Adobe · Spatial research · 2024
Designing for the body
Hands-on research on how the body shapes spatial video editing: five headsets, a market benchmark, a team alignment exercise and accessibility work.
- Role
- XR domain expert
- When
- Summer 2024, Adobe
- Tested with
- Five headsets, led by Apple Vision Pro
- For
- The Project Pulsar team
Gucci app on Apple Vision Pro, shown as third-party research. All rights reserved to Gucci.
Overview
In summer 2024, I was a graduate product design intern at Adobe on Project Pulsar, a 3D and spatial video editor then taking shape, which later launched on Samsung Galaxy XR.
As the team's XR domain expert, I spent the summer in headsets and saw gaps in what the physical experience asks of people while they craft video in headset. I took the initiative to conduct research that filled those gaps and explored accessibility and product strategy.
The problem
Headsets make the body the controller, and the body has limits.
New headsets lean on native gestural control, so hands and eyes do the work a controller used to do. The same devices also carry symptoms with real bodily consequences, which matter for both accessibility and comfort.
The body is the input. Physical symptoms become the output.
Research
Five headsets showed that eye position and micro-gestures matter most.
I tested hand and eye tracking on Apple Vision Pro, the main device, then on HoloLens 2, Quest 3, Quest Pro and Magic Leap. I compared the operating systems, passthrough, the absence of hardware controls, tethered battery packs and the gestures each headset expects.
Apple Vision Pro
Apple · visionOS · Main device
Hand and eye tracking carry the input, with voice available. There are no hardware controls, though external inputs can connect. Passthrough keeps you in your surroundings, and a tethered battery pack adds weight.
HoloLens 2
Microsoft · Windows Holographic
No hardware controls, so native hand gestures do the work: touching holograms directly, a hand ray with air tap and hold for far targets, and two-handed grabs to move, resize and rotate. Its see-through lenses show your surroundings directly, which gives it some of the strongest augmented reality of the set, alongside Vision Pro. Hand tracking only works inside a frame a few feet either side of you.
Quest 3
Meta · Horizon OS
Hardware first: Touch Plus controllers are the main input, with TruTouch haptics, and hand tracking is there when you put them down. Color passthrough shows your room.
Quest Pro
Meta · Horizon OS
Also controller-led, with hand tracking available. Touch Pro controllers carry three haptic motors and track themselves with built-in cameras, so they stay accurate even out of the headset's view. Eye and face tracking are built in, and its color passthrough is lower quality than Quest 3's.
Magic Leap
Magic Leap · Magic Leap OS
A limited field of view, a tether to a waist-worn compute pack and a single hand controller that I found hard to navigate with. The UI also lagged in responsiveness. I rated it inconclusive and focused my headset research on the others.
Eye position turned out to be the key factor in how well input worked. Micro-gestures helped reduce fatigue.
The Vision Pro gestures I tested:
- Tap
- Double tap
- Pinch and hold
- Pinch and drag
- Zoom
- Rotate
Process
Recording myself in the headset let the team watch the task step by step.
I filmed myself replicating the core Pulsar task, adding an object to spatial video. The team could watch each step and spot where problems might arise.
Methods
Using Apple's immersive video as a new user showed where people get stuck.
I screen-recorded myself trying gestures in Apple's immersive video mode as a new user. Anywhere I experienced difficulty, I noted it as an opportunity for Pulsar to improve the spatial video editing experience.
I also looked at a third-party app, Explore POV, which plays 180° immersive video. It served as an example of well-designed immersive video playback.
Alignment
The team saw the app as low effort and casual; I saw it as high effort and more technical.
I ran an informal stakeholder poll. I asked teammates to place Pulsar on two scales, casual to technical and low to high effort, then added my own placement. The gap showed where the team's picture of the product and mine differed.
I placed Pulsar higher effort and more technical based on external research into headset ownership, onboarding, and the complexity of video editing. Video editing is hard before you add a spatial layer, so I projected a bigger lift for new users to get started and get comfortable with Pulsar's interactions.
I used this as part of the foundation to push for additional use cases and workflows for both power users and casual creators.
Market research
I broke down Gucci's Vision Pro video as third-party market research.
I studied Gucci's immersive Vision Pro story as a benchmark for what well-made spatial video can do, noting where the video breaks the frame and where it invites the viewer to interact.
Gucci app on Apple Vision Pro, shown as third-party research. All rights reserved to Gucci.
The Gucci app provided inspiration on how to make spatial video more digestible for new users, so spatial storytelling would be a pleasure for both creators and viewers. I gave suggestions on how I felt certain interactions could convert.
Outcome
I closed with recommendations on comfort, input and audience.
The deck ended with recommendations for the team on how long people can comfortably stay in a headset, how they control it, and who else the product could serve. Those stay with the team. I wrote them to help shape Pulsar long after handoff.
Sources
19 sources · Show
Research papers
- The effects of target location on musculoskeletal load, task performance, and subjective discomfort during virtual reality interactionsPenumudi et al. · Applied Ergonomics, 2020
- Consumed Endurance: A metric to quantify arm fatigue of mid-air interactionsHincapié-Ramos et al. · CHI 2014
- Understanding, addressing, and analysing digital eye strain in virtual reality head-mounted displaysHirzle et al. · ACM Transactions on Computer-Human Interaction, 2022
- Toward optimized VR/AR ergonomics: modeling and predicting user neck muscle contractionarXiv preprint, 2023
- Cybersickness in current-generation virtual reality head-mounted displays: systematic review and outlookVirtual Reality (journal)
- Accessible by design: an opportunity for virtual realityMott et al. · ISMAR 2019 Workshop on Mixed Reality and Accessibility
- SeeingVR: a set of tools to make virtual reality more accessible to people with low visionCHI 2019
Articles and news
- 'VR is not the next smartphone': Jeremy Bailenson highlights VR research, and how to evaluate use of immersive techCenter for Academic Innovation, University of Michigan
- Magic Leap is Google's new mystery partner for XR headsetsThe Verge, 2024
- These gestures are how you control Apple Vision ProMacRumors, 2023
- What happens to your body when you've been in virtual reality for too longBusiness Insider, 2018
Headset details
- Getting around HoloLens 2Microsoft Learn
- Compare Meta Quest devicesMeta for Developers
- Quest Touch Pro controllers: haptics and index trackingUploadVR
- Meta Quest 3 comes with new controllers and support for Touch ProMixed News
- Meta Quest 3 vs Meta Quest ProTechRadar
- Meta shows first glimpse of Quest 3 mixed reality gameplayRoad to VR
- Magic Leap 2 specs revealedUploadVR
- HoloLens 2 vs Magic Leap 2: enterprise AR comparedReality Atlas
© 2026 Faith Gowen. All rights reserved.