#virtual-reality
1 messages ยท Page 61 of 1
yeah i want a 1080ti
we need amd to be real competition to nvidia
and with intel!
buys the protest card, back to Matrox
at least with the Matrox you can do like 20 displays
matrox is nice for that
That's because AMD's stuff is so much more worse
I bought a GTX960 (regrets) when that 970 business happened because the AMD cards were crappy. I upgraded to a 980Ti right before I got my Vive and luckily caught the 90 day upgrade window to the free 1080 โค EVGA
free?
yeah
it's a huge problem that some people think amd is "crappy"
i thought you paid the difference
EVGA does a 90 day step up program. So if a new card comes out after you buy one you just pay shipping and price difference. They send you the new one. For me to go from 980Ti->1080 was just $15 in shipping.
AMD is behind with stuff like multi res shading, and power efficiency, but that's it
and currently they don't have high end gpus
but that was different before 10xx release
AMD needs to stay around to light a fire under Nvidia's and Intel's ass... but problem is those two companies are very good at shutting out competition through anti-competitive means D:
noise and heat put me off amd, i run everything stock with good cooling to keep it as quiet as possible
Thankfully all the consoles and Apple are using AMD right now to give them some money
a gpu has to eat at least 300w or it isn't a gpu
Also Arizona Sunshine devs rescinded their ill-thoughout exclusivity for i7
'it was supposed to be a reward to the people that heeded our recommendation guys!'
'please believe us!'
well maybe I should make my game exclusive to i7 lol
it will not run good on CPUs with less than 8 threads
If I ever take a money-hat, I'll straight up say - I took a money hat! woo!
i though they were blocking older i7s
And then start a kickstarter to a value exceeding the cost of the money hat.
being douchebags
Then when gamers are all - why'd you take the money hat, I'll say... because it let me make a better game then without it.
CPU progress is even less than gpu
horrible
no idea how we should ever be able to run 8k vr
Problem is... when you don't even get the money hat, gamers aren't going to buy your game more just because it's not exclusive!
I have no problem with people taking money for games as long as they are up front about it.
It sucks, but if it pays for the game to happen that's cool.
Right. And I think that's true for the most passionate/vocal of the community.
why use expensive screens when you could project an image straight into the eye
They understand economic rationality. They just don't want devs treating them like idiots.
i thought dlp uses spinning mirrors
The magic leap stuff
still not cheaper
uses a type of retinal display tech
vr needs to be 8k
Virtual Retinal display
but no gpu can handle that
a little dlp mirror isa lot lighter than an 8k panel
vr headsets should make you look like a bono fan, not like you have a box stuck to your face
But yeah, VRDs is essentially fiberoptic steered by DLP style mirror motors
Or something to that essence. I can't remember the deets
It was from an interview with Tom Furness on the voice of VR podcast
Any of you guys listen to that podcast?
nope
they all seem to be stuck with the box on the head cos that's what palmer demo'd
I hope vive 2 is announced soon. need 4k and eye tracking.
i saw an 80s VR system
should be doable
it was whacky
It's pretty good stuff... lets you know what other VR devs are doing and working on...
your whole head went inside it
vive said they've halved the component count
release galaxy s7 and make s6 cheaper
release vive 2 with 4k and eye tracking and make vive 1 cheaper
but steam also said the steam controller was amazing, apparently its not
they halfed lighthouse
I Love the Steam controller, but definitely isn't for all games
why is the lighthouse so big?
Rushed to market I think
or is it a lot smaller than it looks
Just because they've reduced components and in essence reduced the price it takes to make them, does not mean that it will trickle down as much as people are hoping
I really doubt it will be 50% pricing next year
I bet v2 of VR HMDs isn't coming for the next 2 years or so
Maybe...MAYBE... 25% if we're lucky... But, I think they reduced the price on Vive $100 for Black Friday as a test
and we already know what Rift 2 might be like
If they saw any sort of big jump in sales because of that price, you may see the new price
If they didn't, they will reduce it once again
another $100 or so
They probably wanted to milk all the early adapters first that will pay extra. I know I would have paid more.
The first iPhone was the same way. I think it was like $1000 or somewhere near there. I didn't buy one myself though.
oh you are right, I thought it was more. I just looked up some pricing
I think it didn't use to have as good of a cell phone provider plan discount though
unlocked here
I'd like to see the next version of the OSVR
Yeah
yep
SO...
๐
So that's how they're rebranding Windows mobile
Fuck it. We can't crack the mobile phone market, but we can sure as hell crack the mobile PC market!
What's the difference? The difference is one can run full fledged Windows apps*!
*that are compatible with mobile specs.
Yay?
btw, when is MS going to announce VR ? We are in December already ๐
That's pretty impressive it runs the games. I wonder if you could connect a game controller to the phone and play games that way. Also RIP battery life
meh
Cellular PC has been a thing
But, this is what is going to cause AR to overtake VR
AR will be cool when it happens but I feel like we are still a ways off performance and battery wise before it gets as good as VR is now.
For games that is...
If I can wear a light headset that does high quality video pass through with 3d overlayed and not have battery die after 30 minutes that would be pretty cool
I don't see AR being cool for games
microsoft already announced vr a few months back
they are releasing their plans to OEM's. integrating the hololens sensors into a vr headset similar in design to the PSVR headset
@pearl tangle yeah, but they said they will give us details in December (specs and whatnot)
AR games are already quite good, especially if you get the holograms at the magic leap level too
yeah its still early in december. having the full windows 10 on arm architecture opens up quite a few possibilities too
cya ppl o/
Today at the Microsoftโs Windows Hardware Engineering Community event in Shenzhen, China, the company announced that VR/AR (mixed reality) headsets from top manufacturers are due to hit the market in 2017. Whatโs more, theyโll run on integrated Intel graphics without the need for a dedicated GPU. Back in October, Microsoft announced that mixed reality headsets for Windowsโ forthcoming native โฆ
Ha... ok.
rather interesting
So Microsoft doing their own thing.
well all the stuff comes with the embedded hololens inside out tracking in there
Essentially a mobile-VR without the smartphone kinda deal?
from the pic though it looks like everybody else has a tethered setup, or at least a box separate from the headset itself
Hopefully to a portable box.
on the belt
or some such.
That's acceptable wireless.
or wired.
whatever.
๐
yeah id say a smartphone sized thing you just chuck in your pocket is a good bet
just need the headset themselves to hurry up and be the size of regular glasses and people will use them all day for sure
Heh. Need those next gen meta-material optics
I read about Havard designing new lenses
completely flat, using diffraction grates to focus light
Only a narrow wavelength for now.
yeah its not going to take too long for some big shifts in that field now that there is strong demand for it at a consumer level
Exactly. Havard aren't the only players in town as far as that stuff goes.
seems likely that all those headsets are running something like the intel NUC box at the moment. if its the skullcandy 1 then thats decently powerful for a full integrated box
nah even glass manufactures are 1 of the most trade secret heavy industries in the world
nobody patents any of the techniques or anything so its all kept completely secret, moreso than coke recipe even
But the coke recipe is more of a marketing thing
then it is a real 'so nobody can guess our formula' thing ๐
Given the simple fact that coke has a variety of tastes!
And I'm not talking about the zeroes, stevias, diets, etc.
TBH, things like this is a good thing
We will get the good from whatever we can
And improve upon it in the next gen
I mean, the precise formula for the syrup that they make might not have been guessed by their competitors... but the taste delta of coke is variable enough that competitors can get 'close enough' to be 'like coke'.
Such that the only thing that seperates them is... the marketing/branding on the bottle.
and that nice pesticide flavoring
hah yep but its a lot harder to have exact tempering techniques and injecting the perfect amount of particular oxides into glass at the exact right time to get the perfect lens
My dad has glass charts that he can use to order specific types of glass
it's pretty cool
this microsoft stuff is going to do a hell of a lot more for the industry as a whole than all the other current headsets combined. it's going to start getting it into mass for commercial uses
I hope they don't pollute the VR market with low quality headsets
SO we can expect ASUS ROG version for $600
the fact that there is the pure AR, pure VR and MR all using the same general hardware and interface gives a lot of options
obviously not going to have the same level of stuff that we can do with the vive and rift for entertainment purposes but opens a lot of options for stuff.
did I miss an announcement?
will be interesting to see how apple comes along with their AR stuff in the next 12-24 months. microsofts current jump into hardware again is actually very impressive
Do we know if the MS VR stuff will at least do positional tracking?
I can dev for it if they have motion controllers.
yeah its all inside out tracking
all the specs have the full hololens capabilities built into it
so full hand tracking capability and spatialization
I see.
seems like leap motion is doing a mobile thing too now
Working in Silicon Valley, Iโve come to realize that everything that seems futuristic was actually on the drawing board a startling number of years in the past. Technology is like a flash in the darkness โ long nights building towards an instant when suddenly everything is different. Thatโs why I think all of us work โฆ
That's a pretty good FOV on their camera
definitely an improvement. that was the main problem with the other 1
aside from the pretty shitty tracking quality that is
yeah the orion 1 makes a big difference
but it still loses tracking too easily with a single sensor and they never opened up the API to multiple sensors at all
well when it loses tracking it loses the entire hand essentially. so whatever you were holding is dropped. and with a low field of view you have to be looking right at your hand all the time, which you dont really walk around with your hands in front of your face the whole time for anything
except maybe that 1 where you are grabbing 12 year old girls tits
Haha.
Well, the software can be worked around.
Instead of registering hands disappear as a drop
just register it as inventory
yeah still doesn't quite keep up with something like lighthouse either way
True.
maybe with some haptic gloves or something, poking at air with nothing in your hand is always a bit odd
Yeah. I mean, appropriate visual and audio feedback goes a long way....
but haptics are a nice catalyzer for cementing the tactility of those other cues
i.e. it's easier to associate the audio/visual cues with feeling, if there's already a little feeling to go with it!
yeah having something in your hand makes a big difference still. camera tracking stuff always seemed like a good idea in theory but its just not quite right in reality. same reason the kinect failed
Well, both hand tracking and motion controllers are substantially more compelling in VR then not.
But the Kinect failed because you can't really make compelling games that people really would like to buy with it.
I mean even basic tasks was a total chore.
Like selecting the next item in a menu or saying ok
so is stuff with gear vr
didn't stop a lot of people using it still. just gotta develop better ways to do things
Whereas in VR, that's actually a lot easier with camera tracked because of the third dimension.
And in VR... controllers will still allow you to do small repetitive things easier and faster.
But the delta isn't quite as significant as gamepad and kinect camera.
i.e. 10s of ms vs 1000s of ms.
the kinect v2 is pretty decent with its latency actually
same with realsense and tango and leap motion, they are all similar tech
I don't mean latency though. I mean time to execute a function.
Cancel, accept, left, up, down, right... basic menu navigation.
well if your tracking is perfect then i guess you dont need to hover over the button for 1.5 seconds for something
but yeah it was always too finicky to make shit happen
With 2D kinect, you had to.
kinect can track in 3D space without any problems
That's true.
But it's a harder task to represent that depth information to players
With a 2d display
i.e. you're swiping where you think the ball is... but you're not forward enough.
In 3D that mismatch is obvious.
But in 2D, it's not. So kinect basically failed at basic human computing interaction requirements in the primary application in was designed for.
What's funny in retrospect is the amount of hype it had
Given that simple fact! ๐
So much smoke and mirrors to dress it up!
oh yeah when you have shitloads of marketing money you can do some nice stuff
just look at all the magic leap marketing stuff
4.5 Billion in venture cap raised.
Rony Arbovitz is a master of smoke and mirrors.
That first reveal at TED... legendary
$1.4 billion not 4.5 right?
not sure how you could have $1.4billion in cash and only be worth $4.5 billion... its insane
im sure they must have given away at least 75% of the company
its just because they do it in multiple rounds so the actual company growth figures and multipliers are all out of whack
Yeah true
Things I learned about the VR community today:
Would rather not see the content, than to see it be a timed exclusive content.
strikethrough
Doesn't know shit about the economics of indy game development
hah did you learn that from the community in here or just the VR community in general?
Reddit ๐
About the Arizona Sunshine exclusivity debacle.
Lots of angry pitchforks not going down after devs rescinded the exclusivity.
The the developers come out and say something?
oh yeah i saw something on that. I don't quite get what they did there. locking just to i7 or something?
Basically intel paid them to make extra content
they took the money and made it, added it to the game
and told players only i7 players could access it.
A really dumb way of doing it mind you.
What they should have done was put an unskipable intel overlay / start image for people that don't have the processor but let everyone access the content
Players understandably got angry. Backlash, blah blah, then the devs woke up and patched in the fix.
then they get their money, intel gets their ads and people are happy
Intel should have known better
so it was just extra content in there locked only to i7?
as should ahve the dev
Timed exclusive
Locked to latest i7s. Even the 4790k I have apparently didn't make the cut
Dayum
but not the full game, just some of the content or something?
And it was announced after people had already bought the game
@pearl tangle yep
It was the "hard" mode with extra weapon unlocks
Well, it wasn't announced so much as discovered.
It sets a bad precedent if we just let hardware manufactures time gate game features or games
weird way to do it. why not just plaster intel i7 all over in game assets or something
Yeah that would have been the smart thing to do, or just do a long intel ad in the beginning
Don't know. Dumb deal the devs made.
well they have been doing that for ages with nvidia and amd. hair fx, turblence etc etc
yeah but that is more opt in
game looks decent though actually
Setting dumb deals aside...
If the i7 actually did make an improvement I think it would be different
Essentially the point to be made is, devs take this money so they can add more features to games.
well so was this by the sounds of it.
So that kinda makes sense right?
Where as consumers see it as taking away features because someone paid for its exclusivity.
they didn't stop people doing it they were just saying you can do it tomorrow but i7 guys can do it today. same as going guys with a 1080 can have fancy hair and particles today and if you go and buy a $1000 GPU you can do it tomorrow
Which also kind of makes sense.
But even when unrelated devs came out and said that... yo, we need this money to survive!
People were still - fuck that, just make the game not exclusive!
yeah people bitch a lot thats for sure
I think it wouldn't have been as big of a deal if it was announced ahead of pre-orders
it's annoying but thats how they survive. i wish i could play ps4 and xbox exclusives on my computer too
People think outrage and anger can be mitigated.
Someone was saying - if only devs were more open and honest
oh just reading stuff on it
and I was like, yeah, I agree, open and honest viewpoint
its bonus content anyways that they didn't even advertise. so not like they were stopping people doing stuff
its like hey people with this here is a bonus
and the guy was all - well, it doesn't matter what your situation is, tough titties.
So... I get why devs just shut the fuck up and hope it all blows over now ๐
you can't win!
I think going silent is worse
yeah have you been following kingspray?
But at least you can pay to stay alive a bit longer ๐
Yeah Kingspray should have just been more forthcoming from the get go
hah yeah complete silence the entire time
They saw what happened to super hot and figured they should shut their mouth but I don't think it really helped
Well, superhot and kingspray are fantastic examples of what I'm talking about.
I rather know something is exclusive from the get go so I don't have to think about it anymore
Damned if you do, damned if you don't. Oh well!
superhot told people
kingspray got a day past its intended release date on steam and rolled back to coming soon. didn't say anything. then 3 months later releases on occulus store
then 1 day later comes out on steam. devs never said anything at all
Just let the community make the noise. Don't even have to do marketing!
hahahah
๐
i think if they just came out on the release day and said sorry guys we are going to have to delay by 4 months to make some fixes. oh by the way you will also get multiplayer and etc etc when we launch now
I think the key thing is to set proper expectations. No Man's Sky is another good example
No Man's Sky is not in the same league
yep. people don't deal well with dev silence either
No Man's Sky is amazing in how fucking bold a dev can be in lying like a motherfucker
similar. they didn't say anything at all until they came out and go oh heres a bit of an update
I'm just really glad Steam allows for refunds now.
So far I've only had to refund VR games but only because they made me motion sick or just weren't fun.
Yeah, I've refunded VR games for shitty locomotion.
Before VR I never really had a reason to refund anything.
To be fair, I refund a game if I didn't enjoy it.
never tried to refund something. but probably could for a lot
You have to do it before 2 weeks or 2 hours played
At least within a few minutes.
Like some stuff is just... yep, this is trash.
Like there was this one zombie game using the simple art assets
Stuff popping up 1m away from you
bad locomotion.
I was just... welp. No dice.
Stuff popping up 1m was ridiculous. You walk towards it thinking it's a path. Nope, it's not.
oh yeah that was garbage
soi much vr crap just jumping on the chance to make a few bucks
and yet steam told me they wont release my projects on there because they are marketing stuff. even though Qantas did a shitty app which is purely just playing 360 videos and I did a cool 1 thats proper room scale with interactive elements and realtime shit going on
Really?
Wow, there are so many crappy VR games on there too
Fuck. Wonder if I'll be able to get my stuff up on steam now ๐
Just do another wave shooter those seem to go right in
It's not a game. It's a locomotion system!
yeah i just sent them a link to the 360 video saying we are looking to release a free version on steam. and they said they don't think its appropriate since its marketing
the hawaii thing and the qantas thing and the shirt shopify thing are all just marketing. at least mine is an interesting experience for people that isn't pure marketing like the others
Were you going to eventualy do paid?
The shirt shopify thing was great
Ikea is basically just marketing too
the qantas 1 is a 360 video player that is just the videos they have on youtube
I'd like to put what im working on steam much like budget cuts did. A short demo to get feedback and allow a way to update it. I don't want to just go into early access to get stuff onto steam.
and potentially this 1, modified a bit more and adding in high scores and leaderboards but so far it would still be the most comfortable skiing game on steam even though its a brand thing
thats just a quick 360 video i threw together this morning, not properly done yet
Well, definetly more worthwhile than the qantas stuff ๐
I guess the question is more... how the hell did qantas get that stuff on steam?
hah yeah exactly. there is the shitty basketball 1 with gatorade crap all over it or something too and it got there
thats pretty much what i wrote back to steam
Yeah, I should be safe with my demo.
It's not marketing in that sense.
But it is 'marketing' for my own company/services.
@clever sky do you have a demo on there now or you are currently working on one?
anyone seen working vr implmentation in 100% c++?
Gonna package it up soon
and upload it and get you guys to test it
But going through the process of setting up my company
i'm working on one now but converting from blueprint can be tricky
and getting it on steam soon
viveport will take a lot more stuff a lot easily anyways so I will throw my stuff up there first and then get steam to stop bitching about it once they see the proper interactive 1s
Hey folks! Check out this mod for the VR Funhouse that @fresh canyon, @neon egret and I made for the VR Funhouse game jam back in September: https://www.youtube.com/watch?v=Oky_uTI2Spk
It's on the Steam Workshop now: http://steamcommunity.com/sharedfiles/filedetails/?id=806212885
Steam Workshop: NVIDIAยฎ VR Funhouse. THIS IS A TOTAL CONVERSION MOD AND DOES NOT SUPPORT ANY ADDITIONAL MODS
PLEASE MAKE SURE TO UNLOAD/UNSUBSCRIBE YOUR OTHER MODS
------------ Story ------------
We bring you back to the times i
Haha... that moment you go to build your project and it all fails so spectacularly!
As a novice developer, this seems like an oversight that could've been remedied by more frequently building the project during development...
Yep. I'm just going to rebuild the project by migrating it
and see how it goes.
I suspect I'll be asking many questions over the next couple days!
Maybe even better is have a migration project and a editor project.
Because I have a shit load of extra content in the editor project
What's a good file sharer for distributing a multi-gig project?
Directly relevant to you guys because that's how I'll be linking my demo to you ๐
AWS if you don't mind paying for bandwidth
Any good free ones?
mediafire is decent
Gonna try get it on steam for public distro
I don't like Mega myself, but others might
onedrive?
ive used google drive to send builds
a few times
another good idea is to put it on steam XD
if its a VR game you can put in on steam without greenlight
then you just send steam keys
pros include that you can add steam DRM, and updating
Ah ok. Well... I guess I gotta go watch some tutorials on how to go through that entire process.
google drive is great for private releases
Why's that?
I just looked up 'best file sharing sites'
and got a site called... sabercat hosting?
Some of you were speculating and asking recently how many copies of Gunjack CCP has sold. I told you it was way over the speculated 30k you were throwing around..
Well my CEO recently said in a talk that we've sold over 500k ๐ so take that to your estimation models for the market.
Nice
Tipatat tweeted this straight from one of Hilmar's talk
https://twitter.com/tipatat/status/806674192221413376
whaaaat?!
wow
well, hopefully getting all parts for my new PC this weekend and then I can get back to UE4 dev for Gear VR ๐
I have been touched ๐
or is that be-touched?
I got my Touches.
No need to be touchy about it
I think they're a touch better than the Vive wands
...ok I think I'm done
EveGunjack sold 500k copies?
Welp.
Gotta get that mobile money.
Go buy me a Pixel and daydream tomorrow. ๐
there are people who pay for mobile games? still don't get it ๐
VR mobile games
@clever sky except 500k+ copies were sold on Gear VR ๐
Yes, and that gold rush isn't going to happen again on the GearVR ๐
and with now 2M+ userbase you have better chance making money with Gear VR than Daydream
actually Gunjack was criticized from day 1 - it's basically a Galaga with nice visuals
Seriously though, Gunjack was early in and a decent quality at that point in time.
yet, that didn't stop people from buying it
well, good luck with Daydream ๐
Now? A lot of users have stopped messing around with GearVR, so you want to think about daily active users rather than total userbase
less devs on Gear VR - better for me ๐
Well good luck with GearVR! ๐
The other big reason for me to develop on Daydream would be because of the motion controller.
The work I've done essentially requires one. Without it, I'm just making crappy nothing special games, and I'd rather just stick to my day job if that's the case.
but Leap Motion is coming to Gear VR
Integrated?
I doubt that
Because developing for a peripheral on top of a peripheral is just not good business.
either Leap Motion doesn't want to be sold, or no one wants to acquire it ๐
while I really want motion controller for Gear VR, I don't see why one can't make a good game without it
Because I want to make a dungeon crawler hack and slash that involves movement and moving your hand around?
You can make a decent gamepad only game.
Sure.
But that's not what I want to do ๐
"According to Leap, its goal for the new module is not to package and sell it in Best Buy as the โLeap Motion 2.โ Instead, the group wants to work with VR headset manufacturers themselves to integrate the hand-tracking sensor directly."
Best of luck to Leap Motion. Really. Great tech when it works. Just... for some reasoon hasn't been incorporated yet, even though it'd be great supplementary control scheme.
it would be good for gearvr where you dont have other controllers
the question is - does it even work ?
I hear all kind of stories ranging from "wow" to "garbage"
Since the Orion update it's been pretty good.
i have one but theres not a lot of software out there for it
So long as you're ok with the limited tracking FOV
i have yet to stick it to the front of my rift
I mean, UE4 + Orion
It's just not robust enough to base any real development around.
that jelly fish demo was cool
It's only good for experimentation work at the moment.
And I think most devs are thinking... that's neat, but do I want to spend the time and effort playing with that tech when there's no market for it?
not sure if I'd use it for picking up items and holding something
And while leap motion can do some stuff you can't do with motion controllers.
The most important things like picking stuff up and moving it around can easily be done with motion controllers.
Hell, I'm getting a good deal of hand presence going by aligning the hand mesh correctly to the controller (so it's oriented to your hand positions as you'd hold the controllers)...
and having physics collision on them so that when you touch something, things react to the virtual fingers.
which in turn feels like things reacting to my own fingers
Like... the area in which finger tracking would excel... are also sadly the areas that would most benefit from tactile feedback.
Like typing on a keyboard.
Even if you track fingers perfectly, it feels whack typing with no resistance.
You can try how bad that is by getting a laser keyboard ๐
Or spinning a pencil. Just not happening if you can't feel the pencil!
how good is it at gesture recognition?
seems like creating good gestures for it would be key
Not sure. Haven't played around with it enough.
that's why I'd only use it for interactions (and gestures if I was making a game with spells)
by interactions I mean pressing buttons, pulling levers, etc. sort of things
also attacking AI with bare hands should be cool ๐
(and if they would allow to use custom meshes for hands, then weapons can be used too)
You know what... leap motion would be a great addition to Google Pixel now that I think about it.
daydream doesn't do positional tracking and only has a single controller.
no way.. there is no room for it
Daydream was meant to be non-expandable
and Gear VR has more room where to put it
๐
But with leap motion and daydream controller, you get both hands positionally tracked within a cone and some buttons still.
I don't mean from a it can be done now point of view.
Only that the two has decent synergy.
but Google has Tango.. Probably even better than Leap Motion
just massive and expensive
AMD have announced the release of their latest Radeon software suite, snappily titled โRadeon Software Crimson ReLive Editionโ and, among many other features, the new suite brings with it support for Oculusโ newly introduced VR performance enhancement feature Asynchronous Space Warp plus VR focused rendering features โMultiResโ and โMultiViewโ. AMDโs โRadeon Software Crimson ReLive Editionโ โฆ
Yeah. Great stuff.
can someone with Radeon test and confirm?
ha
good ol' AMD
I still wonder whether I should focus on VR, or port my old game to UE4 for Steam and XB1, and then focus on VR ๐ค
lol, talking about fragmentation
I guess Steam isn't a good place to be, so they decided to make their own storefront
Anyone here spent a lot of time doing networking in VR ?
vblanco probably is the One
OK, so I have this setup: pawn has motioncontroller components that have child actor child compoments that are set to a custom actor type to represent hands
I'm trying to do a Server_sethandtransform(Ftransform) in which i check of the owning pawn islocallycontrolled() and if so doing that call
I was lazy and used this: https://forums.unrealengine.com/showthread.php?116782-VR-(OpenVR)-Expansion-Plugin
VR (OpenVR) Expansion Plugin
Updated: 12/08/2016
11/13/2016 - Added template project.
10/10/2016 - No longer OpenVR specific, if platform does not support OpenVR the plugin no longer compiles out the OpenVR specific code.
It's not OpenVR anymore
Works a charm
but the server call doesn't seem to be fired, I tried doing a "SetOwner(the pawn)" to the hand class inside my pawn, but that doesn't seem to do the trick
Although, I'd like to hear other answers if people have any
So i looked at using that but decided against it for various reasons, its really good tho
Ok
so the RPC ain't on the pawn itself
No the HandActor (that is held in a child actor component)
so, your structure is
Pawn
- Motion Controller Component
- Child actor component
anyhow, is the child actor replicated
yep, panw is bReplicates = true and the child actor is too
so it all works fine on a listen server, but not dedicated which makes me thing its just a lack of net connection on the client
err
hmmm wait, i think i have a hunch about something i might be doing wrong
is the child actor component set to replicate itself, not the actor it represents
I had a weird issue like that iirc
oh, no it isnt!
good catch, let me try that first
TBH the child actor is so tied to the pawn that i could theoretically just route an RPC through the pawn like Server_MoveHand(bool IsLeft, FTransform transform) or something, but its more just trying to get an understanding of networking as a whole
@spring pond You should put every "Motion ControllerComponent" to -1 player index
that way it doesnt work
but if you are possessed by a LocalPlayer, you set it to 0
so it works
wait, then replication of the motion controller will just magically work?
then, from the owning client you basically call "server_updatetransformations()" with the transformation for head and hands
hmmm ya
and that updates the transforms
that's just so that you don't end up moving everyone's motion controllers
exactly
ya
and controllers are only moved locally, so you need to send that info to everyone
generally UE4 networking isn't so bad, but this motion controller/HMD pos stuff is clunky
my recomendation is that those replicated transforms are not "directly" added to the actual transform
but you interpolate to them
from the server_updateVRTransforms()
you set "desiredLeftHandTransform" and so on
and on Tick
you interpolate into that desired location and rotation
of course, DesiredLeftHandTransform has to be a replicated variable
for the child actors...
they are a huge damn issue
not sure in 4,14 with the improvements
but they werent replicated on my end
so i had to spawn them server-side
and then the player "attaches" the actors themselves
I can tell you a mockup of how I ended up doing this entire replication thing
ya my pawn manually init's them in code
I don't use variables nor extra stuff
what I do is,
I set motion controller components / camera to replicate, and then I just send a transform to the server and then it sets components' and camera's transform
that way you just use the components or the camera and avoid extra variables
variables let you do interpolation so it looks smoother
well yea
@blissful heath So, yeah thats what I'm doing for the HMD pos right now and its working perfectly, it was just fucking up with the hands and I assumed it was because they are this child actor (that is spawned on the server)
although, bare in mind updating, i.e., camera's transform server-side won't visually update the position on clients' end but the position and rotation variables
(wrong adam ๐ )
oops
OK, thanks for the help/info guys I plug away at it a tiny bit more and see where i get
hopefully i can knock this out so nobody else on the team ever has to worry about it ๐
โค
I don't get though why you'd do that on the child actor itself
rather than handling it on the pawn
do you call the RPC on tick? @wicked oak
fine, just read you mentioned you do earlier
Kraver that you? Or different Adam? ๐
different I guess ๐
too many Adam's in VR ๐ hey
@sullen stirrup Yeah, I think I may rearchitect this to be more like your way and just have a single RPC on the pawn. I'm always going to want to mirror those three things anyway
just wondering, it feels a bit more handier having it this way imo ๐
https://twitter.com/cognitiveVR << they just released UE4 support
The latest Tweets from cognitiveVR (@cognitiveVR). Analytics for Virtual Reality. Our free Unity SDK is available now: https://t.co/wxlEeGCOoy. Vancouver, British Columbia
(I am guessing as an integration)
Immerse yourself in an epic sci-fi space shooter set in a cinematic VR world.
I wonder if it's any different from the one on Gear VR
its a whole new game
I figured ๐ But how really different is it ? From video on YouTube looks somewhat exactly the same (same mechanics for sure), but with controller support
I havent been following it that closerly, but I know they been working on quite a lot of new mechanics. So its more than just "different enemies"
Im not sure what has been publicly announced/shown, so I'm being careful here about what I say ๐
it's been released already, so ..
(I don't know how "embargo" works in these cases)
Im not sure what made it into the final game from what I've heard ๐
aye ๐
when will Epic finally merge Dev-VR to master again...
the last merge was 69 days ago now, I finally want to get all the new stuff from there ๐ฆ
hello fronds, where do you all go when you want to find a decent animator/rigger? polycount for instance?
(for freelance / small jobs etc)
you can also try #career-chat
FYI dudes, moving everything to a single function to replicate HMD + L/R Motion controllers worked just fine and made way simpler code
๐
@heady parrot how many people worked on Gunjack 1 ?
and how long did it take to build ?
so did epic say anything about having any plans to support multi view / multi res rendering directly in ue4 (support both nvidia and amd) for rift and vive on pc?
not that I know of :/
and on gearvr it's called MultiView too and UE4 supports that on gearvr, right?
since nvidia supports both of it in their ue4 branch I guess that it can work with both deferred and forward in theory, right?
Nvidias "Lens Matched Shading" only works on Pascal (GTX 10xx), but does Nvidia have something like the MultiView stuff?
@full junco yeah, UE4 supports multiview on Gear VR, but it was done by Oculus, not Nvidia
ok
so on the ARM website they write this:
What is multiview rendering and why is it useful?
Virtual reality applications need to render all their scenes twice from different view angles in order to create the illusion of depth. Doing this by simply rendering everything twice with different view and perspective matrices is not optimal, as it requires setting up a mostly identical draw call multiple times. The GL_OVR_multiview extension adresses this issue by allowing one draw call to render to multiple texture layers of an array texture, removing the overhead of setting up multiple draw calls.
Nvidia seems to call the MultiView stuff "SINGLE PASS STEREO": http://puu.sh/sJ6Gb/7dd6891d58.png
but I can't really find any information from nvidia about that
also not mentioned for the ue4 VRWorks branch
well, I don't know the detail, but I know 4.14 should have multiview for Gear VR as an experimental feature
ok here's a thread about it I just wrote: https://forums.unrealengine.com/showthread.php?130950-Native-support-for-MultiView-MultiRes-rendering-on-Rift-Vive-VR
Nvidia and AMD both support MultiView (render both eyes at the same time, like instances stereo rendering for GPU) and MultiRes (render the edges with lower res than the center) rendering for Rift and Vive.
Nvidia offers their VRWorks branch of UE4, but that's stuck at 4.13 currently.
Now it's nice that Nvidia offers that branch, but honestly it's quite useless when its stuck at 4.13 while I prefer to use the master branch for VR to get optimizations and new features as fast as possible.
anything there that I should edit or does it make sense?
epic always makes such decisions dependent on "community interest". so its important that people ask about it
why they don't support it? Because Paragon, UT, Fortnight and RR don't need it
RR would need it
ehh, I don't know what makes you think they do what's in community's interest
because they do
VR isn't that big yet and UE4 VR community is tiny compare to the rest
VR is the future
so, right now interests of major part of the community and Epic cross. And it's not VR, sadly
and epic is very focused on VR currently
then wait and see
UE 4.99 might have it all
๐
I was in the same boat with GEar VR (and still is to some degree)
now UE4 has almost everything for Gear VR
see, so they listen to what people need
I was asking it in 4.10, so... Just wait ๐
4.10 wasn't too long ago
feels like it was
feels like yesterday to me
lol, ok
Has anyone else had issues when using the oculus plugin in your project and performance?
so, then by 4.18 you shall receive
If I enable the plugin, package, creates a session and opens it, performance is horrible
If I disable the plugin and do the same thing, everything works just fine
This is using both the Vive and the Rift btw, no difference
what so the vr side of things still works with the plugin disabled?
yeah totally, the rift runs on openVR
But I can't get access to some stuff like ASW and Touch specific input
And obviously, can't package for oculus home
hah bloody oculus
There's also the small issue, that if I don't use the oculus plugin, when the engine receives "HMDType", it just receives "SteamVR"
ah yeah that is an issue
you doing custom controllers for the different 1s or have you just got it locked to the gun still?
Hmmm not sure I understand the question
well are you showing a vive wand or touch controller in your game
Atm I'm pulling the data from steam and fills a procedural mesh
Which won't work on Oculus Home...
So Ideally I want to know which HMD/Controllers in use - set BP_ControllerTutorial Mesh to the correct one
Using steamvr is cool because it's a generic API. IE "Show Trigger Button Tooltip" can be dynamically shown in the right location of the button
And it doesn't matter which controller you're using
I just want to be able to have the consumer decide on buying through Home or Steam, whatever they're comfortable with
depends if somebody is using the stupid frying pan controller or something that would be rather weird in the game hah
Oh man.
Magic Leap... I was slowly getting suckered in by this shit too.
I mean, I was still skeptical of the entire business foundation... but man alive, didn't think Rony Arbovitz was that good with the smoke and mirrors.
But at least one of these videos โ showing an alien invader game that let the wearer of the supposed headset or glasses make use of real-world objects โ was created by visual effects studio Weta Workshop. Prior to today, it was believed Weta had simply created the visual assets for the game. However, The Information reveals the entire video was created by the studio.
8\
Oh well. Looks like Michael Abrash was right.
AR is a long long time from becoming mainstream
And it'll have to go through VR to get there.
i thought it was always obvious weta did the whole thing?
magic leap just has the display tech ahead of the curve
microsoft has the spatialization stuff pretty nailed but not the display
ha, I told you so..
@pearl tangle they had a kind of display tech ahead of the curve.
One that they couldn't minituarize apparently ๐
yet.
Yeah
thats why people have put hundreds of millions into them
nobody expected it was going to be something consumer ready in 12 months
Well, that's what they were going around telling people.
We've got this ready to go to manufacturing
they have developer kits out there
yep
i just had to get my entire team to go through a bunch of crazy over the top NDA for Samsung too for something that im doing for something at some time....hah
Hehe
hopefully get to do some stuff with Space X next year though ๐
@clever sky @pearl tangle comment here and tell epic you wanna see it in ue4 ๐ https://forums.unrealengine.com/showthread.php?130950-Native-support-for-MultiView-MultiRes-rendering-on-Rift-Vive-VR
Nvidia and AMD both support MultiView (render both eyes at the same time, like instances stereo rendering for GPU) and MultiRes (render the edges with lower res than the center) rendering for Rift and Vive.
Nvidia offers their VRWorks branch of UE4, but that's stuck at 4.13 currently.
Now it's nice that Nvidia offers that branch, but honestly it's quite useless when its stuck at 4.13 while I prefer to use the master branch for VR to get optimizations and new features as fast as possible.
hah amd only came out with the drivers today?
@pearl tangle yeah ๐
@full junco Good post
@real needle thanks!
I'm letting my team now that they should chime in
great ๐
I will throw up a query on the UDN about it as well actually
Awesome
@pearl tangle great too!
done
UDN is just like the answerhub, right?
with the difference that it isnt public and I guess epic responds faster
@full junco Cheers. Good man.
Was playing Smashbox.
Bloody awesome game that kinda got lost in the whole Arizona Sunshine kerfuffle
thanks!
I tried Arizona Sunshine earlier. i7 stuff aside it seems like a very cool game. I'm sure the teleport haters are going to hate on it though.
Teleport hater here
there are "teleport haters"?
The weapon aiming seems off to me though
It's a good game inspite of its weak locomotion mechanics.
teleport is the only good way to move so how can anyone hate that
Like I look straight down the sights line it up and no headshot
But really... it's the first of its type. A lengthy FPS SP campaign.
in VR.
That's not a mod.
(Doom 3)
Eh... certain parts were quite grating ๐
I'm only an hour in
But not terrible!
overall that voice acting that is.
@full junco Well, even among teleportation options, there are ways to do it better and worse.
Dash = more immersive and slightly more motion sickness inducing... but not for the majority of people.
@clever sky all teleports feel the same, what options are there?
Blink = less immersive, more disorientating, but slightly less motion sickness inducing.
You get a quick smooth motion from point A to B
Dash is like what Raw Data uses
blink is putting a black out frame between points
No. You have to understand the physiological mechanisms at play
It works pretty well in raw data, its not very many frames between point a and b
not just blanket rule out vection whatsoever because that's what they said at the start.
But dash works because... there's a lag to vestibular activation
i.e. visual motion under a certain threshold is meaningless
And that threshold is around 50-100ms
Where as blink, is not as good as it intuitively seems, because it is more likely to cause disorientation.
hm
In the sense that you lose a sense of spatial awareness
e.g. if you teleport behind a wall and lose visual sight of your previous points of reference
you become temporarily disoriented until you integrate the rest of the spatial data (by looking around or remembering where you were)
well then it's probably best to just implement both and let people chose what they want
Exactly.
But currently, the messaging is off target, and the appreciation of the nuances is lost to many devs.
So we end up getting a lot of disdain for teleportation as a whole, when it can probably be implemented in a way that's acceptable to more.
well I currently use linear teleportation in my game, so not that Parabel thing. so you always have visual sight to where you came from
I like that
Unless you're pointing in a direction with the controller that you're not looking in.
is yours like Vanishing Realms?
I haven't played that
Or your arm reaches around an obstacle
That is a great game you should play it
Or a few other edge cases
I only really played the lab and audioshield and hm, probably some more, yeah tilt brush
I actually do prefer the direct LOS teleportation as well
it's easier to point and aim than the arc teleportation.
I prefer straight but the arc works good for hills or stairs
I liked how Arizona Sunshine prevented teleporting over certain things
It was both good and bad.
yeah the natural length limitation is good
It also handled blacking out the screen when you click your head through something. I liked how it slowly faded.
I do that too
But kinda a good way to limit gameplay futzing from players.
the closer you come to stuff the darker it gets
There was some other game that did that a while ago but didn't do the slow fade and then i would be lost in the a solid color and had a harder time finding the way back
don't want people to be able to clip through stuff
but my implementation isn't perfect, what's the best way to do it?
Well... I don't know if I've perfected it.
I use a sphere trace but that can be a bit buggy
But I've got an extra feature that makes it extra... spicy? :p
I was thinking while playing to put up invisible walls over things you don't want the teleport to go through and make the line trace fail
Seems like that would be a lot of work though
That or build it into the actors like the road blocks they had
I'm using a sphere trace too.
the problem with the sphere trace is it doesn't pick the closest result
not the closest point
Oh. What's it do?
I guess I must've made my system robust enough to hide that fact ๐
not exactly sure what it did, was months ago that I implemented it
and it works well enough
just not perfect
Yeah. It's not perfect like you say.
Main problem I get is when you continue shoving your head through objects
and there's just no more stuff to collide with
it sometimes shows through
even though I'm actually pushing you back
at the same time.
no more mesh
I mean, it doesn't show completely through... but it fades back just enough to let you see that you're glitching.
At this point, I'm just... eh, that's good enough. If they're glitching the system because they can't take no, then that's not on me ๐
well yeah
going out on the other side is a problem for thin stuff
solution : don't have thin stuff ๐
Yeah... D:
I've been trying to solve that thin stuff problem.
It works if you're going slow
But you can 'bash' through
Problem is, the clipping code is from the plugin.
Oh wait no.
I did solve that problem
But introduced another one!
hahaha
Yeah. Probably best to design around those limitations.
I'm not at a stage where I'm balancing against those sort of things yet though, so not too worried about it yet.
no idea at what stage I am...
Well, I'm not developing a game yet per se ๐
I really would like to not wait forever with releasing it
but I don't want to release anything that feels unfinished
True.
the nice thing with just waiting is that ue4 gets better and better
and other software like ATW and ASW
That and you don't have to brave the wildfire of a vociferous userbase if you've slipped up somewhere! ๐
Well, if you take an exclusivity deal, they will wreck your shit.
I can't agree with that. VR needs devs of all shapes and sizes to grow.
And some have to seek funding from any source they can get.
Especially medium sized studios
I hate all limitations that aren't purely technical
But yeah, the i7 lock out was... bad optics.
Here's how I would've done it; made it DLC
and have intel give out free codes for i7 owners.
Bam, same bloody thing achieved.
Without the massive outrage boners.
and that game probably doesn't even need so many cores I guess?
I am actually filling all 12 or more threads with AVX2 stuff so yeah I should make it i7 exclusive, but that would be a technical reason then
In all the kerfuffle, comments about the game's buggy ass second half have been lost in the fray ๐
@clever sky intel with dlc codes would have been a way better way to handle it
@storm vortex right? Such better optics. Use established market norms to achieve the same end!
still think they should have just had a splash image instead though
and then not have it for the people that already own it
Most annoying advertisements are for things I already have
Well yeah. But I'm talking about more in the sense of; if that's what they had to do, then that's the way it should've been handeled.
Intel advertisements are getting kinda annoying.
Yeah, gamers hate day 1 dlc and they especially hate being locked out of stuff
why does Intel need advertisements anyways without any competition
Day 1 DLC has plenty of precedence though.
They would have been wise to just not introduce the feature till a few weeks from now and have it be an intel special, i think that still would have gotten some backlash though
i7 core lockout has no precedence.
dlc in general are bad
yeah, and was stupid in this market. That's such a small audience
I don't mind DLC if it is good DLC and worth the money
Yeah, delaying it would've been a wiser move
Deus Ex Human Revolution had good DLC. Dishonored had the best DLC I ever paid for
Put some gap between launch and DLC
Don't try to sell me horse armor though.
I think I have never ever bought a dlc. I bought expansion packs like anno 1404 Venice though
Skins though I don't care about
I don't mind DLC; as long as its cosmetics or mini-expansion kinda stuff.
Stuff that isn't pay to win
The whole paying for addiction mechanisms are morally iffy for me ๐
That's why I stopped playing Battle Field after Battle Field 2
Like buying keys to unlock boxes
I never had to deal with that stuff luckily. My gripe was battle field 2 had DLC that had better guns and if you didn't have them you'd get slaughtered
Well... not everyone can be minecraft.
And strike the zeitgeist at just the right time
I haven't played them but I have heard the Witcher 3's DLCs are very high value
that are probably good dlcs
That's from a more old school developer that gets things though. No DRM and great value for a product
Well, you say that, and it's true to a certain extent.
But what people don't mention typically is that the economics of working from poland
is very different from most of the rest of the world.
why?
Yeah that's very true
I'm not sure what game devs in the US make but I'm sure it's much more
they are not just some small studio on Poland. they are a major thing there. even the politicans are proud of those Witcher devs
Point is, holding up Witcher 3 in its situation like the model for the rest of gaming development
is misleading to fans and harmful for devs ๐
I mean... it's easy to hold up the best of the best and point to their success.
But the reality is, not everyone can be the best of the best.
And without a sufficiently large/healthy industry, not even the best can be the best the way they are.
Like the goal is to get the highest peak. Well, if you're building pyramids you need a wide base.
don't like that metaphor
hahah ๐ ok it's not my best.
But you kinda get what I mean right? The tools we're using and taking advantage of now.
That doesn't exist because of VR.
They all exist because it was built off the wide base of game development.
yes
Which is great. But there are something that we can't borrow from the existing gaming industry
vr would be super lame if normal games wouldn't have existed
yeah
So, to get to Witcher 3 in VR levels, or to find enough indy developers to strike at the anvil
so one sparks the zeitgeist like Minecraft did.
needs a wide base in the market.
I still don't know what minecraft did. it's lame.
It made the best digital LEGO game
I don't like Minecraft either myself.
But I can sure as hell recognize that they totally captured the moment in gaming.
Sure. But counterstrike alone doesn't an industry make ๐
I think Counter-Strike was the game that kinda trigger FPS ESports though
valve is like a whole industry ๐
I don't want to think about what vr would be without valve now
yeah, Valve hardly makes any of their own games
Well... it'd be a lot more Oculusey.
but they make tech others stand on
And ironically, probably a lot more peaceful.
lol
Due to the lack of competition ๐
I think had Facebook not bought Oculus there would have been much more peace
we would not have any room scale
True, true.
And also slower pace as you say
But facebook has accelerated VR development.
yeah if it were up to Valve would be another 3 years
I mean, we have the Vive BECAUSE facebook bought Oculus.
vr without room scale is like no idea, politics without lies or something like that
heh
vr without room scale is like no idea, politics without lies or something like that
This analogy doesn't feel right ๐
why? I like it
Hahaha... I mean politics without lies is desirable.
well yeah I get it now it's bad
But that's not the reality ๐
I would love to have a third big vr company (desktop)
A better metaphor would be - VR without roomscale is like... taxes without representation? ๐
taxes are something people don't like and even with representation they still don't really like them
vr is something people like
Haha. I think the best bet is to move away from politics as metaphor.