#virtual-reality
1 messages · Page 54 of 1
well instead of adding a component you just spawn it in the world.
but yeah simplest method is you do a hide on the particle mesh itself. spawn the sound and particles, put a delay the amount of time that the sound and particles play for and a destroy actor after the delay
That's my default behaviour...
And it seems that the destroy actor node is causing the judder
because the judder will simply delay by whatever amount I put in before the destroy actor
how about if you dont have the audio and particles going off?
@clever sky i spawn tons of projectiles..with sound and fx..and in some demos leave them around.. have no judder
doesnt seem right that that is the issue..
with low end spec card for testing
@pearl tangle Turned off explosion and sound, still happening D:
Really seems at this point to be destroy actor itself causing it
regular projectile component you are creating as your bullet?
Ok.
Can confirm that it is destroy actor causing it. Either the automatic lifespan destruction or the manual call
@pearl tangle Yes, first person projectile
doesnt make sense..
I know. It's weird as hell.
But I turned off the manual destroy actor
And set lifespan to 15 seconds
Fired 30-40 rounds
and 15 seconds later... it starts juddering 😛
what i did for a few cases was replace the impact actor with another with nothing more than the same mesh with a delay to destroy
ive had 30+ actors laying around after fire..and got no juddering on a weak card
Yeah, it doesn't seem like the number of actors is the cause here.
The destroy actor call itself is causing the juddering.
To be clear I'm talking about judder in the motion controls
i even have special logic, and lots in some cases, to pick them up and refire..
yea, have had no impact.
try spawning the projectiles from somewhere else in the level with nothing to do with your attached mesh
what card are you using?
just setup a spawner somewhere and spawn them from the trigger
well its probably more likely that the projectiles are still in some way linked to the attached gun somehow and its doing something odd there
could be..
The spawner definetly causes judder
hmm
I mean, it's spawning once per tick
But yeah, the HMD is fine (ish) but the motion controllers are juddering.
well, per tick could be dangerous
so not linked with your gun anymore, just something in the level and it still does it but only when its destroying it?
do like many a second, not a tick
What's easy code to limit rate of fire?
delay
a delay on your trigger
just set a bool "canFire" or something after you pull the trigger and on delay toggle it
yea you need a multi fire mode or delay mode if trigger held down..
Oh yeah
Delay worked fine
And the judder is even better
More defined somehow than per tick!
Probably because it's not continuously causing lag
well, tick is/can be a lot..
Specifically, I'm saying its much easier to notice the judder when I set delay (to 200ms), because it's not ticking/juddering continuously.
It looks like... it almost forgets to do late update when it calls destroy actor
again, no idea why...there has to be another factor..
i destroy shit like a projectile all the time, even rapdily, and dont experience this with a shitty card even
@winged shale I know you've had problems with late update! What are your thoughts! 😛
Yeah. Weird as hell
Well, I'm just going to work around it if I can
by spawning the explosion and sound off actor
But how?
try the profiler?
Because the standard nodes just spawn stuff like particle and sound as a component
Oh... put them in their own actors
and spawn them as actors D:
destroying the component shouldnt be anything on the gpu anyway since it should be freeing up space on the gpu. unless you have a tonne of glass or something in your scene and bad materials that its trying to re render a bunch of stuff
Yeah, I don't think it's a GPU thing so much as it is late update latching problem
Because everything else renders fine
Just the motion of the hand jerks
Ok. So it seems to be specifically
An issue with epics motion controller
Because VRExpansionPlugin's gripmotioncontroller component...
is fine!
with your same firing thing?
i have had judder issues with the motion controllers too but usually came about when there was other issues
i think i may have gone overboard with the marketplace
Yeah... I figured, well, if I want those things, they're good deals. But chances are, I don't actually want/need a lot of it
they come in handy for us to prototype and put in a bunch of things
have done a bunch where we have ended up just pulling apart a specific asset from it and using with other things
ended up using the navigation system from the Fish 1 to create realistic flying birds which worked much better than the actual bird packs we have
and a lot of it is just for learning things too
if $5 saves us an hours time or 2 then its well worth it. I think I have put about $1500 into the marketplace now
@pearl tangle looks like my library now after the sale. I bought quite a few of the bigger scenes. They take a long time to load for me and don't perform too well unless async reproject is enabed. Are you seeing similar performence in any of the large scenes? (I wasn't expecting them to run smoothly to begin with). My SSD was also almost full when downloading so I think all my assets installed might have gotten fragmented on the disk.
what large scenes are you talking about? I haven't actually tested any of the stuff yet. still downloading all the things
The Lake Side Cabin, Modular Building Set (City), Urban City, Modular Scifi Props (I think because of heavy reflection fx), Miliary Base, and especially the Cave one. (Underground Cave and Bunker) I cannot even get that one to run.
The best one I have tested so far perf wise is the Urban Construction pack. I don't think any of those scenes are likely optimized for VR, but I was curious how well they would perform.
what gpu are you running?
1080SC with 4790k @pearl tangle
tried lowering the display settings in editor with them? I assume their demo scenes probably aren't very good with the stuff but im sure plenty of the assets should be fine to use
They run perfectly fine in the editor, just when trying to view in VR. As I said though I didn't expect any of them to work that well but was just curious if you had any better luck. They do run alright with async reprojection enabled. I rather use that feature as a safetey net though
well yeah they should run fine on a screen. Tried with the new renderer in 4.14 at all to see if there is any difference?
I haven't yet, takes a while for me to load each scene. I just wanted to check them all out first. When you check that box (forward render enable) it takes a long time for scenes to convert the shaders to it.
yeah its not the fastest thing ever
but im sure if you dropped the display settings down and put shadows lower and scaling lower then you would be fine with them
If you change those settings in the editor will it also effect the headset as well? I have't tried playing with them yet. Up until today I have mostly been doing stuff with blueprints and didn't have any intense assets to work with.
yep it does if you are doing the play in vr in the editor
that's good to know thanks for the tip
Cool thing about having a bunch of assets from the marketplace is you get to go to a bunch of places in VR that most people have never tried 😛
well if your machine can run it evidently :p
anybody done much testing with the vr works VR SLI?
i picked up 1 of the new sli bridges yesterday to start testing out
I don't think anyone's done much testing with it, too small of a use case
Yeah. Anything where you can control the machine that the app will be used on.
yeah pretty much 90% of my work seems like it should benefit from it since I have 3 1080's lying around
Should be pretty nice.
Don't have to work as hard to optimize 😛
Or can have a high screen res setting without batting an eye.
i really need the day i can comfortably and accurately model in VR to come
much potential!
yessir
It's so meh
i need to be able to do it on a professional level
There's a poly limit of 25k polys
any current solutions are just "NOPE!"
And you hit it within 4 seconds of waving your hands around
but it takes time for all this stuff to develop
I know right?
@pearl tangle I haven't tried much with the hb sli bridge
I kinda want to work at one of these 3D package companies just so that I don't have to put up with the dumb UX decisions that they make.
@pearl tangle I'd love to know if you get any cool boosts and how. So far most of my stuff is pretty simple, so it's probably why I'm not seeing much boosts from pascal vr sli
getting some boosts at least?
no 😄
@silk lodge Have you tried set res of 300?
yeah!
No diff with SLI or no?
I was doing tests of msaa vs txaa vs fxaa over sli vs single
I had instanced stereo rendering on
I'm wondering if that defeats vr sli
I built a rig for vr sli, but then I got busy and haven't had much time to test it
I have had fantastic results with flex and flow and dual gpus
the scene I was testing had about 10 objects in it
so I'm guessing vr sli helps more with more objects
But I wants more resolutions D:
titan xp lets me run most stuff really high res
Just spawn a shower of physics cubes
high res + txaa looks really good in certain situations
you are definitely running the vrworks build for your stuff right?
smp?
simultanious multi projection
the new titan x's over here cost $1800 USD so really not worth it. waiting for the 1080 ti before i upgrade my home machine
Titax X is the 10 series titan right?
yeah titan x pascal
Hasn't it been called Titan X before? 😛
Goddamn confusing as hell.
so they're titan xps
got 2 of 'em the day cuz I didn't want to wait for 1080tis
and the 1080s were all sold out
Threw down the cash monies!
aww yeah
Useful so far?
I have a 1080 in for compute / physx / flow
Ah
it's been nice for gpu accelleration in non vive stuff
and it's been good in vr games in general
but probably isn't making much difference
for most things
4.0 super sampling 😃
Well, you'd want to take advantage of them.
But at the same time, you're sucking down power
I mean electricity for the luxury of it 😛
You running them on a 4k?
Fair enough. I had that setup. Kinda miss it.
Damn. You must've been running with CRTs for a decade+
Slight radioactive emission is nice 😄
I'm gonna get super powers!
I'm recompiling shaders for my project, and I will find out if turning off instanced stereo makes vrsli work faster
6.6ms with sps on
SPS?
Ah
@silk lodge did you manage to get flex fluid simulation working with multires?
@native cedar I haven't used multires - does it mess up the smoothing pixel values?
Sadly I have weird behaviors when using them together (funhouse source installation)
I'm struggling 'cause I need them fluid particles but also performance
it really takes a lot of power to compute
well, you need 1 compute (flex / flow / physx) and 1 graphics
I want to believe vr sli works
on the flip side, if it does not, I'll have 4 vr capable machines
'cuz I'll break up my titans 😃
so what is the best rendering settings you have found so far? 220 screen percentage + txaa?
@silk lodge Isn't the recommended machine for max settings on funhouse... 2 1080 SLI? or is it 3?
1 is def for physx
If it's 2... means they couldn't even get SLI working for their own thing
here's the real deal
funhouse does not take advantage of unreal's hdr rendering
and does not have much fire
you can get really rad fire
Dayum. That's pretty.
What's it look like in VR?
it looks great
volumetric?
Far out.
you can tweak the shader in realtime as well
I've done clouds of smoke 😛
Interactive?
yeah
you could probably do clouds, but the problem is that it's not shaded well
so it's best for emissives
lol
I made beach ball valley 'cuz hitting beach balls is fun!
You could loose weeks
it's really fun
they have the other system that can work well for clouds but not sure how it would like interactive 1s
people who demo it get lost in there
so are you running the 3 cards in there? 2 in sli and dedicated physx
nah i mean part of the nvidia stuff
Ah
i don't think its included in vr works
oh they do have it in there
nvidia flow
turbulence is in there too
wait no its not. thats just in the fun house 1 sorry
lol crazy
also, just to make sure I'm getting top results, I've got a board running those 16x 16x 8x
ah that might actually be why you don't get the full performance boost from SLI too if you are running them like that and for physx. Have you tried just 1 monitor from the SLI 1 and dedicated physx from the other titan?
I could try disabling monitors right now
i have seen multiple monitors do some odd stuff with the vive before so you never know
yeah, if you're rendering your primary display and vive on different cards, you're gonna have a bad time
also if the 1080 is running like that doesn't it bring down the max memory available on the titan x's to the 8gb or have they solved that problem?
well you cant run different cards in SLI anyway but i know previously there was issues with things but i feel like maybe im remembering it a bit out of whack since it was like 7-8 years ago that i was screwing with that stuff
maybe it was back doing cuda stuff
I'm getting 4ms with a 1080, stationary directional light, no skylight and a map with just a skysphere and some spheres painted as foliage. 220 sp + r.postprocessquality 3 (faster txaa) and forward renderer
did anyone get a better result?
6.60 instanced stereo / single pass stereo / multires rendering / lense matched rendering
6.98 single pass stereo / multires rendering / lense matched rendering
6.21
6.6 multires rendering / lense matched rendering
6.03
6.38 (removed extra monitors)
no sli
6.3
5.5
so with sli off
my scene is working better
=/
wonder at what point that changes. I know instanced stereo gives a performance hit up until a certain point
I never noticed any substantial change with instanced stereo...
epic's VR showdown doesn't even use it
I think it reduces draw calls
so in some scenes it could be great
but not if they're only using a few objects
they didn't have instanced stereo in there when they did showdown. that was made back on dk2
@pearl tangle do you have dual gpus?
Instances stereo isn't a magic bullet and can sometimes make performance worse
You need to do your own benchmarking and find what works for you
i haven't got them sitting in this computer at the moment but i have some 1080's i will rig up in sli when i get a bit of time
alright. I should package some of my funny flex demos and send 'em to you.
lmk when you've got 'em set
every time I change a vr rendering setting it has to recompile 5000 shaders
yeah shoot through a download link and il get them setup for tomorrow
send me a PM or something and I will give them a shot. always fun to try out random PC breaking stuff hah
yeah, benchmarking VR its annoying as fuck
becouse every setting changes everything
with nvidia branch, you have a fuck ton of settings
VRMultires-InstStereo-Forward
those 3 are enabled/disabled independently, and they all cause recompile
Crossfire does nothing to my knowledge
it can do stuff, its not implemented in many games
UE4 cant use it, for example
but it can use SLI for VR IF you use nvidia vrworks branch
I mean Crossfire does nothing for any VR implementation, to my knowledge nobody has implemented Crossfire support
SLI & Crossfire are dead tech anyway
Is there a new version of SLI coming?
Or is it dead simply because devs can't be assed supporting it? 😛
SLI support is close to impossible in modern pipelines, outside of VR
Doesn't play well with deferred?
it's pretty dead for games as so few consumers have multiple cards that it isn't worth supporting anyway
for GPU computing on the other hand, it has it's uses
Yeah fair enough.
still, SLI might see some resurgence with VR
AI training system basically
stereo instanced rendering with SLI seems like a neat trick to have for VR to me
SLI has two giant issues :
- you can find a single GPU that performs better than your SLI at an equivalent price, with better support, until a very high price range
- it's very hard to implement, say, motion blur or temporal AA on SLI with AFR
could potentially really help framerates
It can be super nice for VR, sure, but still not betetr than a single GPU
It does help for research, heavy computation, etc
The dream is dead!
imho there never was a dream. SLI was launched around the Crysis period, when GPUs just couldn't keep up
After 2010 that wasn't an issue anymore
I don't think a single GPU is going to be strictly better than SLI for VR
Well, it was always a dream, never a reality. The dream was SLI delivering solid 50%+ performance
the reality is it delivering spotty performance with compatibility issues everywhere
you can essentially do a considerable amount of the work for each eye on either GPU
And now it's not even a hope anymore 😛
@hard light Probably not, but if you factor in the game engine support ?
I mean, what use is SLI if it's only as good as one GPU ?
It needs to be much better for the price
the Nvidia UE4 branch will support it, if it doesn't already
yeah, it's a new feature / work in progress - I wasn't sure how well that was working yet
The point is, no one wants to buy hardware that performs 10% better on some games, 50% worse on others.
the gains should be somewhere in the region of 40-60% when implemented properly though
so we'll see what Nvidia pulls out of their arse for this one
So you're saying, a $500 SLI will perform 60% better than a $500 single GPU ?
no, compared to the same single GPU
but that's the point - no single GPU is up to snuff when it comes to doing VR well
if you want 90+ fps with modern visuals, you're gonna need better hardware
which is why people are also interested in foveated rendering of course
So a GTX 1080 isn't enough ?
SLI isn't about price / performance, it's just about performance
a GTX1080 is enough to run projects intended to run on a GTX1080
Yeah, basically SLI is a way to throw money at the problem 😛
and they look crap compared to a flat experience
You can't seriously tell me that there isn't a single GPU that works well in VR.
If you need a $2000 rig to run VR, VR is dead too
Haha... it depends on the dev. But a lot of VR devs are happy to target the baseline spec for better customer base
VR is enthusiast grade expensive shit
and because it's just easier to build lo-fi
even the baseline is a £900 PC + £760 VR kit, heh
lo-fi is also cheaper to build too, of course
At this point in time, you'd only need SLI if you were a dev putting out a VR app where you can control the machine.
My point is, SLI is useless unless you're building a computer with a budget > $2k
That the app is experienced on.
VR is niche, very niche, and it's going to stay that way for a long time
Yeah, I'd have to agree SLI is pretty dead. Too hard to make work.
And not enough users to justify making it work for.
until hardware catches up and things become affordable, it'll stay niche
If VR needs a $800 headset on a $2000 rig, it's going to be dead soon too
Nah. It'll be fine. Just a slower burn before a nice ramp up
I mean, I'm a game dev, I'm a longtime gamer, I would never put more than $1k on a PC.
I don't know why you guys are so happy to call everything dead 😛
I'm not saying its dead, I'm saying it doesn't have a future at that kind of price
Especially love the guys that are all; VR is dead! AR is the next big thing!
it's not going to stay that expensive, but hardware is not cheap enough to go massively mainstream
Well, you can certainly get away with spending less than $2800 right now.
it'll take a few years
heck, PC gaming used to cost that much once upon a time, and look where we are now?
But I think it's fair to say that the average VR player right now probably has spent that region.
Basically I'm super happy with VR because it makes GPU makers go nuts
Which I fully support
Modern gamers cry so much about cost D:
I remember when the first IBM Thinkpads came out and cost something like £7500 xD
If you can still remember old school gaming prices, you'll know it's a great time to be alive now! 😛
(adjusted for inflation, that's going to be around £12,000)
these days everyone has a laptop
VR today is where cinema was in 1920 - I think this is a fair comparison
So I guess you think PSVR is shit ? Because it's much closer to mass market pricing.
PSVR is still very expensive when you add it all up
PSVR is great marketing.
and it's okay, but it's very low res and visually the PS4 can't do much with it
It's only $400!
it's not $400 when you add it all up
PSVR is basically $900, so basically the Vive headset
@hard light Exactly!
PS4 Pro (£340) + PSVR (£350) + Camera (£35) + Controllers (£100) = nearly £1000
Which is kinda rich coming from me
but for the first mass-market VR kit, it's not too bad at what it does
Honestly. I think it's actually a good thing that VR prices were higher than expected.
And that's counterintuitive as hell.
if they'd have cheaped out any more the experience would just not have been good enough
But here's what it is; the higher prices set a higher anchor point for VR to work its way down from.
Gives VR more headroom to keep pushing the envelope
because the tech isn't really quite there yet to be - I NEED THIS - for everyone.
IMHO, they shouldn't have supported the basic PS4, since PSVR on the basic model isn't stellar - it's DK1 levels of janky
I'm not sure if VR headsets will be that much cheaper in the long term.
"long" being 10 years
I think they're going to get better before they get cheaper
@cobalt relic Well, we'll have more tiers, mainly from previous generation of tech and equivalent. Not unlike how iPhones created natural tiers through gen progression
that's a fair comparison
So, it'll be a trickle down model of affordability.
new mobile handsets are still £700-800 when they come out
Sure, but I reject the idea that Rifts will cost $200 in 10 years, for example.
they're just better
I'm pretty sure the CV1 equivalent will cost that much or less in a decade
Nah. The high anchor point allows them to anchor expectations of that going forwards.
a lot changes in 10 years
I mean it can always shift up or down. But when it shifts up, consumers bitch and moan
See gaming consoles
Which needed to go up as simple inflation adjustment economics.
But gamers still loved to cry about the price rises 😛
plasma TVs cost $1800 in 2006, now you can get superior ones for $500
not that people still buy plasma TVs, but you get the point
Luckily for MS, the PS3 just threw out the price anchor point and took a hell of a beating for it.
So the $400 is launched at was comparatively ok!
Thing is, unlike plasma TVs, VR don't live in a vacuum
They need games
You can upgrade your TV to plasma, you don't need better content
HDTVs needed HD content
plasma was the example, as it was a new technology
The content will come. I just don't expect big new AAA VR titles for another 3-5 years.
Not least because AAA content just takes a long time to make.
What I'm saying is, VR needs to have a market from the outset, so that developers make games on it.
niche markets are still markets
If there is no VR market there will be no VR market
But we'll get stuff like Fallout 4 and Doom to test the waters.
And it'll be great!
And so so much indy content.
Farm Simulator 17 absolutely crushed it this year, but it's a niche market game targeted at a small enthusiast audience
I'm really not convinced by Fallout 4 / Doom VR
¯_(ツ)_/¯
Remember that the Wii U didn't have games n it, and it sold considerably more than the VR headsets are doing
Hopefully VR will sell more
tbh, I think VR is going to outsell the WiiU given a similar timeframe
the WiiU was a disaster
It'll take time. It's a new medium... Can't expect it to have out the gate performance of established tech during their generation upgrades.
But if you take VR as a whole, excluding cardboard... it's probably already doing better than Wii U
The Wii U sold 6M units in the first year, and it was hailed a disaster
If you limit it to just Rift and Vive, it'll take longer 😛
it was a disaster because the investment proportional to expectations proportional to sales were bad
That's because it has a huge established user base to draw from and resultant expectations to meet.
it sold something like 15 million units over it's life when it was expected to do 100 million
How much VR headsets will be sold in the first year after Rift & Vive get out ?
PC VR is sat at around half a million units
My guess is under a million combined from Rift and Vive.
PSVR may outsell that already
Maybe. Depends on how many Sony decided to ship. Still hard to find I think.
Sony are struggling to meet demand, so it's hard to say what it will eventually ship
I think they underestimated demand
Have you been using VR much @cobalt relic ?
motion controllers are becoming hard to find in most territories
Yeah, I wanted a PSVR at launch. But I couldn't find a bundle for all th parts here in Australia
So, oh well, I'll just wait 😛
@clever sky Nope, I don't really care about VR. I enjoy it as a driver for better gaming tech, but I hate having something to wear on my head.
@cobalt relic Fair enough. Well to each their own. But I think the most telling thing for people that have used VR for a bit...
It's a great concept, just not my cup of tea 😃
I'm enjoying VR, but I'm an enthusiast for this kind of thing :p
I'm the target audeince, hehe
Or highly compatible with it (like Project Cars/Elite Dangerous)
I'm convinced it's a larger experience with infinitely better immersion than the screen
And I say that in the sense that... Doom 3 is a crappy game
Better immersion is not what I'm after though
But in VR with the VR mod, it's magnificent! :p
Doom 3 is neat in VR, but the mod has a long, long way to go xD
True! But the point is, it makes a mediocre game a better experience
Circling back around to my main point; it's difficult to go back to the normal experience when the VR experience is available.
I think it's going to depend on the audience
I mean, you can make the argument that it's not so convenient right now, and that the limitations means it's not as valuable to some people as others.
But in the future... where the tech is mature and evolved enough...
I'm hard pressed to see a future in which standard gaming survives outside of a niche dedicated to by then retro experienced 😛
My own gaming setup is a PS4 + videoprojector, so I do a lot of moving around the room to do stuff while a level is loading, or to say hi to my gf in the bedroom. Not to mention cooking while gaming, a dangerous experience 😄
Not to mention we get to watch eachother play
It's a different way of playing
Like... 15-20 years of tech maturation, where the HMD is light, the FOV is large, it can be tracked robustly, it can connect to your phone or console or PC, it's an AR/VR/MR device, etc, etc.
I'm mostly looking forwards to multiplayer VR games where one person is in VR, and one person is on the desktop
UE4 needs to hurry up and implement that shit, because Unity does it out of the box 😦
Doesn't Keep talking and nobody explode do that ?
it's like some kind of funky asymmetrical split screen experience 😄
it's very fun apparently
yeah, keep talking kind of does it
What I'd like is combined AR/VR
where you and people physically close by
are gathered around a table
with multiplayer vr users
and you're playing some sort of interactive table top board game
I'm looking forwards to proper AR tabletop gaming, but that's a long way off yet 😦
See how magic leap does.
need a decent tethered AR headset
Maybe it could be sooner than expected
none of this all-in-one unit bollocks
So you want mostly VR with a bit of AR? 😛
no, I just want the hardware to actually be capable of doing what it needs to do!
(and it would be cheaper since you don't need to try and cram top end mobile hardware into a headset)
So you'd think right? 😛
Maybe... it isn't because you can't get as much of economy of scale
when you're expecting to sell peripherals
mice, keyboards, monitors, all peripherals :p
Those are integrated components that were modular enough to be swapped out 😛
I'm looking forward for giant borderless flat TVs, more processing power on consoles, and uh, a more resilient dualshock 4
That's it, I'm a conservative now
@cobalt relic They're already pretty big and flat and small on borders
Basically you want same as the last 2 decades of gaming 😛
Come on! Raise your ambition a little!
We've pretty much plateaued on game quality.
I'll disagree, I'm very much limited by processing power on my own game, and that's on PC
Depends by what you mean by quality.
more processing power can theoretically allow you to do more, but nobody seems to really be leveraging it tbh
Mechanics typically don'r require processing power
But even on that front, I'l disagree
there are decade old games that still hold up pretty well today
take a look at Company of Heroes as an example
people still haven't made a better RTS :/
Maybe not flattened out completely, but certainly we're roundly in the slow growth era of improvement to game mechanics on traditional games.
I have been playing Destiny for two years now, and the overall game design is superb, especially the raid mechanics.
The exception here is VR. Because it changes the basic paradigms of input/output
The art could certainly use 4K and HDR
I think the only thing we've seen that's been disruptive in games is the idea of the large scale survival genre
Like Day-Z?
MOBAs have been around for a very long time
I still don't understand who these people playing MOBAs are.
it's only recently they got a label
Well... DOTA1 has been around for a long time.
Cooperative gaming ala Borderlands, perfected by Destiny, is the biggest improvement in recent years I can think of
ironically, the more casual audience, hehe
Blending of solo & multiplayer is really cool
It's basically MMOs with persistency boiled away
Nobody cares about the MMO part
Pretty much.
The part people care about is joining their friends after seing they're doing a cool mission
Like you can have the massive part... and it doesn't matter because you just spend a bunch of time with a few people running a dungeon!
Or doing raids with peopkle trying to guess how it works
You never see more than 12 people in Destiny anyway
I'm just really happy about the small-scale cooperative multiplayer model
Borderlands did it well, destiny does even betetr imho
You mean medium scale
Warframe does it too
@cobalt relic Oh. Ok. 😛
4 by default, 8 with a mod
I wonder if we can get VR and gamepad players in a game concurrently.
In multiplayer.
obviously you can
Maybe asymmetrical classes.
It's a great scale, you play with people who actually care and talk to you
but it needs to be mechanically sound
Because it'd be an awesome way to advertise VR
as long as you have good tracking of the VR player
they can be far more expressive
Gamepad vs mouse works well if you're not doing a competitive shooter
Or something that doesn't work on gamepad like a RTS
gamepad vs mouse only works if it's not competitive
but PC games that work best on gamepads are fine
Like third person games
e.g. Rocket League
not all third person games
Gears as an example would have been a disaster if they allowed competitive cross platform play
Stuff like World Of games should be fine too
But it was designed for the gamepad! 😛
who takes 10 seconds to turn around 360 degrees on the spot? lol
it was immeasurably frustrating for me
like trudging through treacle all the time xD
Oh yeah.
I was dual wielding in VR today in my own game.
Fucking hard as hell to hit both spots at the same time consistently.
How you do it game characters?
What hax you using??
I think SPT has a little help 😛
Actually I don't know. I do know that your fnger position can easily throw off your aiming when shooting though
it depends what you're used to apparently
And that some VR games do use a little auto-aim correction
I don't think SPT has any aim adjustment in it though
you'd see it when it does the trace based effects
I am investigating how they achieved the funhouse rendering on ue4.11
I am discovering interesting stuff!
these are the console commands that get executed in the funhouse game
I wonder what does that r.HairWorks.MsaaLevel mean
@upbeat holly get ready to join nolifecity
WELL THEN
SINCE WERE SHOUTING ID LIKE TO SAY THANKS FOR YOUR WARM INVITATION
YOU ARE VERY WELCOME INDEED
May the lighthouse beacons shine down on you, like (laser) light from heaven
I Bathe solely in the glory of vive....
I wanna do a little force pull when I grab the handle of an object, instead of having it just pop into my hand.
What's the best way to move lerp an object into position?
I'm trying to teleport it... but it's buggy D:
what happens if I hook up Vive/Rift to 670GTX ?
Ehm, no idea how strong that one is, but what do you expect to happen?
well, I am sure if I just have an empty room, performance should be fine
I just don't want for Rift software to prevent me from messing with VR if I don't meet min recommended specs
I am not trolling.. It's a legit question. If an empty room can be rendered at 120 fps for 2D screen, theoretically in VR it should yield 60 fps.
Which is what Gear VR renders at with a way less powerful hardware and not in an empty room
I just want to get jump start in development before I get 1070.
120fps flat is not going to be 60fps VR
but anyway, the 670 isn't a great card for VR, but you may be able to do some simple stuff with it
it's mostly the overhead of the rendering pipeline, theory breaks down when you have to set up all the features of the engine for a desktop renderer each frame.
670 will barely eke out enough to hit 90Hz. In the lab before we upgraded, we had a bunch of DK2 kits, and the 780s in them couldn't do it at 75Hz
so ultimately you're gonna have a bad experience on the 670
eeh, I don't know what kind of scene complexity you had there, but I don't see why an empty room level wouldn't render at 75 fps in VR
on 670GTX
ok go ahead then, have fun
I certainly would try, if I knew for a fact Oculus software wouldn't block me from using Rift with 670GTX (which theoretically could happen; I can see a pop-up saying I am out of luck)
if you yo forward with sp 120, the faster txaa and only baked lightning you can do stuff with a 670
you can go real low when it comes to rendering, also none of this matters if you just need to implement programmed stuff
@native cedar that's the idea - to build a BP framework, while keeping visuals confined to a room (or several occluded rooms). Once I have get 1070, I can focus on visuals as gameplay would be partially implemented.
in that case, no need to worry, if you use the proper settings it won't even look bad
aye, thanks
don't listen to the 780 haters, the 780 rulles
lol 780 doesn't exist
it does, I have one. I didn't imagine it
did... did you take your pills today?
wtf
The new GeForce GTX 780 marks an evolution in gaming performance—an insanely fast, smooth, and whisper-quiet experience, every time.
The 8 series doesn't exist. Nvidia jumped from 7 to 9.
But, I'm pretty sure snowbro was just fuckin around 😛
870 wouldn't exist, but 780 does 😃
870M does exist
that's mobile
Well, the 8 series didn't exist in my world!
btw, do Vive/Rift have force feedback in the HMD ?
aye, just asking
of course not /s
Just asking because in Gear VR, since the phone is in HMD, you get to have force feedback on your face
so theoretically (I haven't tried it yet) you can have something slamming in VR in your face and HMD would vibrate
(I wonder if it's safe for the phone - might fall out of Gear VR)
force feedback is not vibration
force feedback is used in things like joysticks and steering wheel accessories - it applies a physical force on the yoke / wheel counter to what you are doing in order to apply realistic feedback
hmm.. yeah, it is.
no, it isn't
small motors in gamepad create vibration
Haptic or kinesthetic communication recreates the sense of touch by applying forces, vibrations, or motions to the user. This mechanical stimulation can be used to assist in the creation of virtual objects in a computer simulation, to control such virtual objects, and to enhance the remote control of machines and devices (telerobotics). Haptic devices may incorporate tactile sensors that measure forces exerted by the user on the interface.
Most researchers distinguish three sensory systems related to sense ...
"Haptic or kinesthetic communication recreates the sense of touch by applying forces, vibrations, or motions to the user."
vibrations
are you being anal about wording ?
No.
Read the article:
"In lighter aircraft without servo systems, as the aircraft approached a stall the aerodynamic buffeting (vibrations) was felt in the pilot's controls. This was a useful warning of a dangerous flight condition. This control shake is not felt when servo control systems are used. To replace this missing sensory cue, the angle of attack is measured and when it approaches the critical stall point, a stick shaker is engaged which simulates the response of a simpler control system. Alternatively, the servo force may be measured and the signal directed to a servo system on the control, known as force feedback."
google for "force feedback gamepad" and youget crap load of products that "vibrate"
highlighted the important bit in italics
I don't know what world you come from, but force feedback always implied vibration
since like first gamepads with force feedback , when the word haptic wasn't even used widely
force feedback is not vibration, it's not even open for discussion - this isn't me being pedantic, they are literally two different things. Force feedback is when the user is given feedback by applying a physics force to their controller
no, it's you being pedantic
gamepads advertised as "with force feedback"
have you even held one of those, especially back in the days ? It vibrates, that's what it does.
if they're advertised with force feedback, they are wrong
the whole point of FORCE feedback is that it uses FORCE - how is this not obvious?
i.e a joystick that literally pulls your hand
it's literally a motor, eccentric motor
Eccentric rotating mass (ERM) vibration motors use an unbalance load on a DC motor. Read about the different types and how to use them here.
but, keep being anal about it... I don't care 🙄
Force feedback, not vibration, is a thing
It's just a super expensive thing that gamepads or HMDs can't have
Like @hard light said
technically you could do it with analogue sticks, but I don't know why you would
used correctly it could be pretty good actually
yeah, I have a HOTAS that has force feedback
hardly any games use it any more though 😦
let it be known, that by falling down the rabbit hole I have discovered that it is theoretically possible to do a kind of force feedback with motion controllers through the use of gyroscopes
in theory you could do it in the HMD too, but it'd be heavy and vibrating all the time
it'd be like having a pair of lighthouses attached to your face in addition to the HMD D:
Wow! My neck really felt that twisting and wrenching! VR is so realistic!
lol
Can you imagine trying to dev for a full immersion exo-gyro-suit-thing from ready player one?
You'd have so many bruises from fucking up your vector maths and shit 😛
Apply force here... no, not that force oh god
In the end, you're like, F-you players, you can bear the same pain I did!
you could probably kill someone with a simple math bug
one inverted vector and it's all over...
ive only seen force feedback on racing wheels and flight sticks
never EVER on a gamepad
basically they do force against your movement
so if you are in a racing game and turn fast, the wheel can do force against you, or depending on bumps on the road
its NOT "vibration"
I agree with you here. I really want to see actual force feedback on finger triggers, not just rumble.
Imagine a programmable actuator that pushes back on your finger
normally, force feedback IS a kind of haptic feedback
the XBone gamepad "finger rumblers" feel quite good, but they dont push back
is just that its not going to happen at the moment, it tends to be bulky
so they just vibrate
also, a traditional rumble motor is just a motor rotating with a weight.. That is not what the Oculus Touch has
really?
what does Touch have?
i think its a piezoelectric thing
Touch uses a proper linear actuator.. Best way to describe it is like a speaker with a weight glued to the membrane
ah
its far better than what the Vive has
Vive only has normal motor vibration to my knowledge
it's not great, but it's kind of functional
its also higher frequency, and faster
Im working on some things to push the envelope of usage of haptics in games, hoping to publish a blog and OS code once I'm done.
Been working on "haptic textures" to simulate the feel of surfaces and give objects a sense of weight when you pick them up
damn
the Touch haptics is quite amazing 😃 its a step above what's been done before
i just do vibration when you shoot stuff
give it 12 months before the Vive has it too xD
or a small strong vibration when you wack something with the sword
thats what 99.9% of everyone does 😃 but you can do so much more
Vive v2
the new controllers
siggi, how?
my haptics are very simple, probably too much, but i would like to make them better
amber I'm counting on it 😃 and thats why I love seeing competition, it will develop the tech faster to mainstream
what are you doign?
to my knowledge you can't do much more than simple vibration in UE4 anyway :/
nothing disallowing you from doing direct API access
Im feeding haptic values procedurally based on what you are doing. Imagine feeding an audio sample to the actuator and then "scrobbling" the playback on that based on user movement and actions
and that works?
amber not true. with the latest SDK you can actually play back an audio sample through the haptics
yes it works. It's a bit tricky atm but I'm hoping to publish my work once done 😃 just too many things to work on atm
im adding the 3rd map and hard mode to DWVR
so it can start to be content complete for the beta/earlyaccess/whatever version
cool stuff
You can use float values to drive rumble with Vive controllers...
So in theory you could so something similarish.
With animation curves and what not.
it all comes down to attack and decay, ie how fast the haptic rumble device can respond to changes
the Oculus once runs at 150Hz if memory serves me right
the API changes are in the engine itself now, but I havent tested if they are hooked up to the SteamVR classes.
they already do animation curves
its playing back a haptic feedback curve
it has amplitude and frequency curves
The Vive ones?
yeah I need to play more with the Vive haptics, we are probably releasing our game on Vive too
How do you call them seperately? I just have a single intensity and duration node 😛
1ms is quite big yes 😃
thats super huge. thats more saving than I ever got by using instanced stereo rendering I think
So we're getting a nice present for 4.14.1?
yeah seems so!
and I'm on newest master and don't even have that yet because epic is so extremely slow with dev-rendering merges lately 😩
They recently said they are aiming to release 1.14.1 in about 2 weeks
who cares about when they release it, you can always just pick the commit 😀
but it is strange that they wait so long with the release of the first hotfix
well, they are prob busy on 4.15 😃 last time they talked about it they were aiming at the end of the year
which places it hella close to 4.14.1
yeah
I'm just using master, mostly because the RMC async collision stuff only works on master currently I think 😄
1st december should finally be the next dev-rendering merge, will be the first one since almost 4 weeks, so it will be yuuuge
oh you working with Chris?
I just use the RMC
its great 😃 I saw massive improvement when I converted my procedural mesh examples over from PMC
yeah the PMC is really outdated now
Chris took in a lot of feedback from me 😃 I was .. very frustrated about the PMC yeah
and the answers I was getting from Epic were basically "thats not the usecase we created it for".. /facepalm
What you using it for? voxel engine or what?
yeah voxel stuff
sweet 😃 got anything to show?
not yet, but hopefully soon 😄
so far the best one I've seen is Everquest Landmark. Really makes me wish we could do a big world with multiplayer like that
I mean more games like that 😃
ok, don't really know that
it was a dual contouring voxel engine
wait, what? EQ:L used voxelfarm? I never realized that
yes
nice!
they did have VERY close integration with voxelfarm team
i guess they licensed it with source and shared stuff
See how others have successfully used Voxel Farm in their game projects. Shows the user creations in Landmark and promise of advanced voxel physics by the Crowfall Kickstarter Campaign.
Damn, I might consider making something with this then 😃
EQ:L is voxelfarm 😛
voxelfarm is pretty amazing, I think @proper pumice has some live production experience with it (I think he mentioned a few problems with how it was being used)
Anyone wanna help me make a VR-RPG with it? 😛
yeah, voxelfarm is actually pretty solid now
Voxelfarm 2 was a fucking disaster
but Voxelfarm 3 is pretty legit
I'd love to hear from anyone using it in UE4
cool 😄
is the perf ok? and supports realtime modifications just fine?
its kind of the point of it
there is a trial, I'm going to check it out soon 😃
hay anyone had any luck with MSAA in 4.14? I have just turned it on, and everything looks aliased, and I do mean everything. Toggling the anti aliasing with "show antialiasing" essentially confirms that the MSAA is not doing anything
what am I missing? D:
i checked! it's 4 by default
@raven halo it only works with forward rendering, you probably try it with deferred
only aliasing I've seen is on specular and wrong normal maps 😃
sure?
deferred smooths over quite a lot of artifacts that become noticable with forward
100% sure of the fact that I'm using forward
out of curiosity, what screen percentage are you all guys using?
I'm using 150
holy crap
default atm, but I've done dynamic quality scaling in the past
high SP is super expensive
Valkyrie uses it for example
Oh yeah. I need to make an option to change that 😛
Does Unreal have a system to dynamically change screen percentage?
no
no, but there is some stuff for it in Oculus's latest integrations
its not hard to build that system though, at least for PC
console is a bit trickier
But with MSAA you got yet another variable to dynamically scale 😃
so scale screen res AND MSAA based on perf
the whole idea is let the perf decide those factors
its the method outlined in this talk
on Gear VR with MSAA performance can be solid 60 fps.. Not sure what's different between that and desktop MSAA
is it fine to basically call r.screenpercentage XX in ue4 every frame?
I wouldnt do every frame but you can get close to it yeah. The method we used was slightly different, but based on the same concept
for console you need to change the res without reallocating memory for the frame buffers
in the latest SDKs I think it does yeah
i havent tested, its probably a thing for the oculsu branch
not sure if its on by default?
its not
would be good if the SDK just does it
no reason why every dev would have to care about implementing that
what about steamvr?
but when working in VR you sometimes have to implement these things before they become mainstream
@clever sky what RPG are you working on ?
Not working on one yet. Still putting together my locomotion demo.
Having a bit too much fun with it, but it's coming along 😄
I'm still waiting on that! 😉
I put collectibles in it!
It's like the best collectibles ever!
Would be cool to do an RPG with minecraft elements in it 😛 especially in VR.
Dig with a shovel! Really dig!
yes 😃
motion controllersr + dual contoured voxels.. see some possibilities there .)
Alternatively, magic effects and explosions would leave nice impact craters all over the place.
then after some time all your are left with is a giant hole in the ground 😃
Valve is continuing to improve upon the design of the ‘Lighthouse’ Base Stations, the laser-beacons that form a crucial part of the SteamVR Tracking system. Future iterations are expected to become dramatically simplified, reducing size, noise, and perhaps most importantly, cost. VR is awesome, but it’s still expensive. Until costs come down significantly, the sort …
Hehe 😛
BTW, does open level unload the map you're moving away from?
Because I seem to be having a bit of a memoy leak issue after loading up landscape mountain...
depends on how you travel
How you travel?
not if you force it to unload
How do your force unload?
how you move between the maps, there are ways to keep things alive
I'm just calling open level on trigger?
don't know yet. I only worked with streaming level in 😃
but more often than not a "leak" is just a hard pointer somewhere
Hard pointer? What's that mean? D:
there is a node in BP to unload level, I am sure
You can unload streaming level
I'll probably have to go the level streaming route. Not sure if that'll fix the problem though.
not sure if you have to worry about hard vs weak pointers in BPs 😃 but basically if you keep a reference to something it wont get garbage collected
Hmmm. Good to know. But I don't think it applies in this situation...
As I haven't made any references to the maps other than through load level.
open level
Actually, it's quite bad. Even after I quit UE, it still stays open in task manager
and even after force quitting, the computer is still lagging and stuttering.
Oh well. puzzle for tomorrow. Going to sleep, night.
o/
Is it possible to make a living on VR market currently ? (game dev)
if you spam steam, for sure
im about to release my first VR game with some effort
ill probably do a couple more fast-games for steam
either you get lucky with a great vr game
or you do work for big companies that pay lots of monies
nobody cares about bad games
nobody
good games rise to the top fast
Good games need to also have single player support though. Some of my favorite VR games are only multiplayer and there aren't always people playing them.
I'm a big fan of Art of Fight, but hardly anyone plays it.
you've gotta get a bunch of vive friends to play the multiplayer games you want
I'd always first do singleplayer and later MP
but if I would release a game on steam then I would expect that I would work on that for a few more years before its considered done
I think that is something to consider though, not everyone has huge friend lists on steam or friends that have VR. I only have a few friends with the Vive and they are obsessed with playing Overwatch so hardly join me.
start hosting matches on http://vrlfg.net/
start adding everyone who plays the games you like
there's 11 in art of the fight right now. best get on there 😃
Working now, but maybe later 😃