Oculus Rift for architectural visualizations

[OPKJoe]

Hi everbody,

my name Joe Wälter and I am currently studying Digital Film Making/3D Animation in Bochum, Germany.
For my Honours Thesis I am going to research the chances of the Oculus Rift to be used as a cheap but efficient Virtual Reality Device for architectural visualizations.

My methodology consists mainly of two different approaches:
I have created an interior visualization of an office, which I am going to present to a small audience in three steps:
1. As “bare” photorealistic renderings, printed on photo paper
2. As photorealistic 360° panoramic renderings, which can be explored with the Oculus rift. This presentation reacts to head-tracking and supports 3d stereoscopy.
3. As an interactive VR-Environment, programmed in Unity3D. The benefits are the same as in step 2 plus the user can move and interact with the environment (e.g. open doors) with a wireless game controller. Disadvantage: Not as (photo)realistic as the renderings.
With this approach I want to work out whether the Oculus is an enrichment to a potential customer by investigating, how they react to different kinds of presentations.

My second approach aims at the technical background – will there be enough support by the developers/communities and are 3d artists interested in producing VR content at all.
For this purpose I am asking you:
If you could spend 10 minutes of your time to answer some questions about VR in 3d visualization, I’d be truly grateful. You can even participate, if you haven’t made experiences with VR yet.

http://www.sogosurvey.com/k/RQsQVUPXsRsPsPsP

The survey is completely anonymous and the collected information will only be used for this work.

Thank you all in advance.

Joe

P.S.: Feel free to ask me anything about this.

[Neilg]

I think the technical issue present here is how to render a 360 degree panorama in stereo. We'd have bought an occulus a while ago but there's no easy way to generate the content for it.
Real time is pointless, nowhere near enough control over image quality for the work we do.

I also love how your question about real time rendering your only option for no is 'no, not yet'...

[cheerioboy]

I think the technical issue present here is how to render a 360 degree panorama in stereo. We'd have bought an occulus a while ago but there's no easy way to generate the content for it.
Real time is pointless, nowhere near enough control over image quality for the work we do.

I also love how your question about real time rendering your only option for no is 'no, not yet'...

Agreed, I'm still looking forward to a stereo 360 rendering method within vray. At the moment you can get away with use a z-depth to fake the depth but it does a hack job. And like you say, real-time is impossible without completely re-building scenes that are optimized for real-time. A few possible solutions around this might be if whole scenes could get their lighting/gi baked into materials and/or a way to output dense point clouds of 3d scenes.

[OPKJoe]

Hey,
thank you for your response.

You're right, there is no easy way to do a real stereoscopic 360 panorama rendering at the moment, but I achieved pretty good results with the z-Depth channel. The VR-Player does a pretty good job here.

I also love how your question about real time rendering your only option for no is 'no, not yet'...

Yeah, but I put "not yet" in brackets. So the answer should contain both options "No" and "No, not yet".
But thank you for this suggestion, I will keep this in mind when I am evaluation the answers.

[beestee]

I have been reading a lot about this lately and is seems that the VR niche is really being taken hold of for internal training and virtual layout purposes more so than as a sales or PR tool. I do believe that eventually more immersive presentations will become the norm, but I think that affordable glasses free tech will be what drives wider acceptance of the more immersive presentation formats. Either that or the Occulus will need to become significantly more sleek and elegant which going by the appearance of the current prototype "crystal cove," it seems a distant secondary goal. Sony's prototypes are more promising in this regard, but still very sci-fi-ish.

[Neilg]

Either that or the Occulus will need to become significantly more sleek and elegant which going by the appearance of the current prototype "crystal cove," it seems a distant secondary goal. Sony's prototypes are more promising in this regard, but still very sci-fi-ish.

Yeah, there's going to be a good few years before we can convince a client to get a couple for the sales center. It's going to happen though.

is there a cleaner way to do the z-depth method - maybe by splitting the image into r,g,b layers and using the z-depth as an offset mask? that probably wont work, i came up with the idea just now.

[vlado]

We did not have time to do a stereo dome camera for the initial 3.0 release yet, but it will come in a service pack.

Best regards,
Vlado

[cheerioboy]

I don't think so... using the z-depth to create the second eye is always going to end up skewing something. I've also tried using 'StereoPhoto Maker' to splice together slices of a 360 camera moving in a circular motion, to capture the parallax. But it takes a very long time to setup & render & stitch. And the result is questionable. So not quite a viable solution at the moment.

At least I've been having some fun with the Poppy3d (http://poppy3d.com/), loading simple stereo renders for previewing with an iphone. Not the full 360 experience but maybe someone will program an iOS app that lets you load a 360 stereo pair, while using the gyroscope for looking around. Would be a much cheaper alternative than the Oculus. If you have an iPhone.

[samuel_bubat]

A little story regarding the Occulus. We recently had a customer who wanted every new technology available for his new project.
AR, Multitouch with a tangible interface and realtime stuff with the occulus (Unity). While the occulus is absolutely amazing the problem why it
was turned down. Nearly everyone in the presentation became nausea Reading throug the web, this seems to be a common problem with
VR Glasses. It needs some time to get used used to it. Unless this is solved (if it can be solved at all) it´s pretty risky to make customer presentation
with the Occulus.

[cheerioboy]

A little story regarding the Occulus. We recently had a customer who wanted every new technology available for his new project.
AR, Multitouch with a tangible interface and realtime stuff with the occulus (Unity). While the occulus is absolutely amazing the problem why it
was turned down. Nearly everyone in the presentation became nausea Reading through the web, this seems to be a common problem with
VR Glasses. It needs some time to get used used to it. Unless this is solved (if it can be solved at all) it´s pretty risky to make customer presentation
with the Occulus.

Good point, that's an issue I have as well with the Oculus. If you follow the news on their latest version of the headset, they're suppose to have made improvements in the area of motion sickness (http://www.theverge.com/2014/1/7/528...t-crystal-cove). The current solution I've had is just sharing static 360 images, you only have the viewer look at it for a few minutes, there's much less sickness to be had. fortunately

I believe there's still not a single renderer out there that can render a stereo 360 pair.

[Neilg]

If we were to use it it would be an interactive panorama - a still 360 image, and when you look around there are prompts to press a button to jump to a new one. probably just fade over, rather than a lurch forward - as fun as that would be there tends to be a lot of drinking at events we put on...

[cheerioboy]

If we were to use it it would be an interactive panorama - a still 360 image, and when you look around there are prompts to press a button to jump to a new one. probably just fade over, rather than a lurch forward - as fun as that would be there tends to be a lot of drinking at events we put on...

Yeah a lurch might make people motion sick :P
Fading from spot to spot is a great idea - Although in the Oculus' current situation I think it's better to have people view in very short periods of time. So removing the headset between transitions works well. Having supplemental material between sessions helps with making that process feel natural.

I'm also interested in adding ambient animation. So having cars driving by or trees rustling outside the window.

[jacksc02]

We have done a few archviz jobs with the occulus and unity.
It works really well and looks good on the Rift. The only let down is the resolution of the Rift. If the HD version ever gets released then it will be more widely received and hopefully there will be less motion sickness.........

[OPKJoe]

Thank you all for this interesting discussion. This will be very helpful for my survey!
And it kinda reflects my own experiences I made with the device.
In mid February I will be doing the field experiment and I am curious about how the participants receive it.

[OPKJoe]

Du to a small delay, I couldn't start with the experiments yet.
Therefore the survey is still online.

But please allow me to push this thread a little bit up (one last time).

I would be really really grateful, if you could take 5 minutes to answer the questions. It would help me so much.
Thank you in advance.