Built for High-End, Played on Mainstream PCs: Should VFX Be Designed for Low Settings First?

Paulina Miciulevičiūtė creates real-time VFX for games, with a particular interest in environmental effects, natural phenomena, and procedural systems.

Before moving into VFX production, she worked in QA and content at Unity. This gives her an unusual perspective on both sides of an effect: how it is created and how it actually performs in the game.

The idea for this interview came from one of Paulina’s recent posts. She raised a simple but important problem: artists can spend a lot of time making a world feel alive, while players may switch to Low settings and never see much of that work.

Main Topics:

• Game VFX Optimization
• VFX for Low-End Hardware
• Graphics Quality Settings
• Real-Time VFX Performance
• VFX Production Workflow
• VFX Portfolio Tips

Interview

With Paulina Miciulevičiūtė, Real-Time VFX Artist whose work includes VALORANT, Invincible VS, and Battle Shapers

Q: You started in QA at Unity before moving into real-time VFX. Did testing games before creating effects change the way you think about performance and what players actually experience?

Did your QA background make you notice technical problems, edge cases, or production limitations that artists sometimes overlook?

I didn’t test games themselves for the most part, my work was focused primarily on the Unity engine. We would sometimes create small games ourselves to emulate the development experience, and I also had opportunities to test client projects, although those were usually fairly isolated cases focused on reproducing specific bugs.

I think the biggest thing my QA background gave me was a much better understanding of what is actually happening under the hood of the engine and how different systems work together. Once you have that foundation, you naturally start avoiding some of the biggest performance pitfalls in visual effects.

A lot of people learn that a certain technique is “expensive,” but don’t necessarily understand why. Without that deeper understanding, it can be easy to end up with questionable workarounds or solutions. For me, knowing the reasoning behind those limitations has made it much easier to make better technical decisions as an artist.

Q: In your recent post, you raised a rare and very important issue: an artist can spend a lot of time creating atmospheric VFX, only for the player to switch to Low and never see them. Should VFX be designed for Low settings first?

Or should the High version still define the visual goal, with the team then finding ways to preserve it on weaker hardware?

I think effects should generally be designed for the highest quality first, because that version defines the visual goal. From there, it is much easier to identify which elements are essential to the effect and which ones can be reduced or removed for lower settings.

Of course, some players will turn effects down or disable certain features entirely, but they are not the only players we are designing for. We still want the game to look as good as possible on higher-end hardware. The important part is making sure that when the effect is simplified for Low settings, it still preserves its core purpose, readability, and atmosphere.

Q: What absolutely needs to remain when a player lowers the graphics settings?

How do you decide which effects support atmosphere, readability, and gameplay feedback, and which can be removed without a noticeable loss?

The main thing that needs to remain is the core of the effect, what we often call the “body.” It is the element that communicates the gameplay feedback or simply makes the effect recognizable in the first place. If you remove it and the player can no longer understand what the effect is supposed to represent, then you have stripped away too much.

At the same time, it is always a balance. You want to simplify the effect enough to improve performance, but not so much that you break the atmosphere, immersion, or overall player experience.

Q: How do you create a cheaper version of an effect instead of simply turning it off?

For example, what can be changed in the particles, shaders, lighting, collisions, fog, or distortion to preserve the feeling of the effect at a lower performance cost?

There are a lot of techniques you can use. You can swap transparent shaders for opaque ones where possible, reduce draw calls, optimize textures and mesh density, move particles from the CPU to the GPU when appropriate, or build quality switches directly into the shader so that more expensive features can be disabled on lower settings.

There are also more specialized approaches. In Unreal, for example, Custom Primitive Data can allow you to use the same material across multiple effects while still giving them different parameters, rather than creating separate material instances for every variation. And if you are feeling a little more adventurous, you can even recreate certain particle-system behaviours directly inside a shader and use stacked billboards to represent the particles.

There are many more techniques, and optimization is a topic you could discuss for hours. But I think one of the most important things to understand is when to stop. At some point, optimization can become counterproductive if you spend too much development time chasing very small performance gains. Development time is also an expensive resource, so optimization always has to be considered in that context.

Q: How does a team check that a game still looks good on Low?

Are tests on minimum-spec hardware and comparisons between quality settings enough? Who decides what needs to be sacrificed when the VFX no longer fit within the performance budget?

When I create an effect, I usually start by building it for the highest quality setting. Once the visual target has been approved, I lower the game quality myself and check which elements still need to remain for the effect to work visually and functionally.

In the teams I have worked with, QA does not really approve the artistic side of the effect itself. Their role is more about catching obvious visual issues, making sure the effect behaves correctly across different settings, and checking that it stays within technical and memory budgets. They might, for example, flag that certain textures are taking up too much memory or that something behaves incorrectly on a lower-spec configuration.

Usually those kinds of issues are relatively straightforward to address. The more subjective decisions, for example, what can be removed, what still needs to be readable, and how much visual quality can be sacrificed, tend to stay closer to the VFX artist, art direction, and the wider team responsible for the performance budget.

For Small Game Studios

If a small studio comes to you for VFX, what should it define in advance so the effects do not need to be completely rebuilt for weaker hardware later?

For example, target hardware, the performance budget, differences between quality settings, and which feelings or gameplay information must remain even on Low.

Ideally, a small studio should know which platforms and hardware it plans to target from the beginning. That said, from my own experience, I have also learned to build VFX in a way that gives them some flexibility across different platforms, quality settings, and camera angles. There have been too many situations where, halfway through development, someone suddenly decides that releasing on Switch is the way to go 😄

Because of that, I try not to build effects around a single very specific hardware target unless there is a good reason to do so.

Interestingly, I think some of the biggest production challenges do not always come from optimization itself. Quite often, the bigger issue is that the studio has not fully decided what visual direction it wants to take. If the visual target keeps changing, you can end up rebuilding effects regardless of how well optimized they were in the first place. Having a clear idea of the desired look, the most important gameplay information, and the intended target platforms early on can save a lot of work later.

For Young VFX Artists

Should a portfolio show not only a beautiful effect on maximum settings, but also how it performs on weaker hardware?

What demonstrates production readiness better: the final beauty shot, or the ability to create several versions of an effect, measure their performance cost, and explain the decisions behind them?

For a portfolio, I would definitely prioritize the visuals first. You want to create that immediate “wow” effect when someone looks at your work, because that is usually what gets their attention in the first place.

At the same time, I think including a short technical breakdown can make a portfolio much stronger. Showing how the effect was built, which optimization techniques you used, or how you adapted it for different performance targets can demonstrate that you are not only able to create something beautiful, but also understand how to make it work in a real production environment.

So I would not necessarily lead with performance, but I would absolutely use it as an additional layer that helps the portfolio stand out.