New year, new tools. Has any one had the chance to test Microsoft Hololens? Vray does a great job with Oculus and Virtual Reality and I was wondering what can we do with 3d models and 3d renderings applied to Augmented Reality?
Before buying those expensive glasses, I wanted to ask if someone in the forum had run a test in first person or participated in any public demo of the Hololens?
I am concerned about the screen resolution, color reproduction, hardware limitations due to small processing units inside, software compatibility (just Sketch Up for now?)
I’ve tried the Hololens, it’s an early version so far, don’t expect the perfect device yet, what it does is incredible. The field of view is small yes, but that tracking and gesture control is immense!
I don’t think we’ll see VRay on the Hololens any time soon… The VRay solutions for Occulus / GearVR /HTC Vive are all about 360 degree images / video at the moment. Hololens is much more of a game-engine environment!
Thanks for the feedback. I am not sure if Vray can’t do stuff for augmented reality. I dont know…realtime rendering pushed to the glasses and overlay to reality?
It will be a whole new market in the upcoming years and for sure Vlado and his team wont stay looking! I hope
Remember all the computing power for the Hololens is in the device itself.. that’s a lot of GPU wattage you’d need on your head to render Vray at 90-120fps!
I know, but I was thinking to something that calculates all in your machine and transmits the result to the glasses in realtime. Cable or wireless… Boh! Just wondering around!
I think the way VRay can join the fun is creating plugins for Unity and Unreal to use VRay algorythms for GI baking.
And after that you create scene for Holo with VRay’s GI solution.
The real benefit to the Hololens is how well it interacts with other Hololenses for collaboration in an AR environment. Unlike closed VR solutions like Vive or Occulus, it projects the AR content on a small box in the center of your field of view. You can still see other people in the room and interact with digital content as if its a model on the table in front of you through hand gestures. The room/background must be slightly dark so the AR content can read through as it can be blown out in super bright/sunny environs. The device itself does very basic 3d scanning of the room and you can lock models in real space. From a horsepower standpoint, Dave is right, you can think of this as essentially a gear VR or google cardboard with smartphone-like performance. Any realtime GI calc seems unreasonable. I have no notion of the ability to stream content to the device.
Baking out GI and mapping it to reasonable polygon-count solutions is the limit for the device as it stands. Again, its target is not high performance, but rather ease of interaction. In the architecture world this is a much better solution than typical VR as it doesn’t carry the liability issues (client puts on VR and falls and injures themselves) or solo use (one person per VR environment) of closed frame VR solutions and allows digital 3d models to be interacted with through multiple devices in the backdrop of the real-world (including other people). It’s geared to a much different market than the Vive or Rift.
Haha. Not bad lag at all. However, it’s all dependent upon so many factors, of course: quality of the wifi connection, polygon count, etc! The exact same model, computer, and router, moved to another part of the building, have yielded different lag results! This is a visualization technology that is so utterly based on every case being unique. People will comment "oh, a lot of lag!, then base their entire opinion on that one experience. Whereas the informed viewer will question everything and do multiple tests. Bit of a nightmare, really. heh.
Have you guys tried it yourself? Any demo? Is it worth purchasing or still to early? We would love to use it for field coordination too, with clients and engineers , remote video calls and things like these.
Hey sorry for the delayed response. Here’s my brief review:
We have the Hololens and have been building content and testing. The promise is enormous, but the 1st gen device does not meet up to the potential. The main issue - and a deal breaker for us in terms of using with clients in any sort of robust way - is that the field of view is impossibly small. Imagine you have an 8 1/2 x 11 piece of paper in your hand. Hold that in front of your eyes about half an arms length away. That is the viewing area in which a hologram will show up when you have the device on… While it’s certainly cool to experience, with large room scale models are absolutely impossible to view as complete entities. You end up having to shrink the model down to table top size in order to understand the whole concept.
I’m sure the next gen will expand that field of view considerably, but that’s a ways off to be sure…
Having said all that, we are wasting no time in developing 3D content and apps in order to stay on top of the tech and learn about the best UX and UI practices. We know our competition is also engaged, and sitting around waiting for a better version of a viewer isn’t exactly a great strategy, lol.
Thank you a lot for the feedback! I knew that FOV is about 30-35’ maybe, and you are totally right when saying that you can’t fit a real size model in your site. That would be the main goal for this device, adding unbuilt architecture in real environments for review and coordination! Unfortunately is almost impossible to get the hands on the Hololens without buying it, very few Microsoft store have it, and they only organize events to showcase it.
Yeah, that is an issue. I believe Microsoft is going to be getting more aggressive about getting out to various events and stuff in order to let people try it out. I would scout their resources for announcements! Microsoft HoloLens | Microsoft Learn They have a great demo team cruising all over…
Not being able to see an entire model is a dealbreaker, and very off-putting when viewing. It’s sort of like a ‘tunnel vision’ effect - seeing your virtual object in a kind of picture frame, while the edges get cut off past that window. I don’t mean to sound too negative, because even after all that it’s a phenomenal piece of kit that’s only going to get better…
The good news is that there’s a few apps that can immediately view any fbx file.
The bad news is that more robust experiences can only be built with Unity. Other game engines do not yet have a native exporter. This leads to a whole rant on my part about real time VR / AR and game engines; namely that, as many of us know, the workflow from Max to Unreal or Unity is absolutely wretched. So wretched, in fact, that hardly any serious tutorials exist on the subject. It’s why I’ve switched - for now - to Stingray. Does Stingray produce the same quality as Unity or Unreal? No. Is it about 100% faster to get assets from Max up and running? Absolutely. Will Stingray someday look as good as Unity and Unreal? I have no doubt. Anyway, mini rant over… heh.
Stingray is supported from Hololens? Nice… In the other side, if I need to see just static models, I dont need a game engine right? My understanding is either use a sketchup 3d file (and pay 1500$ for their app to view models) or use other apps that support FBX format. We work with Revit/3dsMax a lot so Autodesk file extensions are to be exported more often.
what trouble are you having? we’ve exported entire scenes as fbx and they come into unity fine & well organised ready for the materials to be rebuilt from freshly baked textures.
No, no. Just to be clear…Hololens doesn’t support Stingray; at least not yet. A native exporter might be coming to Stingray, but not anytime soon.
There’s a free app (can’t remember the name) that views FBX files in Hololens with zero work involved. No game engine required!
It’s beyond me why the SketchUp people don’t yet have built in Rift or Vive support. Crazy. Instead, you have a choice of a half dozen ‘pay to play’ services. Heck, Sketchup doesn’t even yet have the capability to export a stereoscopic cube map for fixed point VR…
Also how do you get around the memory limit for textures?
I was prepping a small floor of an office building, got as far as sofas in reception and ran out of texture memory. All the maps need to be high res so it actually looks good imo.
Are you doing any extra work in max before sending to FBX?
Fair enough on the first 2 - we baked the lighting.
memory limit - http://amplify.pt/unity/amplify-texture-2/
used unwrella to unwrap and vray to bake. super simple process and very quick. problem we found is that despite being quick, it still adds enough time that most of our clients dont think it’s worth the money.
