Our VR Work
Hong Kong Metaverse VR. A multiplayer cross-platform VR and PC environment where users share one space: basketball, tennis, bowling, card and table games, and a cinema. Under it, physical hand interaction, voice chat, character customization, AI-driven NPCs and an in-game browser. Cross-platform means a headset user and a desktop user occupy the same room, which is a networking problem before it’s a design one. See the project.
Rehabilitation VR with Hacettepe University. 2D and VR games built with a Physical Therapy and Rehabilitation department. Clinical VR has a constraint consumer VR doesn’t: the user may be in pain, may have limited range of motion, and may be doing this because a physiotherapist told them to rather than because they want to. Everything about comfort and interaction design changes. See the project.
VR Porting: What It Actually Involves
If you have a working PC or console game and want it on headset, this is the section that matters.
Remapping input is maybe a tenth of the job. The rest is this:
The camera belongs to a person now. Anything that moves it without their consent makes them ill. Cutscenes, camera shake, forced turns, scripted vehicle sequences – every one of them needs re-authoring or removing.
Your UI can’t be on the screen, because there isn’t one. A screen-space HUD becomes world-anchored panels, wrist-mounted displays, or diegetic elements built into the fiction. Text that reads perfectly on a monitor is often illegible at headset resolution, so type sizing gets redone from scratch.
Interaction gets rebuilt, not translated. A keypress becomes a hand reaching, grabbing, with physics and haptic response. Everything a player could do at arm’s length now needs to work when their arm is actually there.
Level geometry stops lying. Environments authored for a camera floating at eye height look wrong when someone is genuinely standing in them. Door frames, ceiling heights, prop scale, stair risers – all of it gets checked against real human dimensions, and a surprising amount of it is wrong.
The frame budget resets to zero. Quest 3 is a mobile chipset filling two eyes at 72Hz or better. If you’re porting from PC, treat it as a mobile target with double the pixels – closer to our mobile performance work than to your existing PC build. The same mindset we bring to porting between conventional platforms applies, just with far less headroom.
We scope ports as their own project with their own quote, because pricing a port as a discount on new development is how both sides end up unhappy in month three.
Comfort Is an Engineering Requirement
In VR, frame rate isn’t a quality setting. Drop below the headset’s refresh rate and a measurable share of your users feel sick within two minutes and never open the app again.
That shapes decisions early rather than late: locomotion options rather than one imposed scheme, teleport alongside smooth movement, vignetting during motion, and no acceleration curve the player didn’t initiate. We give users the choice instead of betting on the assumption that they have VR legs, because a good number of them are in a headset for the first time in their lives.
Session length gets designed for too. VR sessions are short, headsets are heavy and eyes get tired, so ten to twenty focused minutes usually beats a design that assumes an hour.
Multiplayer and Shared Spaces
Networked VR is where our two specialities meet. Photon handles the transport, the same stack behind our real-time multiplayer work, but VR adds problems flat multiplayer doesn’t have: hand and head pose to synchronize convincingly, physical object handoffs between players, spatial voice chat, avatar systems, and NPCs that hold up when a user walks right up to them and stares.
Cross-platform sessions, headset users and desktop users in one space, need deciding at architecture time. It cannot be added later without rebuilding the interaction layer.
Headsets and Platforms
Meta Quest 3, 3S, 2 and Pro. PC VR through SteamVR and OpenXR. PSVR2 and PICO where a project calls for it. Desktop and browser builds alongside the headset version when the audience needs a route in without hardware, often the case for educational and institutional projects.
On store submission: Meta’s Virtual Reality Checks are where builds get rejected, and most of them are performance and comfort requirements rather than content ones. We prepare for those before submission rather than after the first rejection.
Training and Simulation XR
A meaningful part of our XR work isn’t games at all: education, healthcare, and industrial training, where the measure of success is whether someone can do something afterwards. We build VR training scenarios as interactive experiences for brands and training teams.
That work runs the way our educational game development does; learning outcomes fixed with subject experts before any mechanic is designed, because a simulation that’s enjoyable and teaches nothing has failed at the only thing it was commissioned to do.
Engine and Stack
Unity with OpenXR, Photon for networking, and integrations for avatar and AI NPC systems where a project needs them. One engine, deliberately.
Full process, from scoping through handover, is the same as our full-cycle game development: a written scope in two business days, sprints with builds you can put on your head, and full IP transfer at the end.
Common Questions
Usually yes. Send us the build and the project, and we’ll come back with an honest assessment; including when the answer is that the design doesn’t survive the transition, which does happen.
A focused training simulation is measured in weeks; a multiplayer VR product in months. Our general timeline notes apply, with the caveat that comfort testing adds a pass most projects forget to schedule.
Yes, the EU-funded simulation and the university rehabilitation work are both in that category. More about the studio, and our FAQ covers IP and NDAs.
Start a Project
Tell us the headset, the audience, and whether you’re starting fresh or bringing something across. If it’s a port, a link to the existing build is the fastest possible brief.
Scope, team composition and a fixed quote come back within two business days.