VR sub forum for tech talk
polycounter lvl 20
So with Virtual Reality about to hit the consumer market within the next year i think its safe to say there'l be a bunch of ambitious devs wanting to get into the mix, and since its a fairly unique platform with its own set of challenges for creating art what better place than polycount to act as a central resource!
I know a bunch of people here are already working in this area, so i dont think i'm alone in wanting something along these lines.
I'd gladly volunteer to mod/manage it if needed.
Anyways, chime in if you think its worthwhile (or if you think the idea sucks)!
I know a bunch of people here are already working in this area, so i dont think i'm alone in wanting something along these lines.
I'd gladly volunteer to mod/manage it if needed.
Anyways, chime in if you think its worthwhile (or if you think the idea sucks)!
Replies
Absolutely! if it warrants more than a single thread then we could go from there. Very much down with that
There is actually already a wiki page that could use some more information.
http://wiki.polycount.com/wiki/Virtual_reality
Joost, I got a lot of respect for you but in this case I think you're dead wrong. There's definitely more than a few bullet points to cover for good VR game design - comfortable, fun and pretty.
Have you actually tried to develop for VR yourself?
Edit:
Check out some of the oculus connect talks coming up, it might make you change your mind.
Peanut: yes absolutely. actual screen res is only one factor, optics play a huge role in perceived resolution. I've worked with the dk1, dk2, gear and recently tried the Vive and each iteration has been better than the last.
I have no doubt that this generation of VR will take off, just needs a bit of time to build a user base for the hardware.
Resolution is good enough with first gen HMD (I've got a DK2 and I tried Vive as well - Not Oculus CV1 yet) that you forget about the pixels and just "feel there" when the experience has been designed for VR from the ground up. Many people out there have tried half-baked experiences or even horrible workarounds that pretend to turn regular games into VR experience, e.g. SKyrim, hence the mixed feedback.
And resolution is really just one factor anyway - hand tracking and good art are just as important to achieve presence, and we're definitely getting there.
It will go maintream this time because it's comfy enough to keep using it for the duration of a full movie - and thus the entire entertainment industry is interested. Heck, I've watched the entire last season of Game of Thrones in a giant virtual cinema room! Quite the experience
for what i do i don't fancy upgrading now to a model with only a 1080p display at all, want higher resolution to make the expense worth it. and preferably one without vsync requirement.
yeah DK1 is not supported anymore, and the latest runtime update (0.7) makes extended mode obsolete so your DK1 will not even be detected anymore if you upgrade. AFAIK latest Unreal release is using 0.6 still, which should still work with your DK1, but I could be completely wrong. I haven't touched my DK1 in over a year now.
Yeah unfortunately the DK1 is done as of this month, but if you wanted to roll back to the old runtime like Mant1k0re suggested then you should still be fine.
As far as upgrading your hardware i'd suggest waiting until the first consumer models hit the market. The dk2 was already old tech after about a month of shipping so i think you'd be disappointed by upgrading now.
Unfortunately i dont think you'l find any HMD that doesnt have a vsync requirement as this is required for a low persistence display. With the dk2 you can disable it via code, but you'l get sick quick from the judder. If you're over or under 75fps (90fps for consumer model rift) then you'l have a bad time.
yeah i'm not getting DK2 most definitely. hoping for a 4x resolution increase from DK1 actually. that should be nice for displaying alpha cards in style finally. the immersion is already strong with DK1 for my tests as such, so it will have to hang on till we see proper consumer models emerge.
just wondering if anyone is working on utilizing multiple graphics cards in a system to drive VR applications? like, assign one GPU per eye to help with performance issues?
Nvidia just put out their first beta drivers with this system to coincide with the 0.7 release so hopefully performance improvements will start to happen soon enough.
I think basic UI production has a fair amount of VR related complications, so i might start there.
Other general topics that come to mind:
-hardware restrictions and workarounds for art production (mobile vs. desktop as an example)
-Adapting particle systems for VR (billboards wont cut it anymore!)
-Getting started in VR (i'm thinking a basic 101 for anyone new to the medium. Present all the hardware and software options + getting started with minimal effort)
-Geometry vs. normal maps (normal maps really take a hit when the user can analyze an object from every angle and position so this might be a worthwhile discussion)
Keep in mind some of these ideas stray a bit from 100% art related subjects, but i think a better understanding of the entire process can help anyone involved regardless of a specific production role.
Please chime in with any other ideas, i'm probably overlooking a lot of basic stuff that i just dont think about anymore when doing this kind of work.
Thanks!