merge 3D as everything else, or is it sitting off on its own, technically present but never actually introduced.

## What does Ambient Light do, and when should you use it?

Ambient Light provides flat, directionless base illumination across your entire 3D scene, without appearing to come from anywhere specific and without ever casting a shadow. Per Blackmagic's manual, it "produces a general uniform illumination of the scene," and it softens a scene's darkest corners rather than eliminating them, which is exactly the job it's built for: preventing shadowed regions of an otherwise well-lit scene from going pure black.

Position has zero effect on an Ambient Light's output, since it isn't coming from any particular direction or location to begin with. You can drag its transform gizmo anywhere in 3D space and the scene's illumination won't change at all, which surprises people the first time they try to "aim" it the way they would a Spot or Point light.

The practical use case is almost always supplementary rather than primary. A scene lit entirely by Ambient Light looks flat and shadowless everywhere, since there's no directional information for any surface to react to differently depending on its angle. But a scene lit by a Spot or Point light alone often has areas so far from the light source, or so occluded by other geometry, that they render as pure black, which reads as harsh and artificial rather than intentionally moody. A small amount of Ambient Light mixed in raises that floor without erasing the directional light's actual shaping, the same conceptual role "fill light" plays in a real three-point lighting setup.

Because it never casts shadows and its intensity doesn't fall off with distance, Ambient Light is also the cheapest light type to render in terms of Renderer3D's processing overhead, since there's no shadow map to generate and no distance-based falloff calculation happening per pixel.

Ambient Light's job is to raise the floor on a scene's darkest areas without adding a single visible shadow or directional highlight, which makes it the light you add last, after your Spot or Point light is already doing the actual shaping work. Add it first, and you'll spend the rest of your build fighting a flatness it's specifically designed to create.

## What does Directional Light do, and how is it different from Point Light?

Directional Light lights an entire scene from one consistent angle with parallel rays, the way sunlight does across an outdoor scene, while Point Light radiates outward from a single fixed position, the way a bulb does from wherever it happens to be sitting. That distinction, angle versus position, is the entire difference between the two, and it decides which one behaves correctly for a given shot.

Per Blackmagic's manual, Directional Light "is composed of parallel rays that light up the entire scene from one direction, creating a wall of light," and "the directional light can be rotated, but position and scale will be ignored." Move a Directional Light node anywhere you want in the 3D scene and nothing changes, because a directional source has no meaningful location, only an angle. Rotate it, and every surface in the scene reacts as if the sun itself had moved across the sky, uniformly, regardless of how close or far any object is from the node's arbitrary position marker.

Point Light works in the opposite way. Per the manual, it's "a light source with a clear position in space that emits light in all directions," and it "ignores rotation" entirely, since a bulb radiating equally outward in every direction has nothing for rotation to change. Position and scale, on the other hand, genuinely matter, since moving a Point Light closer to an object visibly brightens it, and moving it farther away dims it, at least once you've set a Decay Type that makes distance matter at all (covered in the next section).

The practical decision between them comes down to whether your scene has one consistent "sun" angle that should treat every object the same way regardless of where it sits, versus a scene where a specific light source, like a lamp, a screen glow, or a practical light visible in frame, needs to behave like it's actually located somewhere specific relative to nearby objects. A logo animation floating in an abstract void usually reads more naturally lit by a Directional Light, since there's no implied light source to place. A 3D object meant to sit convincingly on a desk next to a practical lamp reads more naturally lit by a Point Light positioned where that lamp would actually be.

Directional Light answers "which way is the light coming from," and Point Light answers "where is the light actually sitting," and confusing the two produces a scene that technically has illumination but doesn't match the physical logic of whatever you're trying to depict. If moving a light's position gizmo does nothing to your scene's brightness, you're looking at a Directional Light behaving exactly as designed, not a broken node.

## What does Point Light do, and how does falloff or decay actually work?

Point Light radiates illumination equally in every direction from a single position, and by default, per [JayAreTV's breakdown of the Point Light node](https://jayaretv.com/fusion/point-light-node/), "A point light defaults to No Decay, meaning that its light has equal intensity at all points in the scene." That default surprises a lot of people expecting bulb-like falloff automatically, since a real light bulb obviously dims the farther you get from it, and Fusion's Point Light doesn't replicate that unless you explicitly tell it to.

To get distance-based dimming, JayAreTV's breakdown notes you "set the Decay Type to either Linear or Quadratic modes." Linear decay dims intensity in direct proportion to distance, a straightforward, predictable falloff. Quadratic decay dims according to the inverse square law that governs how real light actually behaves in physical space, meaning intensity drops off much faster near the source and levels out more gradually further away, which is the mathematically accurate model for how a genuine point source of light behaves in the real world.

Two other controls do the rest of the work. Per the same source, Intensity is set on a slider where "a value of 0.2 indicates 20% percent light," making it a straightforward percentage rather than an unbounded arbitrary number, and Color uses "this standard Color control to set the color of the light," letting you push a Point Light warm, cool, or tinted to match whatever practical source it's supposed to represent.

The choice between No Decay, Linear, and Quadratic isn't just a technical preference, it changes how believable a scene reads the moment more than one object sits at different distances from the light. A scene with No Decay lights every object at identical brightness regardless of how close or far it is from the source, which looks obviously artificial once you have foreground and background elements that should logically receive different amounts of light. Quadratic decay is the physically accurate choice for anything meant to look like a real point source, a lamp, a candle, a practical bulb visible in frame, while Linear decay sometimes reads better for stylized or motion graphics work where a more gradual, controllable falloff matters more than physical accuracy.

A Point Light with No Decay, Fusion's default, lights every object in the scene at the exact same intensity no matter how close or far it sits from the source, which is almost never what a realistic scene actually needs. Setting Decay Type to Quadratic the moment you add a Point Light meant to represent a real-world source saves you from later wondering why a background object looks just as brightly lit as something sitting inches from the light.

## What does Spot Light do, and how do you set cone angle, penumbra, and dropoff?

Spot Light is an advanced Point Light that shines within a defined cone rather than in every direction, and it's the only light type in Fusion's 3D toolset capable of casting a shadow. Getting a Spot Light to look right depends on three interacting controls that most people either ignore or fight with individually instead of understanding together: Cone Angle, Penumbra Angle, and Dropoff.

Per [JayAreTV's breakdown of the Spot Light node](https://jayaretv.com/fusion/spot-light-node/), Cone Angle "refers to the width of the cone where the light emits its full intensity," and it has a hard ceiling: "Maximum value is 90 degrees." That cap matters if you're trying to build an unusually wide flood of light using a single Spot node, since past 90 degrees the cone simply won't get any wider, and a scene needing broader coverage than that needs either a Directional Light or multiple Spot Lights aimed at different angles instead of one enormous cone.

Penumbra Angle handles the soft transition beyond that full-intensity cone. Per the same source, it "determines the area beyond the cone angle where the light's intensity falls off toward 0," and "a value of 0 produces hard-edged lighting" while larger values soften the visible edge of the light's coverage. This is the control that decides whether your Spot Light looks like a crisp theatrical spotlight with a hard-edged pool of light, or a softer, more naturalistic lamp glow with no visible boundary at all.

Dropoff governs the shape of that fade, not just its width. Per JayAreTV, "Dropoff controls how quickly the penumbra angle falls off from full intensity to 0," meaning two Spot Lights with identical Cone Angle and Penumbra Angle values can still look meaningfully different depending on whether that transition happens in a sudden, sharp curve or a gentler, gradual one.

The three controls work together rather than independently, which is why isolated tweaking of just one often produces confusing results. A wide Cone Angle with zero Penumbra looks like a hard-edged circle of light with a crisp boundary, useful for a stage-spotlight or interrogation-room effect. A narrower Cone Angle with a large Penumbra Angle produces a soft, glowing pool with no clear edge at all, closer to a desk lamp or a window's light spill. Dropoff then fine-tunes how gradual or abrupt that transition reads once Cone Angle and Penumbra have set up the basic shape.

| Control | What it does | Effect at low value | Effect at high value |
| --- | --- | --- | --- |
| Cone Angle | Width of the full-intensity light cone (max 90 degrees) | Narrow, focused beam | Wide flood, up to the 90-degree ceiling |
| Penumbra Angle | Width of the soft falloff zone beyond the cone | Hard-edged, crisp light boundary | Soft, gradual edge with no visible boundary |
| Dropoff | Shape of the intensity curve within the penumbra | Sharp, sudden fade | Gentle, gradual fade |

A Spot Light's visible shape comes from three controls working together, Cone Angle setting the width, Penumbra Angle setting how soft the edge is, and Dropoff setting how that softness curves, so changing only one of the three while troubleshooting a light that "looks wrong" usually just trades one unwanted look for another. Adjust Cone Angle and Penumbra together first, coarse shape before fine curve, and Dropoff last, once the basic pool of light already reads correctly.

## How do you get a Spot Light to actually cast shadows?

Shadows require two separate switches to be on at once, not one, and missing either one produces a Spot Light that illuminates your scene correctly but casts no visible shadow anywhere, which reads as broken even though nothing is actually malfunctioning.

Per [Justin Robinson's breakdown of the Spot Light node](https://jayaretv.com/fusion/spot-light-node/), Enable Shadows is a setting on the light itself, and it's "the only type of light capable of casting shadows," with shadow generation "enabled by default" once that checkbox is on. But the light's own Enable Shadows switch only sets up the possibility. Per [Blackmagic's Fusion manual on the Renderer 3D node](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part839.htm), "When the Enable Shadows checkbox is selected, the renderer produces shadows, at the cost of some speed." That's a second, independent switch, on a completely different node, and both have to agree before a shadow shows up anywhere in your rendered frame.

Beyond the two on/off switches, a handful of settings shape how the shadow actually looks once it's visible. Shadow Map Size controls the resolution of the internal depth map Fusion generates to calculate the shadow, and per JayAreTV's breakdown, "larger values produce more detailed maps but increase memory and performance demands," a direct trade familiar from anything else in Fusion involving quality-versus-render-time decisions. Shadow Map Proxy expresses that same trade as a ratio, where, per the documentation, a value of "0.5 produces half-resolution shadow maps," useful for keeping interactive scrubbing responsive while you're still positioning lights, then raising back to full resolution for a final render.

Shadow Density controls how opaque the shadow itself appears, ranging from fully transparent at low values up to, per the manual, "completely opaque" at a value of 1.0. A shadow at less than full density lets some of whatever's behind the shadow-casting object show through, which sometimes reads more naturally than a pure black shadow, especially in scenes that already have some ambient fill light present.

Shadow Softness offers two distinct modes, and the choice between them matters more than it looks. Constant softness, per JayAreTV, "uses fixed filter width for uniform shadow softness," producing the same edge quality regardless of how far the shadow-casting object sits from whatever surface is receiving the shadow. Variable softness instead scales with distance, since "filter size adjusts based on distance between shadow caster and receiver," which more closely matches how real shadows behave: sharp and crisp close to the object casting them, progressively softer and more diffuse the further the shadow falls from its source.

A final, sometimes-overlooked control worth knowing about before you spend time chasing a different bug: Multiplicative and Additive Bias exist specifically to correct Z-fighting artifacts, flickering or striped patterns that show up on surfaces very close to each other in 3D space. Per Blackmagic's manual, the recommended workflow is to "adjust the Multiplicative Bias first, and then fine tune the result using the Additive Bias control," rather than treating either as a random slider to nudge until the artifact happens to go away.

| Shadow control | Where it lives | Effect |
| --- | --- | --- |
| Enable Shadows (light) | Spot Light Inspector | Turns on shadow casting for this specific light |
| Enable Shadows (renderer) | Renderer3D Inspector | Turns on shadow rendering for the whole scene; both switches required |
| Shadow Map Size | Spot Light Inspector | Higher values sharpen shadow detail at a memory and speed cost |
| Shadow Map Proxy | Spot Light Inspector | Ratio for lower-resolution preview shadows during interactive work |
| Shadow Density | Spot Light Inspector | Opacity of the shadow, from transparent up to fully opaque at 1.0 |
| Shadow Softness (Constant/Variable) | Spot Light Inspector | Constant is uniform edge softness; Variable softens more with distance from the caster |
| Multiplicative/Additive Bias | Spot Light Inspector | Corrects Z-fighting artifacts between close surfaces |

A Spot Light casting no visible shadow almost always means one of two Enable Shadows checkboxes is off, the light's own or the Renderer3D node's, since both have to be checked simultaneously before a shadow appears anywhere in the frame. Check both before you touch Shadow Map Size, Density, or Softness, since none of those settings matter at all until the shadow is actually turned on in the first place.

## Why are your shadows hard-edged instead of soft, or missing entirely?

If Enable Shadows is on in both places and a shadow still won't soften no matter how far you push Shadow Softness or Shadow Map Size, the actual cause usually isn't a light setting at all. It's which render engine Renderer3D is using.

Per Blackmagic's Fusion manual, "The Software mode is required to produce soft shadows," and stated even more directly, "The OpenGL renderer cannot generate soft shadows. For soft shadows, the software renderer is recommended." That's not a bug or a missing setting somewhere in the light's Inspector. It's an architectural limit of the OpenGL render path itself, and no amount of adjusting Shadow Softness, Shadow Map Size, or Dropoff on the light changes it, because the renderer generating the shadow simply doesn't support soft edges in that mode.

This produces a specific, confusing sequence: you build a Spot Light, turn on shadows in both places, see a hard-edged shadow, and start adjusting the light's own softness settings expecting the edge to blur. Nothing changes, because you're tuning a parameter the current render engine can't act on. Switch Renderer3D's render engine from OpenGL to Software, and the exact same softness settings that appeared to do nothing suddenly start working, since Software mode is what actually implements the soft-shadow calculation those controls are meant to drive.

The trade-off runs in the opposite direction on speed. OpenGL uses your GPU and renders considerably faster, which makes it the practical choice while you're still positioning lights, testing cone angles, and scrubbing through a shot to judge overall composition. Software mode runs slower, since it's the CPU-based path capable of the full soft-shadow calculation, and it's the mode you switch to once you're closing in on a final look and need the shadow quality that only it can produce.

| Render engine | Soft shadows | Speed | Best for |
| --- | --- | --- | --- |
| OpenGL | Not supported (hard-edged shadows only) | Faster, GPU-accelerated | Positioning lights, scrubbing, fast iteration |
| Software | Supported, full Shadow Softness control | Slower, CPU-based | Final render, any shot needing soft-edged shadows |

A second, less common cause worth ruling out if switching to Software doesn't fix the problem: check that the object you expect to receive the shadow, and the object casting it, both have their respective Shadow Receiver and Shadow Caster properties enabled. Per [Blackmagic's manual on lighting types](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part737.htm), individual 3D objects carry their own Affected By Lights, Shadow Caster, and Shadow Receiver controls, separate from anything on the light or the renderer. An object with Shadow Caster turned off will never throw a shadow, no matter how correctly everything else in the chain is configured, and an object with Shadow Receiver turned off will never show a shadow falling onto it either, even from a perfectly functioning Spot Light elsewhere in the scene.

A hard-edged shadow that won't soften no matter how far you push Shadow Softness usually means Renderer3D is still set to OpenGL, since that render engine cannot produce soft shadows at all regardless of what the light's own settings say. Switch to Software mode first, before touching a single softness slider, and revisit those sliders only after confirming the render engine can actually act on them.

## Why doesn't your light seem to affect your object's surface at all?

You've merged the light correctly, Enable Lighting is on, and the object still looks like a single flat color, unaffected by the light's position, angle, or intensity, with no visible highlight, shading gradient, or falloff anywhere on its surface. At that point, the problem usually isn't the light. It's the material describing how that surface is supposed to respond to light in the first place.

Fusion's 3D geometry needs a material, also called a shader, connected to its material input for lighting to visibly interact with its surface at all. Per [Bryan Ray's writeup on Fusion's 3D workspace](https://bryanray.name/2017/08/08/blackmagic-fusion-the-3d-workspace/), a professional VFX compositor, "Materials (also called shaders) determine how the surface interacts with light that hits it." That's the whole relationship stated plainly: the light supplies illumination, but the material is what translates that illumination into a visible highlight, shadow gradient, or reflection on any given surface. Without an appropriate material connected, or with a material set up incorrectly, a light can be working perfectly and still produce no visible response on the object it's supposedly illuminating.

Per [Justin Robinson's breakdown of the Blinn material node](https://jayaretv.com/fusion/blinn-node/), "The Blinn node is a basic illumination material that can be applied to geometry in the 3D scene," and it's the default material Fusion applies to new 3D geometry. Its two core controls, Diffuse and Specular, do most of the visible work. Diffuse "describes the base surface characteristics without any additional effects like reflections or specular highlights," essentially the object's base color under general illumination. Specular Color "determines the color of light that reflects from a shiny surface," and Specular Intensity and Specular Exponent shape how bright and how tight that highlight appears.

The mechanism connecting all of this to your actual lights is direct: per Robinson, "The Blinn model in Fusion calculates the highlight as the dot product of the surface normal and the half angle vector between light source and viewer." In plain terms, the material's math needs a light's actual position and angle to calculate anything at all, which is exactly why a correctly merged light with a correctly configured material produces visible shading, and why either piece missing on its own, a merged light with no material, or a material with no light reaching it, produces a flat, unresponsive surface that looks broken even though each individual node is functioning exactly as designed.

A 3D object that ignores your light entirely, showing no highlight, gradient, or falloff anywhere on its surface, is almost always missing a properly connected material, since Fusion's Blinn and Phong shaders are what translate a light's illumination into anything actually visible on that surface. Check the material input on your geometry before assuming the light itself needs more troubleshooting. If you're already comfortable animating Inspector values and want to bring a lit scene to life, our [guide to keyframing effects in DaVinci Resolve](https://tryuncle.com/learn/davinci-resolve/how-to-keyframe-effects-in-davinci-resolve) covers the diamonds and curves you'll use to animate a light's intensity, color, or position over time.

## Blinn vs Phong vs Cook Torrance vs Ward: which material should you pick?

Fusion ships with four distinct material shaders, and choosing between them is really a choice about what kind of surface you're trying to depict, since each one calculates specular highlights differently.

Per [VFXstudy's overview of Fusion's materials](https://vfxstudy.com/tutorials/materials/), "Fusion uses a model called blinn by default," which matches Blinn's role as the general-purpose starting point covered in the previous section: broad, soft-edged highlights that read well on rough plastic, stone, or wood. Phong is the next step up in surface shininess. Per the same source, "Phong produces slightly sharper highlights than blinn and may be good for plastics," making it the natural pick when you need a tighter, glossier highlight than Blinn's softer default gives you, the difference between a matte plastic case and a polished one.

Cook Torrance offers more granular control than either. Per VFXstudy, "Cook Torrance has more parameter controls and may provide extra flexibility if needed," which makes it the choice when neither Blinn nor Phong's simpler highlight model gives you enough control over exactly how a specific surface should respond, at the cost of a steeper learning curve for its extra parameters.

Ward exists for one specific job neither of the other three handles: directional, non-uniform highlights. Per VFXstudy, "Ward can create anisotropic specularity (specularity with direction). This is typical for brushed metals." Anisotropic specularity is the stretched, oval-shaped highlight you see on brushed aluminum, a vinyl record, or hair, where the highlight's shape depends on the direction of tiny grooves in the surface rather than spreading evenly in every direction the way Blinn, Phong, and Cook Torrance all assume.

Regardless of which shader you pick, the same underlying workflow applies. Per VFXstudy, "When creating or loading a material, first think about the Diffuse color," since that base color decision drives everything else, and "to get more control you can create your material as a separate node and attach it to the material input of the 3d model," rather than relying on whatever default material a piece of geometry ships with.

| Material | Highlight character | Best for | Extra controls |
| --- | --- | --- | --- |
| Blinn | Soft, broad highlight (Fusion's default) | Rough plastic, stone, wood | Diffuse, Specular Color, Specular Intensity, Specular Exponent |
| Phong | Sharper, glossier highlight than Blinn | Smooth plastic, porcelain, polished surfaces | Same core set as Blinn, tuned for a tighter falloff |
| Cook Torrance | Highly configurable, physically-based-leaning | Surfaces needing precise, non-default highlight shaping | More parameters than Blinn or Phong |
| Ward | Anisotropic, direction-dependent highlight | Brushed metal, vinyl, hair, grooved surfaces | Anisotropic direction and roughness controls |

Blinn is Fusion's correct default for most everyday 3D work, and the only reason to reach for Phong, Cook Torrance, or Ward instead is a specific surface quality Blinn's softer highlight can't reproduce: tighter gloss, finer control, or a directional highlight that follows the grain of a brushed or grooved material. Starting every new object with Blinn and only switching when the surface genuinely calls for it keeps your material choices deliberate instead of arbitrary.

## What is Ambient Occlusion, and how is it different from a real light or shadow?

Ambient Occlusion is a separate node from anything covered so far, and it's easy to confuse with Ambient Light or a Spot Light's shadow because all three darken parts of a scene, but the mechanism and the job are completely different.

Per [Justin Robinson's breakdown of the Ambient Occlusion node](https://jayaretv.com/fusion/ambient-occlusion-node/), "Ambient Occlusion (AO) is the lighting caused when a scene is surrounded by a uniform diffuse spherical light source," and "the Ambient Occlusion node generates global lighting effects in 3D-rendered scenes as a post effect." That last phrase matters: Ambient Occlusion is applied after your Renderer3D node has already produced an image, not calculated as part of the actual lighting pass the way a real Light node's illumination and shadows are.

What it produces is subtle contact shading, the soft darkening you see where two surfaces meet or nearly touch, like the shadow tucked into the crease where an object sits on a floor, rather than a distinct directional shadow cast away from a light source. Per Robinson, it "does not capture sharp shadows or Diffuse or Specular lighting," which is the key distinction from an actual Spot Light shadow: AO can't reproduce a crisp, directional shadow shape at all, only the soft, ambient-feeling darkening that appears in tight, occluded gaps between geometry regardless of where any actual light in the scene is positioned.

The reason compositors reach for it anyway is speed. Per Robinson, AO "quickly approximates computationally expensive ray-traced global illumination," the kind of physically accurate light-bouncing calculation that a real-time or near-real-time renderer like Fusion's Software or OpenGL engine isn't built to compute directly. Rather than simulating how light genuinely bounces between every surface in a scene, an expensive calculation even in offline 3D renderers built specifically for it, AO fakes the most visually important part of that effect, contact darkening, as a comparatively cheap post-process.

Setup requires a specific dependency chain. AO reads geometric information from your already-rendered scene, so it needs Z-Depth and Normals channels enabled in the Renderer3D node's output, and the same Camera3D used to render the scene has to feed into the AO node's own camera input, so it can correctly interpret which surfaces are near which other surfaces in 3D space. Per Robinson, the main tuning controls are "Kernel Radius," which "controls the size of the filter kernel in 3D space," and "Number Of Samples," where the practical guidance is straightforward: "Increase the samples until artifacts in the AO pass disappear."

Ambient Occlusion isn't a light and it isn't a shadow, it's a cheap post-process approximation of the soft contact darkening real global illumination would produce, which is why it can make a scene feel more grounded without ever needing a light source of its own or any of the shadow controls covered earlier in this guide. Reach for it when a scene lit correctly by your actual lights still feels like it's floating rather than sitting in contact with its environment.

## Which light type should you use for which situation?

Rather than treating this as four abstract options, it helps to work backward from what your shot actually needs.

| What your shot needs | Best light choice | Why |
| --- | --- | --- |
| Prevent shadowed areas from going pure black | Ambient Light | Uniform fill with no direction, no shadow, no position dependency |
| A single, consistent "sun" angle across the whole scene | Directional Light | Position-independent, only rotation matters, lights everything the same way |
| A specific practical source, like a lamp or screen glow | Point Light | Genuinely positional, with Linear or Quadratic falloff for believable distance dimming |
| A subject that needs to cast a visible shadow | Spot Light | The only light type that produces shadows at all |
| A focused, shaped beam, like a stage light or flashlight | Spot Light | Cone Angle, Penumbra, and Dropoff give direct control over the beam's shape |
| Subtle contact darkening where objects meet | Ambient Occlusion (not a light, but often paired with one) | Cheap approximation of global illumination's most visible effect |
| Realistic multi-object scenes with foreground/background depth | Point or Spot Light with Quadratic decay | Matches the inverse-square falloff real light actually follows |

Most real scenes end up using more than one of these together rather than picking just one. A common, effective combination for a 3D title or product shot: a Spot Light as the key, providing the main directional shaping and the only shadow in the scene, a low-intensity Point Light as a secondary fill from a different angle to soften the Spot's shadow side without adding a second one, and a small amount of Ambient Light to keep the darkest areas from crushing to pure black. That's the same conceptual three-point setup photographers and cinematographers have used for decades, translated into Fusion's four light types.

There is rarely a single "correct" light type for a scene, there's a correct combination, and the fastest way to a believable result is picking one light to do the shaping and shadow work, a second to soften it, and Ambient Light or Ambient Occlusion to handle whatever's left too dark. Reaching for all four at once from the start, before you understand what each one is actually contributing, tends to produce a muddier result than building up from one light at a time.

## Worked example: lighting a 3D logo animation from scratch

Here's how the full workflow plays out on an actual build, start to finish.

A motion designer needs a 3D logo, built from extruded 3D text, to look like it's genuinely sitting in a lit environment rather than floating as a flat, evenly-colored shape.

1. **Build the 3D text object** and confirm it has a Blinn material connected to its material input, since that's the shader that will actually respond to whatever lights get added.
2. **Add a Camera3D node** to control the framing directly, rather than relying on Renderer3D's default perspective camera, since an animated or specifically angled shot needs that explicit control.
3. **Add a Spot Light** as the key light, positioned above and to one side of the logo, with Cone Angle set wide enough to cover the full text and Penumbra Angle set to a moderate value for a naturalistic, not harsh, edge.
4. **Merge the text, camera, and Spot Light into a single Merge 3D node.**
5. **Connect that Merge 3D into a Renderer3D node**, confirm Enable Lighting is checked, and confirm the scene is no longer black or flat before proceeding.
6. **Turn on Enable Shadows** on both the Spot Light and the Renderer3D node, and check the shadow the logo casts onto whatever background plane sits behind it.
7. **Add a low-intensity Point Light** from the opposite side as a soft fill, with Decay Type set to Quadratic so it doesn't overpower the Spot Light's directional shaping.
8. **Add a small amount of Ambient Light** to keep the shadow side of the letters from crushing to pure black, checking the result against the Spot Light's shadow to confirm the fill hasn't erased it entirely.
9. **Switch Renderer3D's render engine to Software** for the final render, since the Spot Light's shadow needs to read as soft-edged rather than the hard line OpenGL alone would produce.
10. **Fine-tune Shadow Softness and Shadow Density** on the Spot Light until the cast shadow reads as grounded rather than either too sharp or too faint to notice.

Total build time on a shot like this: roughly twenty minutes for an experienced Fusion user, most of it spent on step 9 and 10's back-and-forth between Software-mode renders and adjusting shadow settings, since that comparison only becomes visible once you're rendering with the mode that actually supports it.

## Worked example: lighting an inserted object to match live footage

A second, shorter example covers a different common use case: compositing a 3D object into real camera footage so it looks like it was actually lit by the same light sources visible in the plate.

A compositor needs to insert a 3D product model onto a tabletop shot handheld, with a visible window providing the scene's main light from screen left.

1. **Match the object's position and scale** to the tabletop using the footage's perspective as a reference, ideally with camera tracking data already solved so the 3D camera matches the real one.
2. **Study the plate's actual light direction and quality.** The window light in this example is soft and slightly diffuse, not a hard, crisp source, which points toward Directional Light with a fairly wide, soft-edged quality rather than a tightly focused Spot Light.
3. **Add a Directional Light angled to match the window's direction**, since a window's light, while technically originating from a specific opening, behaves close enough to parallel across a small tabletop object that Directional Light's simpler, position-independent model reads convincingly.
4. **Merge the object, the tracked camera, and the Directional Light** into a single Merge 3D node.
5. **Add a low-intensity Ambient Light** to represent the general room bounce that a real interior always has, beyond just the direct window light, keeping the object's shadow side from looking unnaturally black compared to the rest of the plate.
6. **Enable Ambient Occlusion** after the Renderer3D output, feeding it the Z-Depth and Normals channels, to add the subtle contact darkening where the object meets the tabletop, since that ground-contact shadow is one of the strongest visual cues that an inserted object is actually sitting on a surface rather than floating above it.
7. **Compare the object's highlight and shadow direction against real elements already in the plate**, adjusting the Directional Light's rotation until both match, since a mismatched light direction is one of the fastest ways an inserted 3D object reads as obviously fake even when everything else about the composite is well executed.

The core lesson from both worked examples: the light type you reach for should follow directly from what the shot actually needs, a shaped, shadow-casting key for a logo that needs to feel grounded, or a soft, directionally-matched fill for an object that needs to blend into footage that already has its own established lighting.

## Do lights and shadows work the same in DaVinci Resolve's free version as Studio?

Yes. Every node covered in this guide, the Light node in all four of its types, Merge 3D, Renderer3D, both render engines, and every material shader, ships in DaVinci Resolve's free version, with no separate feature gate on any of it.

Per [Blackmagic's own marketing page for Fusion's 3D and VR tools](https://www.blackmagicdesign.com/products/fusion/vr3danimation), "Fusion lets you add unlimited, customizable light sources and control how elements within the shot react to lighting so you can set up incredibly detailed 3D sets," and the same page lists exactly the light categories this guide covers: "Control ambient, directional and point lights, or cast shadows over 3D geometry using the spot light." Nothing in that description, or anywhere else in Blackmagic's own materials, restricts any of it to Studio.

That matches the broader pattern for Fusion generally. Per [Toolfarm's in-depth comparison of Resolve Studio versus the free version](https://www.toolfarm.com/tutorial/in-depth-davinci-resolve-studio-vs-the-free-version/), "Fusion is included in both the free and paid versions of DaVinci Resolve," and the core node-based Fusion 3D workflow, including its lighting and material tools, is part of that shared, fully-featured base rather than a Studio-exclusive add-on.

What Studio actually adds sits elsewhere entirely: features like Magic Mask, Smart Reframe, and Speed Warp, all Neural Engine-driven tools on the Edit and Color pages, along with the Resolve FX Temporal category's plugin effects. None of that overlaps with the 3D lighting workflow this guide covers. Per Blackmagic's own [product comparison and pricing page](https://www.blackmagicdesign.com/products/davinciresolve/compare), Studio runs a $295 one-time perpetual license, and lighting a Fusion 3D scene isn't one of the reasons to pay it.

Every light type, the Merge 3D and Renderer3D nodes, both render engines, and all four material shaders work identically whether you're running DaVinci Resolve free or Studio, so if you've been holding off on building a lit 3D scene because you assumed it needed a paid license, that assumption is the only thing standing between you and building one right now.

## Does lighting performance differ on Mac, Windows, or with different GPUs?

The underlying math behind every light type, and how a material interprets that light, works identically across Mac, Windows, and Linux. Nothing about a Blinn shader's highlight calculation or a Spot Light's cone angle changes based on operating system. What genuinely varies by platform and hardware is how fast that math gets computed, particularly once shadows and the Software render engine enter the picture.

OpenGL rendering leans on your GPU, so its speed tracks graphics card performance directly, the same way it does for any other GPU-accelerated Fusion tool. Software mode, the one required for soft shadows, runs on the CPU instead, which means a system with a genuinely powerful GPU but a comparatively modest CPU sees a much bigger relative slowdown switching into Software mode than a more balanced machine would, since Software mode isn't using the GPU capacity that system has plenty of.

This has a practical consequence worth planning around if you're building scenes with soft shadows regularly: iterate on light position, cone angle, and color using the faster OpenGL engine, and switch to Software mode only for the final check and render, rather than working in Software mode the entire time and accepting slower scrubbing throughout the whole build process.

Multi-GPU systems, including Mac configurations with an eGPU or a machine carrying more than one graphics card, are the setup most likely to see an unexpected performance gap here, since GPU selection for accelerated processing has historically had more edge cases on multi-GPU configurations than a straightforward single-GPU workstation. If OpenGL rendering feels unusually slow given your hardware, or Software mode feels dramatically slower than expected relative to your CPU's actual performance, checking which GPU is actually assigned to Resolve under Preferences is a reasonable first step before assuming your lighting setup itself is too complex. And if a heavy scene, several lights, shadows, and an Ambient Occlusion pass stacked together, starts throwing errors rather than just rendering slowly, our [guide to Fusion's out of memory error](https://tryuncle.com/learn/davinci-resolve/davinci-resolve-fusion-out-of-memory-error) covers the RAM and VRAM settings that stop it.

None of this changes what a finished, correctly-lit scene actually looks like. Identical light settings, materials, and shadow configuration produce an identical rendered result on any platform, since the render math itself doesn't vary. Platform and GPU choice affect how long you wait to see that result, not what the result is once it finishes.

## Quick troubleshooting reference

Work through this table top to bottom for the fastest path to a fix.

| Symptom | Likely cause | Fix |
| --- | --- | --- |
| Light produces no visible change to the scene at all | Light not merged into the same Merge 3D node as geometry and camera | Wire the light into the shared Merge 3D |
| Selecting the light and loading it into the viewer shows nothing | Expected behavior; lights have no standalone image | Load the Renderer3D output instead |
| Whole scene renders completely black | Enable Lighting is on with no light reaching the merged scene | Confirm the light is merged; check Enable Lighting on Renderer3D |
| Moving the light's position gizmo does nothing | Using Directional or Ambient Light, which ignore position | Switch to Point or Spot Light if position needs to matter |
| Rotating the light does nothing | Using Point or Ambient Light, which ignore rotation | Switch to Directional or Spot Light if angle needs to matter |
| Object at any distance is lit at identical intensity | Point Light's default Decay Type is No Decay | Set Decay Type to Linear or Quadratic |
| No shadow appears anywhere, from any light | Using Ambient, Directional, or Point light, none of which cast shadows | Switch to, or add, a Spot Light with Enable Shadows on |
| Spot Light has shadows enabled but none appear | Enable Shadows off on either the light or the Renderer3D node (both required) | Check both checkboxes |
| Shadow won't soften no matter how high Shadow Softness is set | Renderer3D's render engine is set to OpenGL | Switch the render engine to Software |
| Object ignores the light, showing flat, unresponsive color | No material connected, or material misconfigured | Connect a Blinn, Phong, Cook Torrance, or Ward material to the geometry's material input |
| Scene looks correctly lit but feels like it's floating above its surface | Missing ground-contact darkening | Add an Ambient Occlusion node after Renderer3D, fed by Z-Depth and Normals |
| Cone Angle won't widen past a certain point | Spot Light's Cone Angle has a hard 90-degree ceiling | Use Directional Light or multiple Spot Lights for coverage wider than 90 degrees |

Every symptom in this table traces back to one of a small number of root causes covered throughout this guide, a missing Merge 3D connection, a switch that needs to be on in two places instead of one, a render engine limitation, or a missing material, which means troubleshooting Fusion lighting is really a process of elimination through four or five known checkpoints rather than open-ended guessing.

## Is there a faster way to learn Fusion's lighting tools than trial and error?

Fusion's 3D lighting model asks you to hold several unfamiliar rules in your head at once: which light types ignore which transforms, why a light shows nothing loaded alone in the viewer, why two separate Enable Shadows checkboxes both need to be on, and why a correctly lit scene can still look flat without the right material connected. None of that is obvious from the interface itself, which is exactly why "davinci resolve fusion light node tutorial" is a search people type rather than something they simply figure out by clicking around.

The honest starting point for learning Fusion's 3D tools generally is Blackmagic's own free training, which covers the node graph and 3D workspace fundamentals this guide assumes, alongside well-established YouTube educators like Casey Faris, whose Fusion-focused tutorial content has helped a large share of the editors currently working with these tools, and the Blackmagic forums, where several of the shadow and Z-fighting fixes covered above originated as real user reports before becoming documented best practice. None of that ecosystem is something a newer resource should pretend doesn't exist, and skipping past it wouldn't give you the full picture. If your scene involves fast-moving geometry that also needs realistic motion blur once it's properly lit, our [guide to Fusion's Vector Motion Blur node](https://tryuncle.com/learn/davinci-resolve/davinci-resolve-fusion-vector-motion-blur-tutorial) picks up exactly where this one leaves off.

Where that ecosystem tends to fall short isn't the quality of the explanation, it's the gap between a tutorial's screenshot and your own screen. A video shows someone else's node graph, with someone else's Merge 3D structure and someone else's Inspector layout, and translating that onto your own comp, especially the first few times, costs real time even when the tutorial itself is excellent. TryUncle is the on-screen assistant for DaVinci Resolve on macOS - ask in plain words and Uncle points at the exact control on your screen. If you're staring at your own Spot Light node wondering why your shadow still won't soften, Uncle looks at your actual screen and points at your Renderer3D node's render engine setting directly, instead of asking you to reconcile a tutorial's node graph with whatever you're currently looking at.

It's worth naming where TryUncle sits relative to other AI tools built around Resolve, since the category has real alternatives worth covering honestly rather than pretending don't exist. Sottocut, PremiereCopilot, heyeddie.ai, and cutagent.ai each take a different approach, generally automating parts of an edit or answering questions through a chat interface rather than watching your live screen. Uncle's job is narrower by design. It doesn't build your Merge 3D node tree or pick your material for you. It watches your actual Resolve window and points at where the control you need actually lives, in the moment you're stuck on it, which is a different job than automation and a different job than a chatbot answering from a general memory of how Resolve "usually" looks rather than what your specific project actually shows. [TryUncle](https://tryuncle.com/?utm_source=learn&utm_medium=blog&utm_campaign=davinci-resolve-fusion-light-node-tutorial) is a paid macOS app, currently in founder pricing at $29.99 a month for the first 100 seats, and macOS only, so if you're on Windows or Linux, the free training and community resources named above remain your primary path.

If you're weighing the best way to learn DaVinci Resolve broadly, not just this one node tree, our [learning roadmap for DaVinci Resolve](https://tryuncle.com/learn/davinci-resolve) covers free training, paid courses, and where a screen-aware assistant like Uncle fits alongside both, rather than instead of either.

Once your light is correctly merged into the scene, your material is connected, and your shadow is reading soft or hard exactly the way you intended, you're past the hardest part of this specific tool, and if you're mid-build right now, staring at a scene that's still stubbornly flat or black, that's exactly the kind of moment Uncle can point at the specific checkbox or socket on your own screen while you finish it.

## The verdict

DaVinci Resolve's Fusion page gives you a genuinely capable, genuinely free 3D lighting system: four light types covering everything from flat ambient fill to a shadow-casting spotlight, a Merge 3D node that assembles them into one working scene, and a Renderer3D node with both a fast preview engine and a slower one capable of soft shadows. Almost every confusing symptom this guide covers traces back to one of a handful of specific, learnable rules rather than a genuinely broken tool.

Work through the causes in the order that actually matters when something looks wrong. Confirm the light is merged into the same scene as your geometry and camera first, since that single missing connection accounts for the largest share of "my light does nothing" reports. Then confirm Enable Lighting and, if you need a shadow, both Enable Shadows checkboxes, since those switches decide whether Renderer3D acknowledges your lights at all. Only after both check out should you start tuning Cone Angle, Decay Type, or Shadow Softness, because adjusting a light's fine details on a scene that was never actually wired correctly wastes time solving a problem that hasn't started yet.

And keep the one fact that resolves most of the remaining confusion close at hand: a light illuminates a scene, but a material is what makes that illumination visible on any given surface. A perfectly configured Spot Light pointed at an object with no material connected produces exactly the same flat, unresponsive look as a broken light would, and the fix in both cases lives one node away from where the actual symptom shows up.

## FAQ

### What does the Light node do in DaVinci Resolve Fusion?

It adds a virtual light source, Ambient, Directional, Point, or Spot, to a Fusion 3D scene. Each type behaves differently: Ambient fills the whole scene with flat illumination, Directional lights everything from one angle like the sun, Point radiates in all directions from a location, and Spot produces a cone of light and is the only type that casts shadows.

### Why is my Fusion 3D scene completely black even though I added a light?

The two most common causes are a light that isn't wired into the same Merge 3D node as your geometry and camera, and Renderer3D's Enable Lighting checkbox being on with no light actually reaching the scene. Per Blackmagic's own Fusion manual on the Renderer 3D node, when lighting is enabled but no lights are present, all objects render black.

### Why does selecting a light node and loading it into the viewer show nothing?

That's expected, not a bug. A light node has no image of its own to display. It only produces a visible effect once it's connected into a Merge 3D node alongside the geometry and camera it's supposed to illuminate, and that combined scene passes through a Renderer 3D node.

### Which DaVinci Resolve Fusion light type casts shadows?

Only Spot Light. Ambient, Directional, and Point lights illuminate a scene but do not generate shadows in Fusion's 3D renderer. If your project needs a shadow, a Spot Light with Enable Shadows turned on, both on the light itself and on the Renderer3D node, is the only path to it.

### Do I need DaVinci Resolve Studio to use Fusion's lights and shadows?

No. The Light node, all four light types, the Merge 3D node, the Renderer3D node, and Fusion's material shaders all ship in DaVinci Resolve's free version. Fusion itself is included in both the free and paid editions of Resolve, with no separate 3D lighting feature gate.

### Why are my Spot Light's shadows hard-edged instead of soft?

Almost always because Renderer3D's render engine is set to OpenGL rather than Software. Per Blackmagic's Fusion manual, the OpenGL renderer cannot generate soft shadows, and the Software renderer is required for them. Switch the render engine to Software, then adjust the light's Shadow Softness setting between Constant and Variable.

### What's the fastest way to actually learn Fusion lighting instead of guessing at settings?

Build one real scene with a light, a Merge 3D, and a Renderer3D open next to a resource, whether that's Blackmagic's free training, a YouTube channel like Casey Faris, or a written guide like this one. The friction is translating a tutorial's node graph onto your own screen, which is the specific gap a screen-aware assistant like TryUncle is built to close.

## Sources

- [Fusion Reference Manual 18.6: Lighting Types Explained (Blackmagic Design, mirrored)](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part737.htm)
- [Fusion Reference Manual 18.6: Spot Light [3SL] (Blackmagic Design, mirrored)](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part858.htm)
- [Fusion Reference Manual 18.6: Renderer 3D [3Rn] (Blackmagic Design, mirrored)](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part839.htm)
- [Fusion Reference Manual 18.6: Merge 3D [3Mg] (Blackmagic Design, mirrored)](https://www.steakunderwater.com/VFXPedia/__man/Fusion18-6/Fusion18_Manual_files/part835.htm)
- [Spot Light Node, by Justin Robinson (JayAreTV)](https://jayaretv.com/fusion/spot-light-node/)
- [Point Light Node, by Justin Robinson (JayAreTV)](https://jayaretv.com/fusion/point-light-node/)
- [Renderer 3D Node, by Justin Robinson (JayAreTV)](https://jayaretv.com/fusion/renderer-3d-node/)
- [Ambient Occlusion Node, by Justin Robinson (JayAreTV)](https://jayaretv.com/fusion/ambient-occlusion-node/)
- [Blinn Node, by Justin Robinson (JayAreTV)](https://jayaretv.com/fusion/blinn-node/)
- [Understanding Materials, VFXstudy](https://vfxstudy.com/tutorials/materials/)
- [Blackmagic Fusion: The 3D Workspace, by Bryan Ray](https://bryanray.name/2017/08/08/blackmagic-fusion-the-3d-workspace/)
- [Fusion 21, Virtual Reality and 3D (Blackmagic Design product page)](https://www.blackmagicdesign.com/products/fusion/vr3danimation)
- [DaVinci Resolve product comparison and pricing (Blackmagic Design)](https://www.blackmagicdesign.com/products/davinciresolve/compare)
- [In Depth: DaVinci Resolve Studio vs Free, Updated for 21 (Toolfarm)](https://www.toolfarm.com/tutorial/in-depth-davinci-resolve-studio-vs-the-free-version/)