#virtual-reality
1 messages · Page 84 of 1
nope.
Dang, what a pity. Thanks!
"One such platform that’s launching at MWC and GDC (since both shows are happening simultaneously), is Qualcomm’s new Snapdragon 835 VR development kit. This new Snapdragon 835 VR development kit features a 2560×1440 AMOLED display, 6DoF tracking, eye tracking, foveated rendering and many other performance and power saving features."
really not a bad thing
so like that standalone oculus ?
leap motion stuff for such a headset is a great thing
inside out tracking adn stuff
of course it has the FOV limit for the tracking
yeah
I don't know if it's going to be a retail product or if it's just for R&D and available only to selected devs through Qualcomm's dev network.
pimax, review said that's 60hz and is only 1080p upscaled
I mean, you can use cheap Cardboard HMD, any phone and Nolo VR to play PC VR games (Steam VR)
with motion controls and positional tracking
S8 may have a selling feature for me afterall
has an iris scanner
I apparently have very worn fingerprints, that's why the s7 fingerprint reader is near worthless when I use it
yeah
but I am sure S8 itself won't have all the bells and whistles as Qualcomm's VR dev kit
I'm super excited for that dev kit
the specs are impressive
I'm guessing that even if I could get my hands on one of those, we would need to wait for epic to add support :/
still 4Gb of RAM 😦
wow the qualcomms sdk is very reasonably priced
only 300$
😮
I wonder how much more the 835 version will cost
I wonder if Qualcomm offers UE4 plugin for that dev kit
I registered and asked them 😃
Hi guys, Need a quick advice. I'm planning to develop an application for Android Using AR [Agumented Reality] in UE4 what is the best plugin that has good support and documentation
@mighty carbon let me know if it turns out that they do have one, because if so I'll be ordering one! 😄
will do 😃
btw, just want to say that UMG sucks
that is all
I've taken a look with renderdoc...
and jesus christ
it already has aliasing in the render target
even before it's rendered on a plane and on to the final render
:C
the reason why is because the background is black
and it creeps into the white parts of the alpha channel
haven't tested it yet, but I think it will look better if I add a "background" UMG plane behind the text
at least then the only aliasing it will suffer is from the final render. At least the render target will be fine
it looks ok in Gear VR, but yeah.. It's one of those unresolved issues for mobile VR and AA
also, Create Widget is performance taxing operation 😦
I am guessing for mobile VR it's better to simply put all UMG icons in place and hide them. Then when needed - unhide them. The paid on this is there is no way to correlate inventory item in the array with widget/icon in the UMG
is everyone building their own VR keyboards? not seeing anything in marketplace or open souce
looks great
do they use some special fork of UE4? I am surprised performance is great (supposedly) with all that foliage
They've always had a custom build of UE4, since they added their own Global GI (and tanked engine performance in the base game at the same time)
dunno what they changed for their VR showcase, but hopefully a lot
I recall all improvements trickle down to Epic's master, don't they ?
Do y'all keep your textures at 2048 for the most part? Or bump everything down to 1024?
i have everytihng at 4k
then i push stuff down when packaging
but textures are created at 4k
Hmm, so how do you push it down at packaging?
@mighty carbon Arks engine improvements trickle down? doubt it for the bug stuff like their GI. They might commit bug fixes back to main but their engine version is far behind now.
I see
Valve today announced the Steam Audio SDK, a spatial audio plugin that the company says is made is designed to “enhance all interactive products, specifically VR applications.” The company has launched a beta of the SDK today supporting Unity, and support for Unreal Engine is on the way. Realistic sound is an important but often …
hi guys, are any of you guys doing hand collisions for vr? so the hands can't just move through meshes and stuff?
@eternal inlet that would create disconnect between your physical and VR body - not good
yes i get that... so bad practice?
sounds like it
just feels weird that you can wave hands through meshes
make them draw over
although it's also weird when they're not where they are supposed to be
what do u mean draw over?
that sounds even more confusing tho
that's what FPS games have done with weapons since the dawn of gaming
but this is VR
(so that weapons don't go through walls)
whole new set of rules apply
yeah, but it's still to a certain point
without anything to stop your physical body, you have to accept certain limitations beforehand
either going through geometry or drawing over the scene is IMO a way better than moving your physical hand freely, but seeing your virtual hand being stopped by virtual world
IMO
hmm
with ie. head movement
i feel its weird to get pushed back when moving into meshes
therefore i went with a solution that fades camera to black
with hands it's something different
i did a test some time ago with a push mechanic
that basically also allow you to push hands through walls, but to gain the feeling of pushing
think you tried a demo of it earlier? the fullbody stuff i made?
or maybe it was someone else?
ah
Interesting !
http://www.roadtovr.com/valve-launches-free-steam-audio-sdk-beta-give-vr-apps-immersive-3d-sound/
Valve today announced the Steam Audio SDK, a spatial audio plugin that the company says is made is designed to “enhance all interactive products, specifically VR applications.” The company has launched a beta of the SDK today supporting Unity, and support for Unreal Engine is on the way. Realistic sound is an important but often …
I think 4.16 is all about Audio!
seems so 😃
personally i look forward to play around with new audio features for sure
same! It seems like a new door to some amazing stuff to discover
from the article video it sounds similar to the occlusion stuff epic games demoed earlier
but im sure there are more going on behind the scene
exactly! I guess if these two are going to pop at the same time I'd rather user Epic's solution
yeah me too probably
just regarding to all the amazing stuff heard about their new Audio Engine . I have to say it was a feature that I've been waiting for a year
@eternal inlet I'm watching over your tutorials about VR collission, it's really awesome man! just one point about the Eject from Wall
yes?
I found that It can become a bit of a problem when you change the offset of user in Roomscale
because at some point you're dragging and offseting user's position inside the roomscale space
u mean because u end up in a corner of the room?
true, but couldn't really figure out a better way
as far as i could tell, it's about the same that would happen if i just left it to the pushback
or maybe im wrong?
right , it actually is ok for some projects ... but for example for my case ... I have a room in VR and I don't want user's offset to be changed in the room by colliding with walls of the room
yes I'd rather skip it for the room ... but !
if the user is not inside the room . then I'm actually free to mess around with it's location
so this ejection trick I think works best for Standing experiences
not for Actual RoomScales
my ejection trick or yours works best for standing?
Ah I haven't implemented that yet . but will do when I did
- doing some sound reverb effects in conjunction with the Fade to black thing makes it more awesome and less confusing
or even fading the general audio to silence
#makeCollisionsGreatAgain
jup
adding sound occlusion or reverb will make a huge difference indeed
might add that to the series eventually, once im done recording how to build the core functionality
Cool!
then again, it might be a better idea to postpone it to 4.16 when the audio stuff is more settled
but i guess it could be done already with what we have now
im hoping to be able to post a short grab/interaction series first though... as you may have noticed, im working on some of it now
Yeah that's a thing. I'm postponing everything audio related to 4.16...
good plan
Nice, I'm interested in testing some Force field stuff with motioncontrollers
like attracting objects to your hand
yeah maybe
i seem to remember seeing people do that using physics handles
like the portal gun
or hl gun thingie
portalgun=gravitygun
It might also work with radial force thing
setting it's strength to negative
and attaching it to hands
might work too
GDC is on the horizon, and we want to give devs and fans a preview of our plans. This year’s attendance is expected to top 27K game designers, developers, and more. With an amazing lineup of 500+ conference sessions—not to mention the Game Developers Choice Awards and the Independent Game Festival—GDC has something for everyone. We can’t wait to hit the show floor!
interesting.. seems like a lot of focus on mobile VR
and mobile + UE4
Steam Audio woo
@dusky moon think you're right about doing a force or add manual delta since using physicshandle alone, will more or less snap it to your hand
@eternal inlet ah good to know! because I was about to dive into physicshandle..
will do then : )
I really wanted to go to GDC this year 😦
@dusky moon Something like this? https://gyazo.com/bb04d0fedf202e41410a9072778b10a7
@eternal inlet yes something like that...
only if it could pull more than one object
at a time
Should be pretty easy to expand to a sphere trace and grab array of hits
Or create a influence sphere of a kind where it hits to pick up actors to attract
Nice! and you used the add force thing ?
Yes
Could actually be a cool effect to have like some magnetic gloves that would pull in stuff in a general distance and direction of hands
Based off distance to all actors of a class and dot angle to forward vector
And once released they will all drop to the ground
Ha yes thats neat! It feels great to vacuum stuff in VR
Hehee
thats awesome! if you rverse it will it just do a force push?
later 😉
Just gotta play a bit with multipliers to get it to work
And linear damping and gravity
yeah the force powers stuff is pretty easy to do
I remember doing a lot of fiddling with that back in the leap motion and dk2 days
Still having problems getting my steam demo to upload. Anyone on here that may be willing to help me out?
At this point it only shows up in the non VR section of the steam app still, and when I install the file is empty. When I try to launch it says bad configuration.
Have you filled out the VR part ?
Sweet baby jesus I found the problem. Steam is only showing my main app and not the demo for download haha. Anyone know how to make the demo appear in my steam app?
Just realized this is moving away from UE4 related, sorry
Hehe.. sorry.. I've not gotten that far in the steam process yet
I'm pretty sure it's a separate submission
You'll need to be issued another app Id. They only give you certain type of account for VR stuff
Yeah, Im not the brightest crayon I guess. I emailed the steam guy to ask how to expose the demo to my steam app.
It had its own app ID and depot already, but I may need them to grant me permission to test it.
Most VR teams are small ATM. Hard to know it all heh
yeah, just me on this one
Let's see.. I need to be a 3d modeler, coder, audio engineer, video director, social networking expert, web designer, logo designer, 3d optimization expert, multi store submission expert, press kit builder, animator, vfx expert, material builder and computer builder. I'm sure I missed a few things
So yeah.. hard to do it solo :p
Plus the day jobs
Translation.. good luck getting anything out
Haha, Im pretty stoked Im this close, just to have something release. Demo will be out next friday after this is figured out, and final game late may. Just 6-10PM everyday plus weekends
Then ppl complain that you little game for 3 dollars which gives them far more fun then anything else they can get for 10x the price is overpriced
What you making
Calling it trance VR. Just trying to make something big and fast
Thats the latest video. Final game will have boss fights, and pickups for speed and boss shards to expose the boss. Most of it is written, but I have to grind on the work needed to make real good levels
and my buddy is making all the music
If it all works out Ill have wall running and rail grinding, but at this point I can't get either to work. Game is more for the hardcore free locomotion crowd
Lol.. the puke crowd
haha yeah. Gonna hit up the blueshift and onward fans
haha, thats what everyone says it reminds them of
I had to look it up to see. In my head I was thinking SSX tricky with guns and robots when I made it
@glossy agate do you actually have a live build up on the steamworks page that you can make available for download?
@rugged abyss not quite yet. I have it all working now, just need to make the page nice and submit for approval. Demo will go like I think next Thursday or Friday. Probably friday so I have all weekend to fix any bugs.
When it goes live ill post it up. If you add me on steam "fatmoth1" I can DM you a link when its live too
ok cool. yeah I will check it out for sure
let me know if you have any other issues getting the build live. its not very clear on their site how to go about it, but its easy once you have it figured out.
Will do. It's figured now and runs good. Just want to release on a weekend so I can fix stuff
whats your game anyways?
does anyone know what the best way to package a game for VR is?
google tells me I should run fullscreen, stereo on and possibly some other console commands on begin play to do this
but that sounds pretty hacky
isn't there some better way to tell the packager to generate a packaged build as VR, or some fancy command line args that does the same
I've tried -vr, doesn't seem to work
also tested the bStartInVR bool in the general project settings, doesn't seem to do anything either
blueprints?
there is a toggle in the project settings that puts in the -vr 1 so you don't need to do that
VR is fucked 😦
Zeni doesn't leave Oculus alone, supposedly going after Gear VR too and there is a rumor that another small company is going after Palmer
I wish there was healthy alternative to Gear VR, in case it goes down
Meh
Companies will be companies, the tech will be fine
VR won't die because Oculus does
eeh, it definitely will be hurt big time
AAA companies won't develop for Vive for a long time, contributing to catch-22
Daydream is pretty horrible right now and there is nothing else in mobile VR field
If some company gets fucked, another will do the same thing the next year
I can't think of any tech that died solely because of legal issues
If that happens, it means nobody really gave a fuck
Well, Zeni seeks to stop sales of Rift, Gear VR, games/apps using SDK in any form and also seeks to make Unity and UE4 remove any code that supports Oculus platform.
perpetually
and I can tell you that people who hasn't tried VR (majority of layman) don't care for VR
Oculus isn't the only one laying down money for VR, HTC and Playstation are doing it too. Steam is just the only one not going that route.
so, how many AAA companies develop for Vive? I haven't seen any announcements. All AAA project are being developed for Rift.
Oculus is paying out the ass for first play on AAA software
good
lol
if Valve doesn't do it now, what makes you think they will after (if) Oculus goes down?
HTC already does
and one instance would be Fallout 4 VR btw, witch is currently vive only
AAA isn't about passion for VR.. It's a business and right now there is no money in VR
No, see it the other way around, instead of who will pay for being exclusive VR, who wont pay for being exclusive vr, so then they'll develop for vive as they can get more money there from consumers who will also all have to use Vive if oculus stops
A lot of those games aren't locked to oculus forever
no, they aren't.. But Zeni seeks to stop sales of them.
And no one wants to go to court to prove the code used is not related to Oculus and is 100% for Vive
Zeni is seeking a stop to sales of the rift itself
the SDK can be ripped out easily if you use OpenVR
and the SDK is the only thing that oculus has hands in
The company proposed Oculus be “permanently enjoined, on a worldwide basis, from using…any of the Copyrighted Materials, including but not limited to (i) system software for Oculus PC (including the Oculus PC SDK); (ii) system software for Oculus Mobile (including the Oculus Mobile SDK); (iii) Oculus integration with the Epic Games Unreal Engine; and (iv) Oculus integration with the Unity Technologies Unity Game Engine.”
Most of the games are either Unity or UE4 based
beleive it or not, anything made in unity or ue4 will work with openVR too
what I saying is Zeni can just continue on their path of VR destruction after Oculus is dead and go after other devs
I think Unity already uses openVR instead, don't they?
no they cant
oh, but they can
Zeni is known for their stupid lawsuits and C&D letters to indies
so I have no reason to believe they will settle on something that benefit devs and users
@tired tree Do you mean through revive or are the Oculus store UE4 games working with vives out of the box?
I mean that any game that uses UE4 or Unity can switch to OpenVR and run with the vive if they have to
if its entirely an oculus platform only integration then its a problem sure
yeah, it's pretty trivial if people are using the engine constructs to even have revive-like hacks to make it work too
i mean outside of any huge api changes it'll be pretty much relink one game, relink them all
also, I am more concerned with mobile VR than desktop VR
any game on UE4 works both on oculus, vive, and psvr at the same time
the only thing that would need change is the imput mapping
wich is quite similar in vive vs Oculus anyway
Mobile VR will live without Oculus, its the real future of the medium
real future is AR
and its also the largest consumer base currently
and AR is mobile
AR is a different medium :p
but its closely related
@wicked oak yeah i understand that. i was just asking because i would have assumed that oculus would force you to strip SteamVR libraries from Oculus Home build
I would argue that fully immersive VR and true AR are relatable in importance though.
they do
a poll showed that people don't want to use AR for gaming
because there's no place for open standards in mordor
yah
I personally wouldn't even bother playing AR games unless they are in some theme park
you wanted to room map with tango
i have really cool ideas i want to try with AR
and run AI around the room
that wasn't for a game, @tired tree
like a real world PvP shooter game that is played in a simple field
the headsets maps a X by Y square of terrain, and adds holographic cover
decent AR is farther out than really good VR anyway though
then its basucally paintball/lazertag
@wicked oak I'd rather play real painball than AR bs
goddamit the cloudflare thing hit even the sony developer network
the ip whitelisted ultrasecure one
gotta change pass....
can you add PvE on the AR bs?
@mighty carbon would you not rather play paintball on the surface of the moon with space artillery shelling the landscape?
my grandfathers deck...
AR will completely stomp the smartphone market
Yeah i'd rather pick up a Mac10 and shoot people in Real life than in counter strike /s
and i mean completely destroy it
becouse why look at the tiny screen
if you can make a big screen appear in the air
oh, you cant make selfies with that
sorry then
no, @prime iris , I'd rather not. I prefer single player games in totally fictional worlds. I don't play MP. And after police academy where we had tactical games with real guns, but painball-like ammo, I can tell you that no VR/computer game can match that.
ouch that is a lot of BitCoin sites affected by the cloudflare issue.....thats not good
@mighty carbon and what i'm trying to say is you can still shoot each other with paintballs while playing AR/VR!
you could use AR to augment the scenery without changing the gameplay
AR is nowhere near the level where it can show you moon surface (or whatever) in a satisfying way
sure, but one day it will be!
and it won't be there for decades to come
in a decade I'll be around 50 o.O
then you'll appreciate the augmented gravity of the moon!
Motor that is a silly argument, there are plenty of ex mil / police that play mil sims and FPS and really enjoy it
you cant run around in pc vr
becouse its a inside thing
with AR, you are completely unthetered
I've made an AR game once for a project which was somewhat like pokemon battles only set in jurassic park/world, where you had dino's you could collect on real cards, and then use the cards to battle other dino's, etc., which was kinda cool and worked way better as AR than it would've had it been VR, yet I've worked on VR games that work way better as VR than they would as AR, I think it mostly just depends on what kinda game you try to make
and can run around with it
you can do a game that plays like Raw Data, in a open ish field
doing that with full VR is ridiculously dangerous
AR can add robots and shit to that empty field
oh, I enjoy Crysis, Doom, etc. Don't enjoy military shooters. Know a lot of ex-mil who don't play games at all, and also who love CoDs. All of them would rather play real thing than stupid AR.
nah Blanco, you need Augmented Virtual Reality Reality
VR with real world overlays overlaying AR overlays
I, personally, see AR as non-gaming tech to enhance our every day lives
The cool thing about AR is that VR research will directly impact it, so its likely to come along for the ride
and before AR gets where it's of today's VR quality and complexity, and is accessible, I might not even care for gaming any longer
and since AR has a higher tech requirement it needs the boost, people don't like to invest as much in something that won't see real results for a long period
AR can be cool, but people call stuff like POkemon Go AR, which it isn't really, and that seems to make people think of it as something it isn't really
afaik, advanced AR can cover your field of view 100% with virtual geometry, in essence it becomes VR.
To me AR is "a feature" of VR of sorts. You control to what degree you want to replace the real reality
first we need to get VR to a place where it's affordable, useful, and cheap
AR should come after that
AR, if you don't limit it to just gaming, also includes stuff like the dog ear filter on photos, etc., so it already is pretty big
and affordable, since most people already have a smartphone
I was specifically referring to "proper AR" I guess
however, it'll become way better when we have glasses for AR cheaply available
it's funny to read remarks like that
VR isn't cheaply available yet and per VR heads it won't be any time soon
it will take time
a lot of time 😃
what is a lot of time for you?
I'd say in 3 years time we will have pretty decent mobile VR
with what platform ? :/
it doesn't matter
like you were arguing before, even if Oculus fails, somebody else will step in
in terms of the technology.
in terms of funding, that's different
and I do agree it will be harder
well, it does... Daydream isn't moving forward as fast as Gear VR did. And Gear VR was a pioneer tech. Now mobile VR should be moving forward pretty fast, and yet Daydream isn't moving.
Daydream hasn't even been released worldwide yet
It's an open platform, supposedly
I had to go to Germany to buy a couple of Daydreams for example 😃
so, nothing stops any vendor from making their phones Daydream compatible and release their own Daydream HMDs
that's not true
and yet, there is no rush it seems
they need to get approved by Google
there is a certification process
and only if the phone's specs meet the demands will it get approval. I don't have any insight on the specific requirements.
but i'm sure it's not just hardware
well, it is just hardware since OS is from Google
and Android 7 has native VR capacity
as long as vendors don't mess with that part of the OS, there is nothing else
whatever the case it may be, there still a process you need to go through and you can't just release daydream compatible devices
UE4 support for Daydream is still miles off compare to Gear VR support. HRTF audio doesn't even exist for Daydream. I saw people can't get controller to work with UE4 and can't use conventional gamepad to be a substitute .
I want to think Google is at least trying to make sure the experience is as streamlined as possible to costumers. That being said, GearVR is the supperior product. And if the rumors are true, then the new GearVR version will blow Daydream out of the water.
dude
patience!
GearVR plugin has had a lot of time to mature
couldn't you just buy daydream obline?
It just seems that Google puts as little effort as possible into Daydream.. Kinda like what they did with Cardboard support.
It was way more expensive online and to get it shipped to my country. I had a trip planned for germany anyway, so I bought 2 while there.
I do agree that Daydream is a rushed product
but it's Gen1 and a rushed response to GearVR
to get their foot in the door
I can't say much about what I know, but let's just say Google is investing a lot of money on daydream
on another topic
I've done some more tests with UMG and text
one thing to avoid completely is to have just floating UMG text
if you add a background to it through UMG it improves aliasing A LOT on mobile
the reason why it still might look a bit jagged is because the render target it generates has no mip maps
the render target is drawn on the plane as is
@raven halo so, is there Daydream gen2 coming soon ?
I know nothing about that specifically, I was referring more to funding
I am waiting on Asus ZenFone AR.. This way I can mess with Daydream and Tango, assuming they price for the phone is ok
I'm a bit skeptic about AR right now, admittedly I haven't tried any proper AR device. But the Hololense stuff, even though promising, it doesn't seem to be very compelling at the moment.
well, gotta start evaluating AR somehow.. I am not going to use it for any game stuff. Just simple R&D for non-gaming applications.
plus, it's affordable. Hololens is not.
right! yeah
at this point I just wonder if it would be wiser to go back to conventional game dev until dust settles and competing solutions mature more (could be years from now) 🤔
(I don't have desktop VR and going with Vive isn't an option for me at this time)
I think GearVR is the sweet spot
it's affordable to develop
and you have 5 million people with one
well, yeah.. until it's killed off by Zeni
then you port to whatever comes next if that happens
which I highly doubt
they have facebook money
facebook lawyers money
So, let's say I want to have my Gear VR project ported to Vive/Rift. How hard would it be to add motion controllers and whatever else is needed into my project ?
How to make a lever move along with the motion controllers
set its transforms based on controller's ?
@mighty carbon can you share me an example
I am just using logic here.. I don't have desktop VR
the way I see it, you need to either convert location of the controller into rotation of the lever, or you need to lerp between max/min rotations at variable rate based on how fast you move controller
@mighty carbon lookat node will do that rite?
@mighty carbon i don't have Gear VR, do they have motioncontrollers?
@eternal inlet nope, not yet.
so how do you control "hands" with Gear VR?
i see
you'd use either toucpad on the HMD for simple gestures or bluetooth gamepad
in that case, i assume it will be a matter of adding hands for Vive/Rift
and mapping movements to whatever buttons you like
do you move (have locomotion) with Gear VR?
I teleport
ok in that case, you can just borrow the code from the vrtemplate
aye, cool. Thanks
uw
Kind of, there aren't room scale boundries, he'll have to rip out all of the roomscale parts.
It has tracked bounds support by default
you don't have a need for that
just the arc / teleport point nodes is all you need
you lost me on this one...
I was asking about converting Gear VR project into Vive project. Not other way around.
so I'd need roomscale, controllers and chaperone
oh
sorry
yeah you'd be good just ripping it out then
they store the controllers in seperate blueprint actors in their template, its kind of perfect for conversion for you
aye, cool
I have a steering wheel and I want it to rotate along the direction of controllers, when the vr controller is in touch
World aligned textures and VR, do they play nice?
yes
Sweet
I have created an explosion effect using a tessellated shader. It performs well in VR. I want there to be a flash of light at the moment of the explosion. I am handling this now by toggling a bright point light on and off. This is not working well performance-wise. (shadows are disabled). What is a performant way to achieve this effect?
ah. i changed to "stationary" and it no longer hitches
should have noticed that earlier
@real needle would a emissive mat work if you set it really high?
like a quick flash of an ord with a super high emissive, then a spotlight that points right in the players face. Seems like spotlights give a lot of bloom in VR
i will test this idea out, thanks!
out of curiosity I ran Vive test on my PC
apparently my Skylake i3 is good to go
only GPU didn't pass the test
(I have 670 GTX)
your going to chug alot man
how so?
I am going to get 1060 6Gb
I don't plan on doing desktop VR with 670
@restive blade ^^
Yeah.. decent card. The i3 is a litle painful however
how so ?
how do I ge the AI to see the VR pawn so it will attack
is the disconnected feel between an IK and motion controllers due to motion controller components having a late positional update? And if so is there any way to trigger the animBP to update as well?
at the moment i'm querying the motion controller positions in my animbp's update
and using them to drive the IK
Which ik are u using @prime iris ? I personally use Ikinema and they have that lag as well
just the two bone solver
As far as i understand, ik calculations cant run on late tick
:<
Not sure of the specific reasons tho
i had a feeling it might be something that was pretty deeply coupled in the engine
Im sure @tired tree can give a better explanation
quite a lot of work is passed off to other threads and working a frame ahead/behind for performance. i guess it makes doing a late update trickier than just asking twice 😄
What ik stuff are u building? Arms?
Nice
everything that doesnt use the late update thing will be 1 frame behind
something like an Ik, probably 2 frames behind
i think if you tick it properly like ticking it before animation, you can get it better
i guess one solution is to have motion controller models be part of the skeletal mesh?
its what everyone does
another way is to do the skeletal mesh for the hands separated from the body or others
@wicked oak how do u tick before animation?
and actually go and ATTACH the whole arm to the hand controller
and use late update on that
that might even work
i think it is possible
but i havent done it so i cant tell you
i use late update
and only a hand mesh
not whole arm with ik
The vr template has the handmesh attached to the motioncontroller so as long as use late tick is checked it works ok i suppose
Just not with ik if u were to add a forearm
@wicked oak are u developing for vive?
Oh dayum
Nice
I ran into a weird issue with the vrtemplate this morning. The left hand is a mirrored version of the right hand, and i introduced a box where i made the hand a child of. But i also wanted the box to be mirrored
Problem is that if i mirror the box, the left hand gets a slight rotation offset for some reason.. ever tried something like that?
The box acts as a collider and the collision also doesnt act as the right hand. It gets snaps in rotations and other weird stuff
Seems like a bug
not a bug
just a thing that happens
my 2 hands are carefully aligned
even if they are mirrored
becouse even if mirrored, the "attachment" point can easly not be the same
unless you use a socket as the attachment point or other things like that
Its just strange it only happens for the left hand
Which is mirrored
I'll post a video showing the issue in a bit
Maybe im missing something
this is bit old already so sorry if someone already linked it
@wicked oak https://www.youtube.com/watch?v=fKGbOLvqCZM
here's the issue i talked about
has anybody figured out a good way of actually seeing the stats when VR previewing? (stat scenerendering, for example)
it is not aligned properly with the screen and is cutoff
custom UMG. I think it's pretty vital to debug out of a custom UMG tbh.
ahh, neat idea but dang...more work lol
a question for the profiling gurus - what is the best way to determine where my draw calls are coming from?
i am trying to figure out why under some circumstances it jumps up by an order of magnitude
the GPU profiler
@wicked oak I've done late update on arms onyl to see how it feels, it obviously requires seperated arm meshes and they shift as a whole with the late update translation. Its not really any better since you have to throw out rotation.
re: GPU profiler - it shows me the times in ms, but how about the number of draw calls?
stat rhi
yeah, i'm thinking of having several skeletons. a torso/body for the head for attaching stuff to (weapon holsters etc) and for querying shoulder poisitions. and then having the hands/forearm meshes directly driven by controllers and using the previous frames shoulder location
@wicked oak - thanks!
but the sooner there are popular consumer priced fullbody mocap solutions the better! 😄
if you change your skeletal mesh tick to post physics or post update you'll get less delay on the animation
yeah, makes sense
but i'm also hopeful that it's only really the constraint points of the IK that really matter
i guess my plan for using a forearm UI might require a little juryrigging on the elbows upvector depending on what the player is doing
IK is kind of where the entire concept of UE4's late update falls apart
You don't really want to run something that expensive on the render thread, and it also doesn't have full scene information.
yeah. i imagine it's something that will be separated in future engine designs. things that the player influences directly will be rendered in a specialized way that allows it to be closer to the physical objects they're constrained to
but for now the engine assumes IK is as important as it is on npcs 😄
"is your animation walking up stairs?"
hmm, should i design my skeleton around a unit height and then scale it at runtime to match the players height?
yes
why are simple things so complicated?!
its not that complicated
make the default 6' height
/182 cm
scale it by the difference
yeah, i know. i just wanted to make a pipboy today and I had to learn about IK instead!
add in a mobile app that people can use with their collectors edition fallout4 pipboy's
instant presence
the fallout 4 collectors ed pipboy was functional?
or do you just mean for physical sensation of it being on your arm?
didn't you just put your phone in it and use an app?
yeah
sweet
has there been any word on fallout vr yet? they going to be showing off things at gdc?
or it still soontm?
they already said that the game is fully playable, start to finish currently
and that multiple locomotion options are included
sounds nearly done
guaranteed to be a junkshow, imo
awesome. want to see what the pros can do
ports just aren't what we need
lol
ports are fine if done correctly
fallout is actually better suited to it than most
VATS and the concept of their screen interactions would all play nicely
well i'm not a huge fan of fallout at least the first person variety. not even played vegas or 4. but if something will get me to play their walking simulators it will be a VR game!
VR and exploration is not compatible at the moment
it is very contrived
VR and exploration is not compatible at the moment
what.....
how can you have not played Vegas or FO4? You scum !
new vegas is the only real fallout of the bunch
F1-2 were best anyway
exactly
i don't enjoy bethesda rpgs, they don't have any real gameplay beyond press your biggest x to continue. but that tends to be a controversial opinion so i'll leave at that
sure but thats fine really
huge fan of fallout 1/2 though
MrBushido - play new vegas
dunno man, in 1/2 i could have gotten away with killing everyone, in 4 even though they don't let you kill everyone I still soft locked my character from trying
same with skyrim
@prime iris I agree entirely but I still enjoy them. I'm a fan of the old school D&D stuff like Neverwinter nights and such. Ever since Oblivion I have seen rpgs go to shit. There are still good ones though like Divinity:OS and Wasteland 2 etc
@sharp swan neverwinter nights was awesome! i enjoyed dragon age too but haven't played the recent titles
yeah dragon age origins and Inquisition were fantastic
NoX still best RPG ever made, come at me
LOVED NOX
pillars of eternity was the last rpg i sunk a bunch of time in to
I was going to play that as the sequel is upon us
NoX wouldn't run on my pc very well at the time iirc so I dont remember playing it
it was the most anarchic pvp experience of all time
NoX was what MOBAS should have been to truly be esports, MOBA Quake
beyond even UO
wivy did you play long enough to see our map editor? full scripting system, and LUA support for the game? was a lot of love left for it
i did not, unfortunately
i was never very good, i just really enjoy insanity in online gaming
that is why i loved Unreal Tournament
jack taht speed up
even upgraded the renderer by injecting in and overriding the directx calls
amazing - i love when "the nerds" keep an old thing going
the world moves on from things too quickly
@sharp swan if you're referring to the new pillars game yeah, the first is worth checking out. I wouldn't say it's a great story but the engine is pretty solid and the gameplay was fun
and by that i dont mean the story was bad, just kind of generic fantasy
they definitely put a lot of effort in to world building story writing. loads of neat details in a kind of tokeinesque fantasy
i'll set some time aside during the next school holidays for it
I should probably play it again to refresh the story if the sequel is going to be related in some way
maybe even try one fo the achievements. there was a pretty crazy ironman challenge i think
something along the lines of not grouping with other characters on the hardest setting 😄
"The Ultimate - Complete the White March and Pillars of Eternity, defeat all dragons, all bounties, and both archmages on Expert, Trial of Iron, and Path of the Damned modes without taking any companions after Cilant Lîs." - 0.0%
@wicked oak why do you say i3 is no good if Vive's test says it is ?
it probably depends on the generation of i3. but you're probably going to be pushing things pretty close to the baremin spec doing vr with the most powerful i3s
you'll probably be okay for a most games at the moment, but quite a lot of simplistic graphics at the moment.
are you having an issue with it or just pricing things?
have you tried Vive's test? It looks far from simplistic graphics
sure, but the vive test is valves lab stuff running in highly optimized scenes right?
I suppose.. What's how people should be making games - optimized 😉
i agree, but outside of a select few studios that's not a priority :<
minimum by oculus is an i5
btw @mighty carbon can you test my game?
i would like to know how it runs on an i3
if I had VR, I could
if you get a vive or a oculus
just debating whether to get 1060 and then get VR, or get i7 and then 1060 and then VR
1060 then vr
the cpu is quite a bit on the limit
is it overclocked?
anyway, most games wont be an issue with the cpu
the absolute mayority are gpu bound
mine isnt due to the drawcalls
but it also isnt very multithread (thx epic)
@mighty carbon you can probably get away with a 1060 and an i3 at the moment because a lot of the games are relatively simple
yup
by the time you want to get a CPU 2 things have happened
1: ryzen is good, you can get a kickass rycen cpu
2: intel lowers prices to compete and you get an intel for cheap
or a 3rd. you grab a secondhand i7 from the guys who are getting a ryzen
expect the secondhand i7 market to get flooded
speaking of multithreading, I don't think any engine has multithreaded game code
Doom 3 BFG had it on a single thread, RAGE was the same, even DOOM 2016 has game code on a single thread I believe
so, same game in VR and not in VR would most likely consume same amount of CPU time
I can't go AMD route as I just upgraded to Intel system.
I got i3 because it was on sale at that time and Oculus lowered lower end specs to i3
so I figured I get i3, 1060 and I should be good to go as developing for the bottom line usually yields product that flies on better systems
(and eventually I can get i7 and be set for a loooong time)
and since Vive's test indicated I only need to get better GPU, I am guessing I was correct
it would be extremely painful to work with threaded game code
(especially debugging it)
all ubisoft games have multithreaded game code
Destiny has multithread game code
pretty much most console focused engines have multithreaded game code
I was wondering if to put this TrackIR in a seperate thread or not. Im not sure if it will benefit from it as it's tied to tick, but I could make it an event in the dll or call functions in unreal through pointers so it doesn't need to be. However it does need to stay consistent at a performant level
@wicked oak well, Ubisoft has a lot more manpower than Epic or small studios to work with multithreaded gamecode
dakka 2.0
im testing lots of enemies at the same time for the hell of it
also becouse optimization
I still don't get why you draw arc for teleport 😃
just draw the end location reticle
nope
it becomes really hard to aim
ill probably end up making the arc less obvous, but you need the arc
the arc helps a lot visualing the path to point
ug
the hell did I spell there
was loading visual studio at the same time
I see
so you need something as reference of where you are aiming
every single game does this
maybe not an arc, but a line
but ALL do it
for a reason
I like the bullet spread btw, makes those guns seem like they would feel powerful
btw, they all do not because of a reason.. They just do because first dev did it and managed to justify it 😃
you don't have laser sight on every weapon
not always
arc trace allows you to teleport to things you normally couldn't, since it has a drop to it
without visualizing that drop its really hard to aim to what you want
yes even then, it is a lot easier
like I said, guns don't have leaser sights in every game.. It sure helps a ton having it, but not all of them do.. You just aim and shoot. Same can go for teleport
you get far more visual feedback from a line + point than you do a point only
the teleport is allways carefully aimed
so you need to know where you are aiming
the arc does that
make it additively transparent maybe?
for me arc breaks immersion
it makes me think I will fly to the end point over the arc trajectory
you are doing it from your head though...that is a lot different
have you ever tried tracked controllers?
yup
and pretending I am in VR and imagining how things would go makes me wonder about the arc
if its gaze aimed, no line is ok
but with the hand controller, you need something
becouse games have different arcs/trayectory
so you cant just learn to do it
its odd as hell if you dont have it
source: i have done it for PS4
until i optimized the spline
its also hard to "aim" down the sight of the controller when its in your hand, its not like a gun where you line it up with your perspective
@wicked oak needs more bunnyhopping!
when I get to it, I'll see first hand if I am wrong or right 😉
too bad its on VR, good luck bunnyhopping there
but Sairento VR does it XD
i really dislike how they do it, i prefer my straight dash
you might be right for yourself, but the vast majority prefer the line, don't think its going away
blanco, you can bunny hop in vr
jump + turn ;p pretty sure people would get tired doing it IRLthough
@wicked oak lol really? not played many shooters though i keep hearing sairento come up
things my game does better:
everything Tech related, too many small things
enemies
i have spells
Sairento has:
well you just need to come up with something to make your game more compelling than theirs! change the art to voxels and add a survival mode? 😄
a waifu on the logo
consistent art style
a mission based system, with difficulty and unlocks and loot
while the core mechanics are better on my game, my game is thin as a paper
sairento has more playable hours due to the unlocks and loot and missions
it sold better at launch due to a waaaaaaaaaaaaaaaaay better trailer
the biggest mistake ive done for DWVR launch is the bad trailer
i see that company is trying to have all oculus sales blocked
I just don't see how it's possible
if it was decided that Oculus didn't steal tech, and if current software has no code from what Carmack took from Zeni, what legal ground Zeni has to halt sales of Oculus/Samsung products and software with SDK in it ?
Share TweetVR revenue from software and services will increase by almost $1B this year. After a mixed year for VR hardware sales in 2016, two new separate reports are painting a more optimistic outlook for the VR market going forward. In a new consumer VR report just released by SuperData Research, the findings show that …
oh those forecasts ...
Why do you guys think that growth curve is unreasonable?
I suppose the difficulty is its predicated on continuous year on year growth which would suggest constant updates/refreshes to hardware
But even so, why wouldn't you expect pretty significant growth to VR from this current point in time?
I hear people say because VR is not essential and doesn't take off as fast as smartphones did, investors will lose interest to VR pretty soon
plus Zeni being a bully to Oculus doesn't help industry overall
Expecting VR to have the same growth trajectory as smartphones is just ludicrous
At least post iphone growth
Given that smartphones were around for nearly a decade before that.
Nah I don't think it'll be overly aggressive is all
I think they'll be a big boost in 18 months time when the first wave kinks are smoothed out.
Wireless, lighter, better resolution, hardware can catch up etc.
and probably the biggest factor - an establishment of content standards.
Yeah... the steam approach right now is... creating a bad narrative in the marketplace D:
and yet, somewhat essential for creating VR skill pool for the future of the industry.
he says prices are too high for desktop VR. I suggested mobile VR. He saying phones are expensive. I said Axon is $400. He says he doesn't want Android VR phone.
naysayers will find excuse to dismiss VR
Some people will lead and some people will trail a market.
That goes for producers and consumers.
To be fair, game devs on average are leading the market. But I don't know why there should be a suprise that some of them want to stay conservative and comfortable in their current niche.
just weird is all :/
"Me: Have you tried VR? I think it's the future of gaming. A dev: Meh, I think it's a gimmick. Me: have you tried VR at all? A dev: Nope."
that's a kind of conversation I've had many times with other indies
I have yet to come across a dev that isn't excited to dev VR
But opinions may vary.
I just got my Oculus touch controllers
It's just time is all. I don't think it's going away any time soon.
Demoed my stuff to tech startup guys a couple days ago.
Hey zap
Oh how'd it go?
Some comments I got was - "oh wow, I thought this was a gimmick"
"But it's the real thing"
ah good!
Dude was getting excited about been able to pick up a gun and fire it 😄
Which in a certain sense IS super cool.
Compared to what we've come from - keyboard/mouse/gamepad to point a cursor on the screen.
@restive blade yo sup
And they were all pretty impressed with my walking-in-place stuff
oh yeah, it's great
@clever sky tried your demo btw
@restive blade Rift or Vive?
As Newell said, we're at the 1981 of personal computing.
Vive
Was interesting
Haha... fair enough.
I think it would work for some people really well
Were you able to try it for long?
No.. I had to catch a bus home
I feel the biggest weakness to the thing is that... it does take people time to adjust and get used to it...
30-60 minutes. So it's not great in a demo setting.
Yeah... Ppl get the teleport right away
Yeah. Teleport is great like that. But a few hours in, you're definetly feeling that limitation.
This thing... is the opposite almost. Takes longer to learn, but after you're used to it... you find most things quite lacking by comparison! D:
For sure... That's not been a major issue yet.. most people don't spend long in VR
Yeah.... it's mainly the users that own the stuff that'd have that sort of exposure.
And VR is still kinda been discussed in the context of new experience.
I think longest I ever spent in VR at one time was 20 mins
I mean I say that it takes 30-60 mins. But that's just a conservative estimate.
Some people put it on and they just get it with minimal instruction.
'Ok. Point the arrow? got that... walk on the spot? walks ok! It works! Excellent!'
Yeah... I got it... But only had a couple mins hehe
And some people are like... 'arrow? heading?' 😛
Ah. Well... hopefully I can get the rift version done in the next few days.
The code is done, but I just have to update graphics tutorials and stuff.
Incidentally, it works better on the Rift IMO - due to the analogue stick.
Easier to understand the stick as a device with which to set heading.
Well I've got my rift now.. I'll try it on that
The current version is highly incompatible with the Rift 😛
it loads... just everything is messed up.
Story of my life
Anyway... so what did you think of your small initial impression? And where did you try it? 😛
At my previous boss s house
He's flying to LA today
For GDC?
hello, I imported this model from C4D and trying to throw objects inside, nothing goes past an imaginary barrier
I can't figure out why
I had the exact same issue with an architectural model imported from 3DS, I couldn't navigate anywhere inside the object, nothing could pass through it
great ways to spend my Saturday night lol
open the model and click the "collision" button
it might have generated collision that encompasses both parts
mmmm let's try that
also concave hulls will probably require you to make your own collision geometry
make a UCX_something in C4D and export it with the mesh ?
How To Set Up Collision
yeah
i think that's how it works
not done it myself so probably check the docs first 😄
aaaaah I seeeeee 😃 awesome thank you Mr 😃
a simple fix is to set it to use complex traces
on the mesh/actors collision settings
but that means it will use your models full polycount of collision and that can get pretty slow depending on how many you have/polycount
yeah, use complex as simple
gz
hehe, been there countless times
here's a video on making concave collision https://www.youtube.com/watch?v=Gy1ZBp_OTVo
woo I need that!!
yes its the UCX_
so this is how UE identifies custom collision meshes
yup
awesome. That will save me a loooot of time. I have objects like that here and there. I just got to go back and fix em
secrets don't come with the beginner's guide haha
explains why you would want to use convex collision https://answers.unrealengine.com/questions/174028/convexconcave-collisions.html
tl;dr you can't simulate physics on complex collision (or at least couldn't in 2015)
so while simulated objects can interact with complex surfaces, the complex collision object would have to be static in the world
mmmm that was exactly what I was looking at at the moment. I was trying to simulate physics
hmmmm why unavailable ?
then I think you're gunna have to make a custom collision out of convex shapes like in that video. double check with #legacy-physics for the requirements, they'll know what is possible in the current engines
awesome. Thanks for all of your help Mr
no idea, why it's unavail
it should come with a can of mace attached for when the muggers try to steal it on the tube. :p
MWC begins tomorrow
and so does GDC I think
will be a lot of VR news coming next week
@eternal inlet thats very strange.. first things first, why are u introducing the box component? just to help with placement or?
also, what engine version are u on, and is that just the default virtual reality template, or one of the many spinoffs?
Box is for collision later.. now i dont use it
Its the defaultvr template yes
4.15
kk ill have a looksie if i can reproduce
Pretty sure it can be reproduced with a non vr setup
Just easier to spot it with the motioncontrollers
so far unable to reproduce
@eternal inlet can u share a screenie of the exact sizing and rotation of the hand mesh (preferably unparented) and the box?
been trying to eyeball similar angles, but have had no issue like the jump in rotation your having
UE4 and mobile are weird combo
I get a ton of CPU stalls when profiling my game for Gear VR
OnlineAsyncTaskThreadNull
can't wait for the gearvr controller 😄
why am I getting so many CPU stalls ?
900k PSVRs were sold
I am about to throw in a tower with Gear VR development 😦
(not because of sales numbers for PSVR but because piece of shit mobile hardware doesn't work well with UE4)
yo anyone in the states with a vive have some time to test something in multiplayer?
tried with jonas but he is in denmark and region lock got in the way :p
anyone up for a discussion about the good or bad of scaling player ingame vs scaling playerspace?