Blender Quiz: Use Render Passes Aovs Conceptually Intro
10 questions · exam conditions
0:00
Use Render Passes Aovs Conceptually IntroQuestion 1 of 10

A compositor receives a Cycles render with Diffuse Direct, Diffuse Indirect, and Diffuse Color passes. The art director wants to replace the surface's red diffuse color with blue while retaining the original direct and bounced-light patterns.

Which compositing approach best accomplishes the requested change?

Add Diffuse Direct and Diffuse Indirect, then multiply the result by the replacement blue color.
Multiply the Combined pass by the replacement blue color, then add the unchanged Diffuse Color pass.
Multiply only Diffuse Direct by the replacement blue color, then add the original Diffuse Indirect pass.
Divide Diffuse Direct and Diffuse Indirect by the replacement blue color, then multiply by Diffuse Color.
← Back to quizzes

Blender Quiz

Blender Quiz: Use Render Passes Aovs Conceptually Intro

Practice Use Render Passes Aovs Conceptually Intro in Blender with focused quiz questions that help you check what you know, review explanations, and build confidence with test-style prompts.

What this quiz covers

This quiz focuses on Use Render Passes Aovs Conceptually Intro, giving you a quick way to practice the rules, question types, and explanations that matter most for Blender.

How to use this quiz

Try each quiz question before looking at the correct answer. Use the explanations to review missed ideas, then come back to similar questions until the pattern feels familiar.

All questions

Question 1

A compositor receives a Cycles render with Diffuse Direct, Diffuse Indirect, and Diffuse Color passes. The art director wants to replace the surface's red diffuse color with blue while retaining the original direct and bounced-light patterns.

Which compositing approach best accomplishes the requested change?

  1. Add Diffuse Direct and Diffuse Indirect, then multiply the result by the replacement blue color. (correct answer)
  2. Multiply the Combined pass by the replacement blue color, then add the unchanged Diffuse Color pass.
  3. Multiply only Diffuse Direct by the replacement blue color, then add the original Diffuse Indirect pass.
  4. Divide Diffuse Direct and Diffuse Indirect by the replacement blue color, then multiply by Diffuse Color.
Explanation: When working with Cycles render passes, you need to understand how Blender decomposes a diffuse surface into its components. The key relationship is: Diffuse=(Diffuse Direct+Diffuse Indirect)×Diffuse Color\text{Diffuse} = (\text{Diffuse Direct} + \text{Diffuse Indirect}) \times \text{Diffuse Color} This means Blender stores the raw lighting information separately from the surface color, allowing compositors to manipulate them independently. When the art director asks you to swap the surface color while keeping the lighting patterns intact, your goal is to reconstruct this equation using a new color value. Answer A is correct because it isolates the pure lighting data — the direct and bounced light intensities — and multiplies by the replacement blue color. This perfectly mirrors the render equation, producing a result identical to what Cycles would have rendered had the surface always been blue. The lighting patterns are fully preserved. Answer B is flawed because the Combined pass already has the original red color baked into it. Multiplying by blue would tint an already-colored image, distorting brightness and producing a muddy, incorrect result — not a clean color swap. Answer C only recolors the direct lighting, leaving the indirect (bounced) light unmodified. This means shadowed and bounce-lit areas would retain the old red influence, creating an inconsistent, physically incorrect appearance. Answer D is backwards and physically meaningless. Dividing lighting passes by a color value doesn't "remove" the original color from anything — the Diffuse Direct and Indirect passes contain no color information to begin with, so this operation only corrupts the data. A good study habit: memorize the render pass equation as a formula. When a question involves pass recombination, ask yourself whether each step mirrors that equation exactly.

Question 2

A scene must receive a compositor-generated atmospheric haze. The haze should transition smoothly at object silhouettes, and the artist only needs a normalized near-to-far gradient rather than exact camera-space distances.

Which pass is the most appropriate source for the haze factor, and what tradeoff does it have?

  1. Use the Mist pass; it provides an antialiased normalized gradient but does not store unrestricted physical distances. (correct answer)
  2. Use the Z pass; it stores raw camera-space depth but is not antialiased and does not natively normalize to a haze range.
  3. Use the Normal pass; its camera-facing component directly encodes normalized distance with antialiased silhouettes.
  4. Use the Object Index pass; its integer values yield smooth depth interpolation when filtered in the compositor.
Explanation: When compositing atmospheric effects like haze, you need to think carefully about what each render pass actually stores and how it behaves at object edges — both qualities matter for a clean result. The Mist pass is purpose-built for exactly this scenario. It outputs a normalized 0–1 gradient controlled by the Mist settings in the World properties (near distance, far distance, and falloff type). Crucially, it's computed as part of the color pipeline, so it receives antialiasing treatment — silhouette edges blend smoothly rather than producing jagged transitions. The honest tradeoff is that it sacrifices raw physical accuracy: you're working with a remapped, clamped range rather than true camera-space depth values. For a haze effect, that's a worthwhile exchange, making A the correct answer. B is tempting but flawed. The Z pass does store real camera-space distances in scene units, which sounds useful — but it isn't antialiased, causing harsh stairstepping at edges, and you'd need manual normalization nodes to convert it into a usable 0–1 haze factor. It's better suited for depth-of-field or metric distance work, not smooth atmospheric compositing. C is simply a misunderstanding of the Normal pass. It encodes surface orientation (X, Y, Z direction vectors), not distance from camera. No component of it represents depth in any meaningful way. D is incorrect because the Object Index pass stores integer ID values per object — these hard boundaries produce no smooth interpolation at all, even with filtering. Filtering blends the integers, creating meaningless fractional IDs, not gradients. Your study tip: remember "Mist = normalized + antialiased, Z = raw + aliased." When a question mentions silhouette smoothness and normalized ranges together, Mist is almost always the intended answer.

Question 3

A mesh uses three material slots. A Color AOV named Wetness is added to the View Layer, but only one of the three materials contains an AOV Output node writing to that exact name. The resulting Wetness pass contains valid color over some polygons and black over the others.

What is the most likely explanation and appropriate correction?

  1. The AOV evaluates per material, so each relevant material must write the matching AOV name and compatible data type. (correct answer)
  2. The AOV evaluates per object, so the mesh must be separated into three objects before its values can differ.
  3. The AOV reads vertex colors automatically, so the missing polygons need a color attribute named Wetness.
  4. The AOV is blocked by multiple material slots, so one shared material must replace all three existing materials.
Explanation: When working with AOVs (Arbitrary Output Variables) in Blender's Cycles, think of each AOV as a named data channel that materials must explicitly write into. AOVs don't automatically gather data from objects or meshes — they rely on individual materials containing an AOV Output node with the exact matching name and data type. This is precisely why A is correct. Since a mesh can have multiple material slots assigned to different polygons, each material renders its own polygons independently. If only one material contains an AOV Output node named Wetness, only that material's polygons contribute color data to the pass. The remaining polygons, rendered by the other two materials, write nothing — defaulting to black. The fix is straightforward: add an AOV Output node named Wetness with a compatible Color type to every material that should contribute to that pass. B is wrong because AOVs do not evaluate per object — a single mesh with multiple materials works fine for AOV output, provided each material writes the AOV. Splitting objects would be unnecessary and wouldn't solve the underlying problem. C is wrong because AOVs are not connected to vertex color attributes in any automatic way. Vertex colors and AOVs are entirely separate systems; naming a color attribute Wetness has no effect on the AOV pass. D is wrong because multiple material slots don't block or interfere with AOV evaluation. The issue isn't slot count — it's missing AOV Output nodes in those slots' materials. A useful rule of thumb: whenever an AOV pass appears partially black, your first instinct should be to audit every material on the mesh and confirm each one has a correctly named AOV Output node.

Question 4

A noisy Cycles render will be denoised inside Blender's compositor rather than during rendering. The compositor currently receives only the Combined, ordinary Normal, and Z outputs. Fine texture detail is being mistaken for noise, and some edges are being smeared.

Which change gives the Denoise node the most appropriate guidance?

  1. Enable Denoising Data and connect Combined as the noisy image with Denoising Normal and Denoising Albedo as guides. (correct answer)
  2. Increase the Z pass precision and connect Z as the normal guide while using ordinary Normal as the albedo guide.
  3. Connect Combined to every Denoise input so all three inputs share identical color and edge information.
  4. Enable Object Index and Material Index, then connect those identifier passes as the two denoising guides.
Explanation: When using Blender's compositor Denoise node, you need to understand that it relies on three distinct inputs: a noisy image, a normal guide, and an albedo guide. The normal guide tells the denoiser about surface orientation so it preserves edges, while the albedo guide describes surface color/texture so the denoiser doesn't mistake fine detail for noise. Standard render passes aren't precise enough for this purpose — Cycles generates specialized Denoising Data passes specifically calibrated for this task. This is exactly why A is correct. Enabling Denoising Data exposes two purpose-built passes — Denoising Normal and Denoising Albedo — that encode geometry and color information at the precision the denoiser expects. Connecting Combined as the noisy image with these two dedicated guides directly solves both problems described: smeared edges (fixed by accurate normals) and lost texture detail (fixed by accurate albedo). B is wrong because Z is a depth pass, not directional surface data — using it as a normal guide gives the denoiser completely meaningless orientation information. C fails because feeding identical data into all three inputs defeats the entire purpose; the denoiser needs different types of information from each input to function correctly. D is wrong because Object Index and Material Index are identifier passes used for masking and selection workflows — they carry no surface normal or albedo information whatsoever. A useful tip: whenever a question mentions artifacts like smeared edges or lost texture detail in compositor denoising, immediately think about whether the correct guide passes are being used. Blender's Denoising Normal and Denoising Albedo exist precisely to solve those two artifact types respectively.

Question 5

An artist wants to adjust motion blur strength after rendering. The scene is rendered without baked-in motion blur, and the compositor will use Blender's Vector Blur node.

Which pass setup best supports this workflow?

  1. Render the Normal and Mist passes, then interpret orientation and normalized haze as per-pixel screen velocity.
  2. Render the Vector and Z passes, then supply the beauty image, motion vectors, and depth to the Vector Blur node. (correct answer)
  3. Render motion blur into Combined and also use the Vector pass, allowing the Vector Blur node to reconstruct sharp pixels.
  4. Render Object Index and Z, then derive motion direction from changes in each object's integer identifier.
Explanation: When working with Blender's compositor for post-render motion blur, you need to think about what data the Vector Blur node actually requires: a beauty (color) image, per-pixel motion vectors, and depth information. Understanding these inputs makes the correct answer obvious. The Vector Blur node reconstructs motion blur mathematically after rendering by reading how each pixel moved between frames (the Vector pass) and using depth (the Z pass) to handle occlusion correctly. This means you render a clean, sharp image and let the compositor simulate blur with full artistic control over strength, samples, and curve — exactly what the scenario describes. That's why B is correct: the Vector pass provides screen-space velocity per pixel, Z provides depth ordering, and the combined beauty image provides the color to blur. A is wrong because Normal and Mist passes encode surface orientation and fog density, respectively — neither contains motion velocity information. Treating them as velocity data would produce meaningless results. C contains a fundamental contradiction: if you bake motion blur into the Combined pass during rendering, the pixels are already blurred. The Vector Blur node cannot "reconstruct" sharpness from blurred source data, and you'd also be double-blurring. D is wrong because Object Index stores integer object identifiers for masking purposes, not motion direction. You cannot derive per-pixel velocity from a static integer ID — it carries no frame-to-frame displacement information. A useful rule of thumb: whenever a question mentions post-render motion blur in Blender's compositor, immediately think Vector pass + Z pass + clean render. Those three inputs are the canonical recipe for the Vector Blur node.

Question 6

A Cycles scene has a key light assigned to its own light group. After rendering, the compositor must reduce only that key light's contribution while preserving the fill lights, world lighting, and emissive surfaces. The Combined pass still contains the original key-light contribution.

Which compositing operation is conceptually correct?

  1. Replace Combined with the reduced key-group pass because that pass also contains all ungrouped scene lighting.
  2. Multiply the entire Combined pass by the reduced key-group pass so non-key illumination remains numerically unchanged.
  3. Add a reduced key-group pass directly to Combined because light-group contributions are excluded from Combined.
  4. Remove the original key-group contribution from Combined, then add back a reduced version of that group pass. (correct answer)
Explanation: When working with Cycles light groups in the compositor, think about the mathematical relationship between the Combined pass and individual light group passes. The Combined pass already bakes in every light group's full contribution at render time — light groups are not excluded from Combined; they are simply also available as separate, isolated passes. To selectively reduce one group, you need to perform a substitution: strip out the original key-group contribution from Combined, then reintroduce a scaled-down version of that same pass. Symbolically: Result=(Combined−KeyGroup)+(KeyGroup×reduction factor)\text{Result} = (\text{Combined} - \text{KeyGroup}) + (\text{KeyGroup} \times \text{reduction factor}). This is exactly what D describes, and it correctly preserves fill lights, world lighting, and emissive surfaces because those components remain untouched inside Combined. A is wrong because the key-group pass contains only that group's light — it does not include fill lights, world, or emissives. Replacing Combined with it would destroy all other lighting entirely. B is wrong because multiplying the whole Combined pass by the key-group pass would corrupt every other lighting contribution. Non-key pixels would be multiplied by near-zero values, darkening or eliminating fill and world light incorrectly. C is wrong because it assumes light groups are excluded from Combined — they are not. Simply adding a reduced pass on top of Combined would result in double-counting: the original full key contribution stays in Combined plus you're adding more on top of it. A useful habit: whenever you need to swap out one component of a composite signal, always remember the remove-then-replace pattern — subtract the original, add the modified version.

Question 7

A scene uses a bright HDRI world for illumination and reflections. The artist enables Film > Transparent so the Combined pass can be placed over live-action footage, but expects the objects to stop receiving all HDRI influence.

What should the artist expect from the render?

  1. The world remains visible in Combined, but its lighting is moved entirely into the Environment pass.
  2. The world becomes transparent and is also removed from indirect lighting, glossy reflections, and transmission.
  3. Only the Combined RGB becomes transparent; its alpha remains opaque because world lighting is still enabled.
  4. The directly visible world becomes transparent, while its lighting and reflections can still affect scene objects. (correct answer)
Explanation: Whenever you see a question about Blender's Film > Transparent setting, the key distinction to hold in mind is: visibility versus influence. These are two separate channels, and the setting only controls one of them. Enabling Film > Transparent makes the world background itself invisible in the render — the Combined pass shows a transparent alpha where the sky would normally appear, allowing you to composite over live-action footage. However, the HDRI world continues to function as a light source and reflection environment for every object in the scene. Objects still receive indirect illumination, glossy surfaces still reflect the HDRI, and transmission materials still refract it. The world stops being seen, but it doesn't stop working. That's exactly what D describes, making it the correct answer. A is wrong because no "Environment pass" magically absorbs the lighting contribution — the lighting stays active in the Combined pass itself, baked into the shading of each object. B describes a setting that doesn't exist in Blender; there is no single toggle that simultaneously hides the background and strips the HDRI of all its lighting, reflection, and transmission influence. That would require manually zeroing out world strength or using Light Path nodes. C inverts the actual behavior — the alpha channel does become transparent (that's the whole point of the setting), while it's the lighting influence, not the alpha, that remains fully active. As a study tip: always separate "what the camera sees" from "what affects shading" in Blender — many settings, including Film > Transparent and ray visibility toggles, operate on only one of those layers at a time.

Question 8

A product render includes a UV pass. After rendering, the client asks to replace a printed label using compositing, without rerendering the scene. The original label also affected roughness and appeared in curved reflections on nearby objects.

Which result is realistically achievable with the UV pass alone?

  1. The label's roughness response can be recomputed accurately because UV coordinates implicitly encode normals, depth, and light transport data.
  2. The label and all of its reflected appearances can be replaced because the UV pass stores every secondary ray's texture coordinates.
  3. The visible label pixels can be remapped on the product surface, but dependent roughness, lighting, and reflected appearances in other objects will not be recomputed. (correct answer)
  4. The label can be replaced only on flat surfaces because UV coordinates become invalid on any curved rendered geometry.
Explanation: When working with render passes in Blender's compositor, the key question to always ask is: what information does this pass actually store, and what can it physically reconstruct? A UV pass records the 2D texture coordinates mapped to each rendered pixel — nothing more. It tells you where on a texture each pixel came from, but it carries no information about lighting, surface normals, material roughness, or how light bounced around the scene. That's exactly why C is correct. Using the UV pass, you can remap new texture content onto the product's visible surface pixels — essentially swapping the label in 2D compositing space. However, the roughness variation the original label caused, the way nearby curved objects reflected that label through secondary rays, and any lighting interaction baked into the original render are all gone. Those effects were computed during the render itself and cannot be reconstructed from UV coordinates alone. A is wrong because UV coordinates are purely positional — they describe surface mapping, not normals, depth, or light transport. None of that secondary data is implicitly encoded in a UV pass. B is wrong because secondary reflections are the result of ray tracing during the render; each reflection ray would need its own UV data captured separately, which a standard UV pass does not do. D is wrong because UV coordinates are perfectly valid on curved geometry — that's a core reason UVs exist. Curvature doesn't invalidate UV mapping; it just means you can't recover the curved reflection appearances from the UV pass alone. Remember: each render pass stores exactly one category of data. When a question describes side effects like reflections or roughness, ask whether those effects live in that pass — they almost never do.

Question 9

A studio must archive Combined, depth, normals, Cryptomatte, and custom AOVs for later compositing and color grading. A test export to a display-referred image format clips bright highlights and cannot conveniently retain the heterogeneous passes in one file.

Which output strategy best preserves the render-pass data for later work?

  1. Save a JPEG for each pass after contrast adjustment, then reconstruct highlight values by reversing the adjustment.
  2. Save a standard PNG after the display transform, then infer depth and normals from the visible RGB values.
  3. Save a multilayer OpenEXR with scene-linear high-dynamic-range data, then apply the display transform after compositing. (correct answer)
  4. Save only the view-transformed Combined image because Cryptomatte and AOVs can be regenerated from its alpha channel.
Explanation: When a question asks about preserving render passes for compositing, think about two core requirements: dynamic range and multi-layer storage. Display-referred formats (JPEG, standard PNG) apply a display transform that compresses highlights and bakes color decisions into the data — you can't undo that math later without loss. OpenEXR in scene-linear space solves both problems at once. It stores floating-point values with no upper clipping limit, so bright speculars and HDR lighting data survive intact. Critically, the multilayer variant packs heterogeneous passes — Combined, depth, normals, Cryptomatte, custom AOVs — into a single file with named channels. The display transform is intentionally deferred until the very end of your compositing pipeline, giving you full creative control. That's exactly what C describes, making it the correct strategy. A is wrong because reversing a contrast adjustment after clipping is mathematically impossible — once pixel values are clamped, the original highlight information is gone. No inverse function can reconstruct data that was discarded. B is wrong on two counts: PNG after a display transform inherits the same clipping problem, and inferring depth or normals from visible RGB is not a real technique — those passes carry entirely different data types (linear distance, direction vectors) that have no recoverable relationship to a tone-mapped color image. D is wrong because Cryptomatte and AOVs are not encoded in the Combined image's alpha channel. Cryptomatte uses dedicated float channels for object IDs, and AOVs are separate render outputs entirely — they cannot be regenerated from a finished composite. Your study tip: whenever you see "archive for compositing," the answer almost always involves scene-linear EXR with deferred display transform — that pairing is the industry-standard signal to look for.

Question 10

An animation contains several overlapping characters with motion blur, depth of field, and fine hair edges. The compositor must isolate one character without manually assigning unique grayscale mask values or repairing partially covered edge pixels.

Which render-pass strategy is most suitable?

  1. Enable the Object Index pass and blur the selected integer value until it matches the beauty render's coverage.
  2. Enable Object Cryptomatte and extract the character by its encoded identity and pixel coverage. (correct answer)
  3. Enable the Z pass and isolate the character by selecting its average camera-space depth range.
  4. Enable Material Cryptomatte and select one shared skin material to recover the entire character.
Explanation: When you need to isolate a specific object in compositing — especially with complex effects like motion blur, depth of field, or fine hair — ask yourself: does this method preserve sub-pixel coverage information, or does it treat each pixel as binary? That distinction is the heart of this question. Object Cryptomatte (answer B) is specifically designed for this challenge. It encodes each object's identity across every pixel using a hashing system that stores fractional coverage values. This means a pixel sitting on the edge of a character's hair strand isn't forced into an all-or-nothing decision — Cryptomatte records how much of that pixel belongs to the character. In Blender's compositor, you simply pick the character by name or click it in the preview, and the matte is extracted cleanly, even through motion blur or overlapping geometry. Answer A falls apart because the Object Index pass assigns flat integer IDs — a pixel either belongs to an object or it doesn't. Blurring the mask afterward is a manual workaround that introduces artifacts, especially near fine edges, and defeats the purpose of a clean extraction pipeline. Answer C, the Z pass, records depth rather than identity. Multiple objects can share similar depths, and depth gives you no way to reliably separate two characters standing close together in the scene. Answer D is tempting but flawed: Material Cryptomatte works per-material, so if the character uses multiple materials (skin, eyes, hair, clothing), selecting just the skin material recovers only part of the character — not the whole figure. Remember: whenever a question mentions motion blur, depth of field, or fine edges alongside clean masking, Cryptomatte is almost always the right tool because it's the only pass built around fractional pixel coverage.