Describe the Houdini side: how you constrained the geometry, managed the sim resolution, and baked the result. Then cover the engine side: how you brought it in as a flipbook or geometry cache, what draw call and memory cost it had, and what you cut or simplified to hit the performance target. Studios want to see you can own the full pipeline from sim to shipping.

Start with the performance budget before you start the effect. Know the draw call limit, the texture memory available, and the GPU time you’re working within. Design the effect in layers so you can cut the most expensive elements without losing the read. The final result should hit the target on the lowest spec machine, not just on your dev kit.

Lead with your best work — not your oldest or biggest. The first effect should show real-time work in engine with clear visual reading. Label each clip with the engine and whether it is in-game or pre-rendered. Include at least one breakdown showing layers separately. Reviewers look for range, technical depth, and that the work is actually shippable.

Define budgets early and get them in writing. Track draw calls, GPU time, and texture memory per effect type. When you exceed a budget, escalate with the data rather than assumptions. Build a test scene that mimics the worst-case scenario so engineers can see the actual impact. The best FX artists treat performance as a design constraint from day one.

Cover your voronoi setup, constraint network approach, and how you handle art direction after simulation. Describe how you balance procedural control with the art director’s visual requirements — studios want simulations that look intentional, not physically accurate. Talk about any custom HDAs or tools you built to speed up iteration.

Set up a shared timing reference early. The effect has to serve the animation, not compete with it. Define the anticipation, impact, and recovery frames together before you build anything. Review in engine on the actual character, not in isolation. When timing conflicts come up, the animation is usually the anchor — adapt the effect to complement it.

Flipbooks are best for high-complexity, one-off effects where the visual result justifies the texture memory cost and you can pre-bake the simulation. Real-time simulation is better for interactive effects that need to respond to gameplay state — hit directions, wind, character proximity. The decision is usually budget: if you have the texture budget and the effect doesn’t need to react, bake it.

Translate their reference and language into technical parameters. “More punchy” means shorter timing, harder edges, more contrast at peak. “Feels cheap” often means the loop is too obvious or the particle count is too low. Show two or three variations quickly rather than building one perfect version from the feedback. Rapid iteration from ambiguous direction is the most valuable skill you can demonstrate.

Base estimates on your most recent comparable work, not your fastest ever. Add time for engine integration, optimization, and revision rounds — these often take as long as the initial build. What makes it hard: unclear performance targets, undecided animation timing, and mid-brief art direction changes. Get those locked down before committing to a timeline.

Document parameters exposed in the emitter that artists can tweak without breaking the effect. Use clear naming conventions in your material graph. Version control everything and note what changed between iterations. The goal is for another artist to clone your effect, change the color and scale, and have it work immediately without digging into your graph.

Mesh VFX is better for effects that need a consistent silhouette — magical shields, plasma spheres, specific destruction shapes. Particle-based is better for diffuse, volumetric reads — smoke, fire, dust. Mesh is usually cheaper on draw calls but more expensive on geometry. For hero effects, a combination of mesh core with particle fill often reads best across distances.

Walk through what made it hard: the performance target was tight, the brief kept changing, the animation had unusual timing, or the engine had a limitation you had to work around. What you solved, how you solved it, and what you would do differently. Honesty about the tradeoffs you made under production pressure shows more experience than a polished success story.

Houdini is the industry standard for FX work in AAA games. Most studios expect proficiency in Houdini for simulation, procedural generation, and technical FX pipelines. Unreal Engine and Unity are required for implementing real-time VFX. Substance Designer is commonly used for particle texture creation. Nuke and Fusion are used in studios with film VFX crossover.

FX artist salaries range from $65K-$95K for mid-level roles at game studios to $100K-$160K+ at AAA studios. Houdini specialists with strong simulation experience are at the higher end. Vancouver and London studios often pay in CAD/GBP with competitive rates. Freelance FX artists in film VFX typically earn $400-$800 per day.

VFX (visual effects) is the broader discipline covering any in-game visual effect including particles, simulations, shaders, and post-process work. FX specifically refers to the simulation and particle systems side — explosions, fire, water, cloth, destruction. At large studios these may be separate roles. At smaller studios, VFX artist and FX artist are often the same job.

Show real-time VFX working in engine — not just rendered flipbooks. Include a range of effect types: fire, water, smoke, magic, destruction, environmental. Show the Houdini graph or node setup for at least one piece to demonstrate technical depth. Break down one effect showing the layers and how performance was managed.

Know your Houdini node graph well enough to explain it out loud. Prepare a breakdown of one complex effect showing simulation setup, optimization decisions, and the real-time implementation. Research the studio’s game and be ready to discuss what FX systems they use and how you would approach a specific effect in their engine.