Designing for HoloLens: A Motion Designer's Perspective


"Do. Or do not. There is no try."

How often do you arrive on the first day of your new job and are asked to design for something you've never seen? This was the exciting task presented to me on day one as a motion designer for the Razorfish Emerging Experiences team, where I'm currently designing for HoloLens. It's an entirely new platform to design for with few examples to learn from. That aspect is both challenging and exciting, and a tremendous opportunity for any designer. As soon as I put the HoloLens on for the first time, I quickly realized there is new UX knowledge to understand, software to learn, and workflow changes to adapt to. 

I'm getting a crash-course on the HoloLens and its interactions.

I'm getting a crash-course on the HoloLens and its interactions.

Put your UX hat on

I realized there is a strong need to develop a UX design mindset, no matter what role you play in the process.

After my first experience with HoloLens, I realized there is a strong need to develop a UX design mindset, no matter what role you play in the process. The physical nature of the device, and the environment it augments, requires new thinking—especially for a motion designer. First, the field of view is limited and requires creative thinking for how (and when) items appear. It's not a 1:1 comparison with how you normally view things in the real world as objects you may need, or are expecting, could be above, below or even behind  you. 

Second, discovery is a big part of the HoloLens experience and I found that initial placement of objects is best done at a distance. This allows users to recognize them so they'll be invited to come forward and inspect details. The same goes for animations. I noticed if the user is standing right next to a huge moving object, it feels disorientating. It's best to frame your scene as you would frame a scene in a camera: things that are close should still be recognizable and carefully considered. The difference in this case is that the holographic "pictures" we create can be viewed in an infinite number of angles—details, textures and other surprising elements are important.

Reducing the polycount in your models is key for a smoother experience. 

Reducing the polycount in your models is key for a smoother experience. 

Failing is learning

Designing for HoloLens isn't the linear process I was used to as an experienced motion designer. It's more of a looping sequence of trial and error—especially at this point of learning the parameters of the platform. It starts with tinkering, sketching, figuring out the possibilities and limitations with a developer, then coming up with a hypothetical concept based on what you think will work on the head-mounted display. You bring this concept to life with your two-dimensional designs on screen and work with a developer to get them onto the HoloLens. Once loaded, things may or may not work. Then the process begins again until the desired experience is achieved. 

The standard motion design software isn't capable of supporting this level of HoloLens interactivity. This meant two things for me: I needed to learn new software and new ways to work to make my creations come to life. So on my first day at my new job, I downloaded Unity and got to work. Unlike other gaming engines I've used, Unity isn't the most designer-friendly program. It relies a lot on scripting—either C# (pronounced "c-sharp") or Javascript—for the interaction work. Learning curve aside, it wasn't impossible to master the basic functions of this new program. Spawning a cube, texturing or lighting a scene still feels the same as Cinema 4D, and Unity has a huge library of tutorials and a lot of support from its users. After a day, I was tinkering, making a bunch of new objects, exploring the asset store and rendering my new creations. And the rendering is the best part: you no longer need to wait 10 hours for a render as Unity does it in real-time (but with noticeable differences in quality from the items I would create on Cinema 4D render engines).

Here is the Unity environment with the models I created for my first HoloLens experience.

Here is the Unity environment with the models I created for my first HoloLens experience.

Be comfortable in being flexible

Because Unity is only able to import geometries, joints, key-framed animations, cameras, lights, and basic shader settings from Cinema 4D, I found the best workflow is to start from Cinema 4D. Do your modeling, bake all animations to key frames, apply basic shaders then export as .fbx. Next, bring that into Unity to work on things like shading and lighting. Non-geometric objects, like particles, are best done from scratch in Unity. There are also volumetric plugins in the Unity Asset Store that look promising.

The HoloLens has a few hardware limitations that will affect what you design. I learned it will not smoothly run apps with high quality models, and have tested that a 2,000 to 6,000 polycount per model is a safe bet. The models will look low-poly, kind of like game models on older generation mobile devices. So remember that when designing a scene for HoloLens, you have to take into consideration the polycount and limit what you place in each scene. A smoothly-running app precedes graphics fidelity... for now.

Prepare to make magic together

The collaborative nature of designing for the HoloLens means learning new techniques and processes, but the end results are truly magical. It is exciting to know that motion designers can now have one more medium to play with—one that can potentially impact the audience's life on a personal level.


Fyn Ng (@fyn.ng) is a Motion Designer for the Razorfish Emerging Experiences team, based out of our New York office.

5 Comments