Two weeks ago, I have traveled a bit around Europe and I have visited some beautiful cities. One of this incredible places has been Prague, that in my opinion is a marvelous city that everyone should visit. Apart from the splendid Cathedral and the tasty Trdelnik, Prague also features an interesting Virtual Reality startup: VRgineers.
VRgineers is the company behind the high-end headset XTAL. XTAL is an enterprise headset with incredible specifications, like for instance:
- 5120 x 1440 display resolution (2560 x 1440 per eye);
- OLED display
- Custom non-Fresnel lenses
- 180° diagonal FOV
- Spatial 3D sound from a built-in sound card
- Embedded microphone
- Embedded eye tracking
- Auto-IPD adjustment
- Embedded Leap motion v2 sensor
As you can see we are talking about a headset that really tries to push that technology at its limits… it really looks like a 2nd gen device. The problem is that this innovation has a price: the cost of the headset on the market is 4900€. So, it is a high-end device tailored at enterprises, a bit like Varjo or the StarVR One.
I was really excited to visit VRgineers because I had never tried such a high-end headset. I really couldn’t wait to try it. Let me tell you what I discovered there.
HQ Visit
I have been greeted by Ivo, that let me enter in VRgineers offices and that made me try the XTAL headset (more on this later on) and then the CEO Marek Polčák arrived and made me try the development version of the device and also let me have a tour of the company.
Having a tour inside VRgineers, that is a startup that is trying to develop almost all the main parts of its headset, made me understand how creating a Virtual Reality headset is incredibly difficult. We, the VR users, sometimes believe that the process is quite easy: “why don’t they add a different display to that headset? This way the resolution would be better!” or sentences like this one are something that we are used to saying continuously. Actually, what we don’t understand is that there are a lot of aspects to take in count every time that a design choice has to be taken.
For instance, when I talked with the guy that designs the PCBs of XTAL, he said me that he has to coordinate with the other guy that cares about the mechanics of the device, because the board has to fit inside the shell of the headset with all the other components. Then there is the weight and the dimension: you can add amazing features, but maybe, in the end, the headset is too heavy. Maybe you have a fantastic display, but then you have also to change lenses. So you craft some new lenses and they are better than the typical Fresnel ones but the resulting image is terribly distorted and so the results is worse than what you hoped for. Or you manage to sort all of this, but the resulting headset is so expensive that no one can afford it.
These are things that I could already think myself before going to Prague, but being there, watching so many passionate people working on so many things, from the circuits, to the lenses, not to mention the aesthetics, the testing and so on, make me understood for real how a headset is a complicated product that really needs lots of different kinds of expertise to be created. Seeing all those different PCBs they were experimenting on in the lab (sorry, I was not allowed to take photos to them), and talking with the engineers that told me that every one of those boards required a lot of effort to be designed, and then manufactured, and then tested… and even some hacks to make them work if there were little problems (sorry, you can’t just recompile and ship an update with hardware)… opened my mind about the effort required in creating a headset.
And designing a headset especially needs lots of compromises between quality, cost, comfort and a lot of other features. We still can’t have a headset that looks like the one of Ready Player One and that costs $99.99 with Amazon Prime. So, some features have to be sacrificed. Regarding the cost, for instance, Oculus can afford to sell its devices under cost, because it has Facebook’s deep pockets covering the losses. But a startup must create a sustainable business, so it must prepare a business plan where it writes how much headsets it forecasts to sell, how much are the global expenses of the company and so calculate a price to make the business sustainable. For XTAL, that is a startup that is trying to sell to enterprise niches, the resulting value has been €4900, that may seem a crazy price to us, but either it sells at that price, or there is not VRgineers anymore because it goes bankrupt. Can you see the problems?
I think that everyone in the VR communities should have a tour of a VR headset company, to see the passion in the eyes of people manufacturing the device and also to listen to their difficulties. It has been very useful for me and I really want to thank Marek, Ivo and all other people at VRgineers to have opened my eyes.
My hands-on with XTAL
Ok, so, I guess that you are curious about my experience with the device above anything else.
I have tried something like 4 demos with XTAL.
My first experience has been with a solution to train helicopter pilots. I seated there, on a chair of a helicopter, worn my XTAL helmet and tried to pilot the helicopter. It has been my first experience with XTAL, so when, for the first time in my life, I experienced a high-FOV, high-resolution headset. You can see a short video of this historical moment here below.
From what you can see in the video, it’s better for this planet if I don’t pilot helicopters… I have tried this demo and I died something like 15 times in a row. For 2 times I have managed to make the helicopter to take off (also thanks to Ivo’s help), but then I started spinning and I exploded. I’ll keep developing and blogging, that’s better 😀
After that, Marek took me to an amazing green room where I tried the development sample of the device (sorry, I couldn’t take photos or videos of that as well), so really the latest evolution of the XTAL. I tried 3 demos within that:
- one was a very high-resolution 360 video that let me feel on the top of a high skyscraper. Thanks to the improved realism offered by the XTAL headset, this felt incredibly scary;
- one was a very elegant interactive virtual tour of a house. It was interesting that they implemented Leap Motion’s interfaces in this demo… also the movement inside the house was performed using hands gestures;
- one was the showcase of a very cool car that I will never be able to afford;
While talking with me, Marek also introduced me to other features of the XTAL that I have not mentioned above. One is the configuration tool, that lets you configure, debug and check the device as you wish. It has been studied for professional customers, so it really gives you a lot of features.
The other one is the modular tracking technology. You can track your position in space with XTAL with whatever tracking system: you just need to attach an add-on to the headset relative to your tracking solution. So you can use SteamVR (even the v2!), you can use Optitrack, you can use Vicon, you can even use a secret-tracking-tech-I-cant-talk-about… you just get the add-on for that tracking system, stick it to your headset and you are done. The XTAL per-se has no embedded tracking sensors, but it features a lot of possible faceplates and accessories that will let you use it to with your favorite tracking technology. See the video above: they just attached a Vive Tracker to the headset and the positional tracking worked! You can even switch the tracking technology on the fly while you are using the device! Just amazing.
Marek has also explained to me that enterprise customers don’t love much the SteamVR platform. The reasons are two:
- Steam is a platform for games, and enterprises don’t want to be forced to install a game store to run a VR experience;
- Steam continuously updates and every update can break something regarding the VR experience. The solution may be the one of not updating, but this means being frozen to using an old platform.
That’s why VRgineers has created its own tracking runtime based on SteamVR: to listen to the needs of its enterprise customers. Cool, isn’t it?
How is the XTAL?
Well, it has pros and cons. As I’ve thoroughly explained above, making a headset is hard, especially for a little startup like VRgineers, and some compromises are necessary.
You may see in the above video that when I try the XTAL for the first time, I am amazed, but I am not completely in awe. How is that possible? It was the first time for me trying a big-FOV headset! Finally I had peripheral vision!
Well, it seems that while the logical part of my brain was amazed by the fact that I could have peripheral vision in a VR headset, the most instinctive part was just saying “ok, there is peripheral vision… but this is something that you have every day, it is not amazing. And then, let’s be honest… do you even care that much of your periphery?”. This is something that I have already read from some Pimax users: while you wear the headset, you don’t feel so incredibly surprised by the high FOV. You feel the difference when you return to wear a standard-FOV headset: at that time, going back from Pimax to Oculus, for instance, it looks like you are using binoculars. That’s a strange mechanism of the brain.
Apart from this digression on my brain, I can tell you that the FOV is really big and you have again your peripheral vision, that can be fundamental for instance in action games to see enemies coming from the sides. It is there and it is fantastic… it is not like looking through binoculars anymore. What is interesting is that you have a great horizontal FOV, but the vertical FOV is not that astonishing, so it is a bit like when you look at a panorama photo (very long, but not that high). Marek explained to me that trying to make the vertical FOV big as well would mean making the device even heavier and more expensive and complicated, so at the moment they are focusing on the more important horizontal one.
The device is quite big and heavy. The Pimax too is quite big and occupies a large volume of your face… and this seems the only solution for now to have a big field of view. One of the biggest issues is weight, in my opinion: even if the XTAL is continuously improving on this side (the VR Hero 5K, that is the previous version of the XTAL, weighted more than 1kg(!!), the XTAL is more than 700g), I think that it still feels heavy. For short usages (like for instance, the showcase of a car in an exhibition), this is not a problem, because the headset feels well on the head, but I can’t imagine it used for a long time (e.g. playing Beat Saber for 1 hour). It would be really too much… I myself started feeling neck strain at the end of my demo session…
The other issues are the lenses. Crafting custom lenses for a wide-FOV headset is really hard. We are seeing Pimax too having problems of distortions in the peripheral vision. The XTAL still features various distortions: so while you wear it, you can see quite well if you look in front of you, but the more you move your eyes, the worse the image becomes. There are clearly distortions in the external areas of the visuals. The good news is that this is something that can be fixed by software and Marek told me that it is the n.1 priority of the company. So, hopefully, in some months we can see the situation fixed.
Apart from these shortcomings, what I really loved was the sharpness of the display. The resolution was very high, and the screen door effect was minimal. The colors were really crystal clear, thanks to the OLED technology. I really loved the display: I think it is really the point of strength of XTAL. Marek has also showed me a tool to compare the visuals of other headsets to the one of the XTAL and I can say that the difference is really noticeable.
Distortions apart, the lenses were also a clear improvement over the ones of the Vive and the Rift, and I noticed fewer god-rays, rings and this kind of annoying stuff, so I think that they are going in the right direction.
Apart from the weight, the headset was comfortable. And it was amazing that thanks to the auto-IPD adjustment, I had just to wear it and it auto-configured its lenses.
Regarding the design, I personally like the design that makes the headset look faceted like a crystal, even if I don’t like that much the combination of black and violet colors… here in Italy, these are usually the colors of funerals, so they don’t make it that happy.
I must give a special mention to the Leap Motion v2 sensor. It is embedded into XTAL and it was the first time that I tried it as well. It features two wide-baseline IR cameras and two IR emitters for the tracking. The performances are much better than the v1, really much better. And while David Holz clarified that the runtime is the same Orion that runs on the v1, the better quality of the IR cameras and the new positioning of them really improved the tracking quality a lot.
The system detected quite well how I bent my fingers, it was able to track the hands also when one was above the other and the tracking FOV was really wide. I could move my hands freely however I wanted and still see them tracked! The nominal tracking FOV is 180°, but I have noticed that actually if you go beyond 120° the tracking becomes less reliable and sometimes the hands get lost. I could move my hand to its maximum distance and it was still tracked, it was really something amazing. Anyway, let’s be clear: tracking is still not 100% perfect (some errors and tracking losses were still present) and I think we are not at the point we can substitute completely controllers. But in all the demos that I have tried, I have never been provided with a controller and I can assure you that I never felt the need of using it. For menus and such, using Leap Motion v2, with the awesome UIs designed by Leap Motion was really great.
In conclusion
The overall impression that I had on the XTAL is that it is an interesting experimental headset. I don’t think it can be ready for mainstream adoption (but it is not even its purpose) because of its price and its present-time issues, but it can be cool for niches where the visual quality is of fundamental importance. Big weight and distorted visuals are its main shortcomings, while the fantastic display and the embedded Leap Motion sensors are its main strength. I also appreciate that such a small startup is trying to push forward the limits of VR.
Being there also made me understand why Oculus, HTC and all the others are not already offering v2 headsets: the technology is not ready to offer an affordable solution that requires a non-powerful PC, that has amazing visuals, eye tracking and hand tracking, and that has no issue for what concerns distortions, comfort and usability. We’ll get there, but we’ll need time. In the meantime, brave startups like VRgineers will try to push the innovation forward to let us arrive there.
Did you find interesting this hands-on of mine? Well, we’re not over… I have also an exclusive interview with Marek, the CEO of VRgineers, where he answers all my critics and talks about the future plans of the company! Subscribe to my newsletter to not miss it tomorrow!!
(Header image by VRgineers)