#virtual-reality
1 messages ยท Page 135 of 1
@fossil stratus the camera gets moved according to the HMD
so the relative location will get blown away I'm pretty sure
if you want to move the camera you need to move the parent component it is attached to
@sturdy coral thanks, I'll try it ๐
@sturdy coral hey it worked! thx
so much activity here today
I usually try to read through everything here, so that took more than an hour now
there was a dev-vr merge to master 8 hours ago but you guys talk about iphone stuff. wrong priorities ๐
theres the fix for the debug line issues with instanced stereo that some people mentioned here: https://puu.sh/xyhYG/6bc316ebc6.png
some gearvr fixes to make @mighty carbon happy: https://puu.sh/xyi0B/429fa6dcfc.png
I need more support on my forum thread
@full junco is it for 4.17.2 ? Anything there about fix for 3D Widget performance being horrible on Gear VR ?
its 4.18
lol
dude
I said merge to master
I think I'll keep 4.16.2 for Gear VR dev
4.18 looks super boring at the moment, there no relevant big new feature. I have no idea what you mean @granite jacinth
@full junco really?
@granite jacinth what huge BP stuff is coming to 4.18 ?
Nothing excited you?
yeah, nothing
maybe that they fixed the left/right issue for steamvr now, but thats a simple fix that I already have added myself now
roadmap for 4.18 looks boring
the only "big" feature in 4.18 I remember is multiple bounces for skylight with lightmass. but I dont care about baked lighting, so I dont care
I love the idea of having better bakes lighting (hopefully it bakes faster too)
it will obviously be way slower
no, it shouldn't
you dont add higher quality and faster baking times at the same time. more bounces = slower
there was a thread on the forums about it and the guy who implemented it went to do internship at Epic and that's how 4.18 got multi-bounce skylight
That's how we got the famous VR Editor too
and afaik it wasn't slower than current single bounce light
the code was written by daniel from epic, maybe he got "inspired" by whatever someone else did, but daniel wrote it
damn, forgot his name already
so, what kind of BP stuff is coming to 4.18 ?
I am looking
nothing relevant
and could we maybe not talk about general 4.18 stuff here in #virtual-reality
lets do that in #ue4-general or #blueprint or whatever
๐
here we can talk about all the stuff that was merged to master in todays dev-vr merge
since its a surprisingly big merge
a lot of mixed reality stuff in there
no idea what they mean with that
if it;s a big merge, how come nothing exciting is coming ?
do they mean the windows headsets with MR?
this sounds nice, being able to read the console variables and stuff like stat gpu in VR
I guess this makes motion controllers feel a bit better: https://puu.sh/xyiqi/8295cfd760.png
I have no idea what a "garbage matte" is: https://puu.sh/xyirZ/fb714bc3bc.png
updated SteamVR to 1.0.9: https://puu.sh/xyitt/6c0b287055.png
that fix for the visibile mesh area of steamvr: https://puu.sh/xyiuB/acc5bc41eb.png
support for up to 11 vive trackers: https://puu.sh/xyixj/d5d3d3ce19.png
what is "Oculus XR"? https://puu.sh/xyiFP/388b90cbec.png
@tired tree you looked at all the stuff already? you probably like the steamvr improvements
typo
typo?
nvm
Actually
You know how MSFT coined MR ?
This is probably Facebook's take on it
lol, only 124,292 lines of code added to master with this one merge: https://puu.sh/xyiNa/ab5b4f43f7.png
its due to updated SDKs mostly I guess
nah
I just googled that hell out of it
the only thing I see
is the VR Kaleidoscope camera
lol
and these are apps
not much for Oculus / Gear VR ๐ฆ
@mighty carbon the oculus integration in UE4 was always better than SteamVR, so it only makes sense they improve SteamVR a bit now
What is the best thing they did for SteamVR @full junco ?
I see
well I guess some people are very happy about support for 11 vive trackers
@full junco do you have any trackers?
no
looking at @pearl tangle and his 11 trackers
ยฏ_(ใ)_/ยฏ
I mean, is there even a list of games that support it?
Cant be that many anyway, I though they were supposed to be $50/each
Unless it comes bundled with future Vives, I just can't see them doing well
Not just for gamers, but devs as well
Even though, they gave out like 1000 of them to devs
Well you know how there is that joke about putting a paper bag on a homely girls head before banging?
11 trackers. All on the important parts of her. Put on HMD, select player 2s character you want to see.....profit.
ok, the mixed reality stuff that was merged isnt related to what microsoft calls mixed reality
its for easily capturing mixed reality videos I think, so have a camera with a tracker attached and then capture video with that in front of a greenscreen
yeah
its automated, with depth planes and everything
looked pretty neat from the beta code
automated in what way?
it even does the green screen removal
I mean....who cares, hang a green cloth and half ass it
well, its not that much fun on 1.5x2 meters
for looking awesome in a video something like 3x3 meters is needed at least
both openvr 1.0.9 and 1.0.10 seem to exist in the UE4 third party directory now
not sure why they would have two seperate versions there
mm
well there are a few things in engine they keep multiple versions around for
maybe halfway through porting?
AZ sunshine supports it now too http://imgur.com/gallery/1B4Ya
its not really nice and clean though to use two different versions of the same SDK @tired tree
@glossy agate that gif is at least 1 year old
yeah i know
Dang I just saw it today
I saw it on reddit a looooong time ago
@full junco XR is probably openxr?
was the first VR thing that got more than 10K upvotes I saw @glossy agate ๐
@sturdy coral I would want to think it is, but why "Oculus XR" then?
it is supposed to allow vendor extensions like opengl, so maybe something like that
but, hopefully it wouldn't be full of vendor extensions right at version 0
@sturdy coral there are some XR files that look quite general: https://puu.sh/xyjUk/07671505ba.png
Hey do I still need to get AR Kit from the source code build?
Or is it activated via a plugin?
The whole open XR is meant to be the middleware API for all the others to talk to right? Rather than steamvr doing it's on implementation for Oculus
And the trackers are definitely good fun, I think once you see more peripherals come out that use them they will start to catch on. A few different glove controllers coming out that they are handy for. And for cheap motion capture
ah openxr
that'd be nice.
we'll see if it happens
@sturdy coral it seems to be openxr stuff, yeah: https://puu.sh/xykJj/d2ec64fee5.png
The green screen mixed reality stuff should be the mode 4 of their recent mirroring changes right?
Garbage matte may refer to background removal without a greenscreen, not sure, never heard the term
Matte is like matte painting in film, sort of proto greenscreen
I suppose you should be able to get the depth from doing something like chaperone to remove the stuff but then you would need to be processing extra depth information through the camera which is costly compared to green screening
Yeah, might just be the kidna hacky thing, subtracting from a known background and then only drawing if there is a difference
Like a lot of twitch streamers use
Just a guess though, garbage in that case would refer to removing a noisy/busy background
Yeah it's not so great with a single camera unless you throw some other open CV stuff in there and do weighted averages across the difference but that's only so good with a moving camera. Like the Kinect stuff those other guys ended up doing. Oh FYI their alternative solution works really well on me when I'm wearing a white shirt. It really really struggles with black objects though and makes it much worse than our solution...and the activation is in Chicago...
@pearl tangle do you think Samsung stopped improving Gear VR for good ? (latest model had no improvements or new features and Note 9 is rumored to be foldable)
is there a way to select which headset the vr preview targets? I have both a rift and a vive hooked up to my machine and so far the only way I've found to do something like this is to manually enable / disable the Oculus plugin from the VR section.
is it normal for a build size of a project to almost double when adding the ability to play on both vive & oculus ?
@twin pulsar I think thats the way to do it
@fossil stratus well double to what? if it doubled from 20MB to 40MB then maybe, but not from 500MB to 1 GB ๐
I'm making a VR experienced with a car. Whoever tries that says that the car is not the real size.
Could It be related to the penso distorsion of the vive? How can i correct It?
Lens distorsion of the vive*
no, its not related to the lens distortion of the vive. its just related to the size of the car
@full junco thanks!
@twin pulsar actually, there a ini file where you can set the priority of the different SDKs
oh really? do you know which one?
still I guess there is no choice within the editor per se right? maybe I could track down in the source where the file you are talking about is being used and put a little ineditor menu too
you will always need to restart the editor for it
I dont remember any more which ini file it was
try to search in the ini files for "priority"
ok
You can verify that 1 metre should be 1 metre in both reality and VR by placing the controllers 1 metre from each other, with a 1 metre cube in the scene for comparison, for example. it should be 1:1 by default
@full junco @dusk vigil
I've double checked the dimensions both in unreal and Maya and the are correct. Users see the car with lower length
Well, it is possible if you have messed with the camera scale
Don't trust the user, test it yourself
Strange....my camera is the default unreal camera with the lock to hmd checked.
why do you say that "whoever tries that" says it would be wrong, do you yourself see it correctly?
hey any idea how to fix this issue ?
it only appear when i play in vr mode
https://answers.unrealengine.com/storage/attachments/202170-greenlinea.jpg
What are we looking at, exactly? @analog topaz
those green rings appear randomly around the scene whenever i play in vr mode
they also appear in static reflection captures
they don't appear in editor
@uneven moon
That's very strange, I've never seen that before... It may be some kind of debug tool
Let me see if I can find anything on it
Try removing all of your reflection captures/volumes, rebuild, restart the engine, re-add whatever you deleted and then rebuild and see if it persists.
ok i will check it out , thanks for the help
from another chat, sounds like maybe the same issue "I fixed it once by switching my directional light and skylight to static. Might also be related to reflections."
shakes fist @dusk vigil I asked about that
Hiya Joe
thnx @dusk vigil
I'm trying to figure out the VRExpansion's template. I suck at this. There's no VRPlayerCharacter in the scene, just a playerstart, so I'm a bit confused, where's the code/BP that spawns the player and takes care of the movement & MCs?
Look at MotionControllerPawn, also look at the different Maps EDIT Whoops didnt realize it was the Expansion
@wicked oak when Oculus sets release date, what time do they publish the app in the store on the said release date ?
(I am not seeing mine in the store yet)
@willow trail its spawned from playercontroller and gamemode
if no hmd is connected it spawns a 2D pawn for testing
also if not running in VR preview
Alright, got it, thanks, I gotta take a crashcourse in... something
the template is multiplayer
dragging player characters into the scene isn't going to work for that
Ok, I don't suppose you got a singleplayer template right ๐
it works single player too
I realize, just without having to go through the multiplayer hoops
its not technically a multiplayer hoop, being able to test out of VR with things and then go to VR only when you want to is useful
Oh right, didn't realize that was something you needed code with. Yeah i expect that'll come in handy
the FPS pawn is a subclass to the master VR pawn with the camera and controllers locked into position and WASD movement
so if you change something like how gripping works, the fps pawn can test it roughly
prior to committing to putting on the headset.
wait do you mean you can test out motion controllers with the "static" camera?
if you locked the camera on the fps pawn into position and turned on tracking for the controllers, yes
they would follow the hands
but its primarily for testing the gripping code
That's definitely useful, first thing I did when making my first prototype was enable fps testing to cut down on 'put mask on head, take it off again' time
Sorry I might be thick, but how do you test the gripping code without the MCs? do you also just stick the hands in front of the pawn and have some other keybind for grip?
ie: the drivable car, I developed it in FPS out of HMD, and tested it in VR after the initial concept worked
yeah, mouse buttons simulate controller buttons on the FPS char
i have them attached to the body, so they roate with it, but you could have them moveable outside of that too
Alright got it, debugging with VR is a real pain in the Editor what from I saw, this is gonna make it much less infuriating
since the character is just that, and actual character, all of the default turn/move logic works normally with FPS
ug, those late update changes
leave it to epic to stick literally everything they can behind Private:
guess I have to do more overrides
wait what did they change?
just encapsulated the late update logic into a class
instead of running it directly in the camera and motion controllers seperatly
wait, can you still turn it on and off?
yes?
I just manage it manually on a per component basis
so I have to mess with more
ahh ok
was worried they might have removed the bool to turn it off and on per controller
mmm
they are staggering read/write index's out of a 2 deep array
wonder if that would help with the high velocity issue
did they release 4.18 patch notes or something?
ahh
was that for 4.18 ?
@weary flame there can be a mental/optical illusion when sitting on a chair at a different height than the chair is in game, or sitting when character is standing, that can make things seem the wrong size
@dusk vigil thanks i will try that out too
But if the car looks wrong even when the user is standing outside of it and is properly on the ground, then it is 99% likely to be your model
Make sure they adjust their ipd too, with the knob on Vive, slider on the rift
@wintry escarp can you please check Gear VR store and see if my app was published? I don't see it neither through website nor in the Oculus app on my phone.
@mighty carbon what is it titled? I don't have gear but I know someone who does
Phaeton Chronicles
I have a note 4, you think I can install the gear store without a gear?
Phaeton Chronicles: Fallen Titan is the full title actually
nah, you need to plug phone into Gear VR to activate VR software
btw, my app would run on Note 4 surprisingly well
๐
(I didn't expect that to happen)
nice, I didn't even know if UE4 would support it anymore
since it never got anything beyond the innovator's edition I think
I wonder if Oculus manually tick off titles that need to be released and they decided to have lunch before ticking off a checkbox.. It was suppose to be up at 10AM PST
Do you have a direct link now?
they have web links for some stuff:
The ultimate zombie arcade experience...
All 3 Epic Chapters now released!
Over 40 action-packed missions, more new weapons and awesome gameplay
"Drop Dead is the golden standard for FPS games on Gear VR!" VRHeads
"โฆthe best visuals and story development that we've seen on mobile VR." New Atlas
The link on your site just points to the gear vr page
I wonder if that number at the end of the url is based on an appid or something
well, how can I have a link if the app isn't on the store, @glossy agate ? ๐
yeah, it's app ID.. But, what good does it do if it's not showing up in the store
well I thought you were thinking it might not show up in your store because of caching or something
Well your storefront page. If its like steam you have the store front page only you can see until you click release
(why you wanted others to check)
so I just thought the web might be more up to date
nah, it doesn't show up in the webbrowser, nor in the oculus app, nor in VR
@glossy agate it's different for Oculus... I don't have any control over when it shows up. If it's approved and you approve release date, it will just show up on said date.
I just don't want to raise hell if it shows up for other people
๐
yeah, I'd give it a few hours but get to them a good bit before the end of the business day
yeah thats why we were asking for direct link. but I guess you dont get provided with one
I do - it's the app ID after the URL to the Gear VR store..
but
what's the point of using it if it's not publicly available
cause you were asking people to check if it was showing up haha
It works!
what's the price on it ?
it was $0.00 this morning
now it shows as "Purchased" for me
0.99
odd
I wonder why it's not in the Gallery yet
I'll give it another hour and e-mail support after that.. It should't take 2 hrs for a preloaded app to show up in the storefront ๐ฆ
I see .99 too
finally, it showed up ๐
\o/
is wiping off cold sweat
now I need to figure out some PR :/
its in gallery apps
yeah, it appeared ~1.5 hrs later
hey guys, i spawned a simple actor in my world, and it's supposed to stay still. However, whenever I move my head, it looks like it slightly follow me. Any ideas why this is?
@atomic spire is it a billboard object?
its an actor with a static mesh component
the less directly i look at it, the more it seems to be moving
hmm, not sure, post a short video if you can
it isn't attached to anything right, just spawned in the world? and other static meshes look fine?
i have another static mesh in the scene, functioning as a skybox, and it remains perfectly still
and its not attached to anything, no
skybox is usually too far away to see any translational movement
my best guess is the camera isn't set up right and it is rotating but not translating or something, but I would need to see a video
hmmm
you might try looking at the built in VR template and doing your setup like it
have any of you built an experience for both vive & oculus ? I'm finding in the Vive whenever there is a new level loading .. it cuts to the vive space and i can see the trackers for a second
@fossil stratus that's normal, there is a way to change the background during that but it requires c++
* Order is Front, Back, Left, Right, Top, Bottom. If only a single texture is passed, it is assumed in lat-long format.
* If two are passed, it is assumed a lat-long stereo pair. */
virtual EVRCompositorError SetSkyboxOverride( VR_ARRAY_COUNT( unTextureCount ) const Texture_t *pTextures, uint32_t unTextureCount ) = 0;
cool thx.. good to know.. only problem is I'm not a programmer and don't know where & how to make this change? I'm assuming open a file and alter/add this line somewhere? is it just this small snippet or the whole code you linked?
has anyone managed to do proper VR scenecaptures?
I'm trying to make a portal-like thing, and I managed to make it work in 2d
but I can't seem to figure out how to calculate the screen UVs for VR
so everything looks really weird and the eyes don't match up at all
this is my scene setup, the scene captures are in the purple and yellow rooms and they reproduce the same movements the player camera does
and I render that into a plane in the main blue room
@sturdy coral thanks for the help! was able to fix it thanks to u. i just made sure my camera settings matched the vr templates.
Hey guys! is it possible to fix/update the ARKit plugin? It seems it's trying to use ARSessionConfiguration class while it's removed from the latest ARKit and replaced with ARConfiguration
almost as much as Rift + Touch, but no Steam support and controller tracking is awful
why such price tag then?!
Only benefit I see is you could possibly use it with a laptop or backpack pc in a huge space and have tracking
And be untethered
That and ease of setup for trade show demos and stuff
@fossil stratus you will need a programmer to expose it to blueprint
aight, good news
them using the double buffering for the late update scene references fixes the high velocity tracking instability
it must have been thrashing the array adding and removing scene references when the render thread and game thread sync'd up access
sooo, I made $5 with my Gear VR app today.. Sounds like I am done with Gear VR..
10 mil users my ass...
You advertise anywhere?
no money for that @glossy agate ๐ฆ
I tweeted and reddited
but apparently no one except Alex Paschal retweeted
i'll keep hitting twitter
@full junco have you ever noticed this in commit messages:
Copying //UE4/Partner-Valve@3629179 to Dev-VR (//UE4/Dev-VR)```
looks like valve has their own branch finally
but I guess not public like the oculus one on github
they've had that for awhile
ah ok, I've never noticed
from what I can tell it looks like for now they are just using XR to mean both VR and AR in the same sense as OpenXR uses it to mean both
and there isn't actually any OpenXR stuff yet
like the ARKit stuff uses the XR naming now, just because the engine interfaces aren't specific to VR
Carmack said he'd do critical review of my app.. Even if he trashes it, it would be a good PR move ๐
and ARKit is never likely to use OpenXR
@mighty carbon nice, g/l with that
hope your text is all readable
aliasing. sampling, etc. always seems to be the main thing he harps on
they are doing the XR middleware in engine
well, it passed all reviews without a hitch
pretty sure its so workflow doesn't have to change when it is incorporated eventually
I've heard a lot of people had to resubmit several times
true
from what I can tell they just are using XR to mean both vr and ar in the generic interfaces, not openxr specific
and without a lot of stuff in it, he will have to nitpick what IS there
what do you mean by middleware in engine?
they are using it as the net interface for all headset stuff
so they are going to wrap ARKit, etc. to expose an OpenXR interface?
likely
the HeadMountedDisplayInterface got gutted
its all routed through the new classes now
even Gengine->HasHMD is gone
I just dont think those classes are part of OpenXR
no...but its a class middleware in engine
they use like: GEngine->XRSystem->GetHMDDevice()
but I don't think that is wrapping anything in OpenXR itself or even referring to openxr
since it is so similar to the old apis
maybe so, I just don't think it is going to be a mapping to OpenXR specifically
in the mapping from e.g. PSVR's api to the new XR classes, I don't think it is going to have OpenXR related stuff and naming conventions
@tired tree You're not thinking about exposing the skybox override that @sturdy coral linked right? (or is it already in your plugin)
I'm just thinking now XR means "AR or VR" to the engine
Did they?
I guess my project was originally in 4.14
So, I never looked afterwards
Let's see
I don't see it in there in 4.16
def not in 4.15
I think that's a stereo layer
cause I saw a changelog note on that in master
that said it was forced to a 16:9 ratio until their fix
you meant the 3d room and not a splash?
yeah pretty sure there is no generic function for that
easy enough to add though
@fossil stratus You can always count on @tired tree to help out the little guys
would have to think of best way to handle up to 4 textures
4?
likely just 4 inputs and only use 1st if null on others
if you pass in one you can use a cubemap
yeah 6
can be one though
so if not all 6 are valid would have to default to 1st or throw error
looks like there is already code for getting the right kind of the texture type enum stuff in the stereolayers implementation
add textures manually OR cubemap
so should be easy to implement
could just do two nodes, one for cubemap, one for 6 texture version
true
that's less true to the api, but probably less confusing
You can do it
what's grid control?
Show Epic that they need to give you more money
oh opacity and stuff of chaperone?
yeah
DevGrantx2
charles got a dev grant?
Seriously though
didn't know that
hah nope
I thought you got one
i didn't get one?
did you apply? (I never have)
Have you even applied?
yeah i did, just because
I only applied once to get me a Vive when that was a thing
I haven't submitted for money
If I could get something polished enough to, maybe
(on my own)
Because while sharing is caring....
my stuff is kind of out of their intended ecosystem
Nah man
not being in the marketplace or part of one of their games
I think it's bullshit
If they don't consider it
You have been doing so much for the community
While Epic's sitting on their booty
yeah you have one of the biggest threads in the VR forum for your stuff
they definitely include community plugins and stuff and not just marketplace
Well, I haven't seen one for community plugins actually
Only skootumscript
and that's commercial
not really community
(unless I missed something)
meh
mitch got one for his vr content examples stuff and earlier templates, mainly just getting him a vive pre but I think a small grant too
plus his youtube channel and book thing though
Yeah, I got a Vive pre also from Epic
so it wasn't all a plugin I guess
so, with 4.18 if the 4.17 project uses you plugin @tired tree , I'd be screwed or would it be a seamless transition ?
I don't know. I will be honest, without his plugin, my Senior project wouldn't have been a possibility
or any earlier VR project will be screwed ?
motorsep, generally...clean transition, I have to make a lot of backend changes for 4.18 but nothing large on front end
Yeah, and he now is going to include skybox control
been doing network optimization instead of front end overhauls
rahahah
4.18 is probably going to cause lots of small merging stuff for everyone
hell VR changes in 4.18 are going to be easier to merge in than the movement component shit recently
already merged the new late updates in
too good to wait on
aye, cool
how all has that movement stuff changed?
to do custom movement modes with good replication you have to basically reimplement some of the functions
I need to make sure I check that the thing I reimplemeted wasn't changed too drastically
Carmack will be talking for 2 hrs from the stage at Oculus Connect next month
it was something in the save/restore move steps, there was no way to just override cleanly because it was a monolithic function that was like save some state, do replay, then restore state
and you couldn't just add to the save and restore steps through overridding
So ah.. how bad does camera shake make people feel in VR?
@tired tree yeah, it was UShooterCharacterMovement::ClientUpdatePositionAfterServerUpdate()
yeah, for me it was for a jetpack that was replay-correct
im working on a UI set up that has a gaze triggered button. The actual image for the button on screen is small but I want to make the trigger area for it bigger
so it wouldn't consume extra fuel if it was from replay movements
I can't figure out where this option is
@digital marlin depends how violent it shakes and what view angles it affects
in my app I do camera shake, but it only affects location, not rotations of the actor
and it's somewhat subtle and doesn't last long
@mighty carbon I could see Carmack dinging you for that ๐
I think he is immune to motion sickness though so maybe he won't say anything
I am extremely sensitive to motion sickness and I made sure it doesn't affect me
@tired tree does that set splash screen help? (takes rhi texture ref, converts to handle in void FSteamVRHMD::UpdateLayerTextures())
I guess it has to manage the lifetime?
texture resource doesn't have a native information
you mean how like it just assumes opengl if it is linux and not vulkan?
i have to pass in the format and also get the handle
trying to figure out which bluerpint compatible texture variable is easiest to get an RHI lock on
meh i'll go over it tomorrow
@tired tree you can look at the set splash screen blueprint function
that takes a utexture and probably sends the same kind of texture to openvr
"As an indie in VR, what business challenges have you faced with Spellbound?" is a very interesting paragraph in the article
@full junco aye
thus why i was waiting till tomorrow, I should have remembered I was just looking at that function
and you were right
its easy doing what they did
doesn't send the same, but it at least shows the easy method to get the rhi ref
still have to check for if its on vulkan or dx11 though
but thats not terribly hard
@tired tree Give me some of your brain cells please
Btw, Doom is slated to be released on Switch. Crazy level of optimization o.O
@tired tree well yeah I guess that makes those changes a bit more annoying now when 2 APIs need to be supported
yeah, no super biggie
i emailed notes to myself to remember tomorrow when i'm not half asleep
:p
yeah
what's the difference between giving it 1 texture or 6 textures?
well I mean I understand how a skybox in 1 texture works, I just don't how it works in 6 textures. in what way would those 6 textures need to be setup to Form a seamless skybox
its two textures or 6
lat / long stereo texture or 6 sides
12 sides (2 cubes)
?
do you have some examples for how the 2 or 6 textures would look like?
6 images woud be like this
How to export cubemaps out of Photoshop using the NVIDIA Texture Tools plug-in.
Except, not just one file
what's the visual difference between a skybox created with 2 or with 6 textures?
It will take it in order
and process it according, like the comment says
I don't think you get the issue
If people already have a workflow for how they do their skyboxes
why would you make them change it
which is probably why they allow the choice
depends @full junco
if you use equirecs, you might have warping at the poles
if you use cubemaps, you will have no distortion
@granite jacinth @mighty carbon @tired tree ah, thanks
lol, I thought I lost registration info for my copy... Good thing we are talking about it! ๐
(found it, updated the app and re-registered)
4.18 should be branched soon I think
will be interesting to see how vulkan performs in 4.18
well, I wish iD Software licensed their id Tech 6 - Epic would have to optimize UE4 really quickly ๐
(hint - Doom 2016 runs on Switch)
people use UE4 because it has "good" doc, it has blueprint, the source is available for anyone , it has a very nice editor interface and so on
even if id Tech 6 would be available for anyone to use and be twice as fast as UE4 most people would still use UE4
yeah, I know.. But there is no excuse to have super slow dynamic lighting
or bogus AI
reading #gameplay-ai makes me wonder why Epic neglects that area of the engine
well, the dynamic lighting isn't "super slow". its average. id tech is just surprisingly good
ha, in 2018 Wolfenstein 2 will be released on Switch.. id Tech 6 is a marvel tech
since I have my gtx 1080 I cant really complain about gpu performance in UE4 any more, it runs great ๐
with dynamic lighting and VR
lol, I played Doom on my old 670 and it ran smooth enough
I am gonna test dynamic lighting on my 1060 in VR, but vblanco said UE4's dynamic lighting isn't suitable for VR
well, not if you need multiple lightsources
if you limit yourself to 1 directional light is acceptable
you should try Doom 3 BFG VR mod (recent one)
lighting is fully dynamic, point lights, spot lights, directional light
lighting/shadows
its also all indoor and corridor for the vr one isn't it?
I wonder what TimS does at Epic.. I am sure he could have come up and implement faster dynamic lighting for VR in UE4 ๐
well, it's Doom 3 BFG - whole game with VR support
it's mostly indoors, yes
which is where you'd use point lights and a lot of them
Doom 3 ?
cooridors
it's portal based
so unless you manually place portals and have sealed level, it won't cull what's behind the walls
doom3 uses UMBRA
it also has weird culling.. Let's say you stand next to a continuous wall and looking ahead the wall is lit. So it will render entire mesh, even outside of the view frustum
Doom 3 uses no middleware at all
Doom 2016 uses Umbra
(Doom 4)
yeah
I worked with idTech 4/5 (Doom 3 BFG has a mix of both engines) for several years..
nah, BFG Edition has rendering from RAGE
id Tech 5
threading, rendering, input, audio and file system
the rest is old from 2004
aye, but what difference does it make if it looks good ๐
thats the point
(and runs fast)
well, a lot of modern Unity/UE4 based indie games look worse than Doom 3 BFG
lol no
hell, worse than Doom 3 from 2004
and I don't mean PUBG
anyhow, I'll test dynamic lighting in UE4 VR myself and then let ya'll know ๐
Planetrism is all dynamic and open world for VR. Runs fine for them. I have ultra dynamic sky running in mine okay, just takes some work.
Not sure if anyone else has tried this but any hdri skylights make everything green in VR for me
one of my HDDs just started to try communicating with me. making a lot of weird noises ๐ It's dead. Its just one where I mirrored another HDD to, so its not an issue, but let me remind you that HDDs can and will die, the question is just when.
I mirror my HDDs to other HDDs and I use two different cloud backups, so I can't really lose anything. and code is obviously online in source control
Does anyone here is working with the UE4+ARKit (iOS) currently?
BTW is this channel ok for AR questions? (not sure since it's titled just a VR)
it was created before anyone used AR with UE4, so its essentially also meant for AR
Hi guys . I am working in a mobile app for daydream and i parented my game gun to the google vr motion controller but its not moving with the Controller
Is there any other way of doing this
I had to attach mine on begin play. It didn't work if I simply patented it in the hierarchy window
But that was for Gear VR
What do you folks think about doing early access for a project?
well I think its the only thing that indies with no funding can do
theres no alternative
'There is only one launch' - so the game damn well better be in pretty good shape even at early access stage
Do softlaunch on Itch or Gamejolt like GORN did
@dusk vigil whats your definition of "softlaunch"?
So called softlaunch is in no way something that helped the game.. I recall that Devolver is behind GORN.
So for indie without funding such move could be very well a suicide move
(Been there, done that)
What amount of content do you expect from early access VR game?
Superhot was a Flash game. Gorn was on Itch. No suicides there...
Gorn was memed to death
its an example of a very "streamable" game
funny videos get published so they spread
Does anyone knows how to implement Secondary camera within vr For example i have a motioncontroller and to that a camera is stuck and i am recording inside the vr preview like holding a handycam
For example Climbey came out with rudimentary multiplayer and level editing right from the start in Early Access
no youtube but caught fire on reddit
I'd go with simrak, you only got one launch on steam. If you want feedback and don't need money, release it elsewhere, when it's already a fun experience in EA, release on steam and try to get some momentum
@mighty carbon it should just be better than any existing game in some relevant area
@dusk vigil PUBG started as mod.. Let's not use outliers that had advantages as a norm in indie game dev world
@mighty carbon Wait did it? What was it a mod for?
the guy behind it started working on battleroyale mods, but it's PUBG for say, just battleroyale mods
And the guy behind PUBG isn't someone unknown, like many indies are
well his pseudonym is PlayerUnknown ironically, but I reject the notion that his "fame" had any traction, most people don't even know why it's called PlayerUnknown's Battlegrounds, it's just a solid battleroyale game
any significant traction anyway, it had some for sure
I have literally never heard of the guy
It's a Sid Meyer situation, sounds silly to me, he could just have called his game Battlegrounds, but w/e, his game is selling better than anything else
Having a name in this business, track record and thus followers has everything to do with success
I've never heard of PUBG or the guy behind it until just recently
at some point people just suddenly started talking about it and I had no idea why
and I still don't know why
Because people have been wanting real bad for a proper battleroyale game for a long time, and every implementation was subpar and broken
H1Z1 was selling really well before it came out, then PUBG canibalized all of it's sales because it was superior in many ways
if you're not interested in shooters or battleroyale game, follow twitch streamers/youtubers and don't read any video game news, you won't have heard about it
well I scroll through popular content on reddit, and at some point gifs from that game started to show up
and a bit later friends started playing that game
I never even looked at any video or gif of it yet, I try to just ignore those hypy games until they are forgotten again
Player unknown made the arma battle royal mod
and basically started battle royals
thats why his name had weight to it
ah, well then he deserves it
@mighty carbon I still think a couple of devs I know would have benefited massively from playtesting their product on itch before launching on steam and nosediving straight into the ground
Arma 3 is the only game I need to be happy
he consulted with H1Z1, and didn't like what they did, much like icefrog and HoN
you can do pretty much everything with Arma
so made his own game
its also fun....too much fun.....
@echo thorn I did that, made an actor that is basically a cam corder with a button and when the button it on it turns on the new texture mirroring mode in 4.17 and draws a scene capture
pretty simple
@tired tree well I wish I would be able to create fun games too haha. I am much better with programming than with thinking about what's fun in a game. I have no idea how to create games that are fun.
I still don't think his name is really the root of it's success. Battleroyale games are just the kind of game that is gonna be successful because it's an excellent game to stream, and just as fun to play yourself. Twitch kickstarted the sales and it went exponential
such is life
I just hope people will tell me what they want for it to be fun
no, but his name is what got streamers involved
and helped get the word out
the top streamers on twitch all knew him from the Arma mod
and played it early, and consulted
yup, that was a important factor
his battle royale was Playerunknown battle royale mod or something like that
when you get Lirik possibly actually funding you...you have a huge boost right off the bat
@tired tree , True, especially that one streamer that kept going on about how it was just so much better than H1Z1, can't remember his name. But PUBG got developped really quickly, and I don't think it could possibly have failed
well it IS better than h1z1
h1z1 has been terrible since launch
which sucks because I had hopes
if he was the one creating the Arma mod then it's a huge thing and the most obvious cause for success. just like dayz standalone sold super well too because it was done by the same guy who did the dayz mod
@tired tree well were you able to record videos and the main thing were you able to make a 360 video of what you were capturing nd if yes can you tell me how this is possible . any projects any tutorial??
his skill and understanding of what makes battleroyale games is yes, I'm just really not convinced his name had much to do in it.
@echo thorn 360 videos real time while in vr? not going to happen
either way, it is really fun.
at least not at any decent quality
yeah i know good quality has to go but before this all you said like making an actor , do you have a tutorial for that or a better explanation
@echo thorn if you really need 360 recording you're probably gonna have to save a replay and possibly render it in video offline. you can't cull any of the geometry, it's gonna be way too hard on your GPU
......its just an actor you hold and when its activated it changes the desktop mirror mode...
Custom 360 ingame camera could be a cool feature though : )
Ohkay guys . 360 has to go
But yeah, most current solutions take a long time to make a still scene of any reasonable quality
Ahan @tired tree
Testing mixed reality support for the camera actor VR Expansion Plugin: https://bitbucket.org/mordentral/vrexpansionplugin Consider supporting me on Patreon:...
pretty simple
don't have to have the working view window on it like that, its overkill
Cool stuff
Marketplace material...
Filmtyme's Hackerfest 24hr hack vignette
@tired tree well i was thinking about this to make
@dusk vigil its free, no need
oh....thats pretty much exactly what mine is
its easy
though if I were doing it for something like that, I would make a steadycam system for it
i asked them what they did basically they just added screen capture component 2d. and yeah yours same
your' s look way better
@willow trail how long have you been an indie dev? I started in 2008 and I can tell you that name and connection in the industry have everything to do with success of your indie game
sometimes even money doesn't mean as much as name and connections
@mighty carbon I'm not a game dev. I never said it didn't, if I hear about Supergiant's new game I'm gonna be hyped without even knowing what it's about. I'm saying that in the specific case of PUBG I doubt it had much to do with it. I might be wrong, but I can't see that game fail
@tired tree free on your github? I mean, having this sort of thing on marketplace would make it more easily accessible though, since people wouldnt have to integrate it into engine builds?
@mighty carbon if you're an indiedev I have no doubt connections are gonna be absolutely primordial
I used PUBG as most recent example.. There are plenty of those indie games that made millions and a lot of folks make mistake of thinking name and connections have nothing to do with their success.
@dusk vigil its part of the example template for my plugin, I don't want to break it out, uses some of my specific things like a dial for the screen so it rotates, ect. Would be one hell of a hassle keeping the functionality and trying to tie it in to the lax VR template that epic has.
Well i'm still torn on this issue, because I've never seen a game that I really enjoy being unknown, and I like to try weird obscure stuff. But I suppose that's only when your game is done.
one can make a mediocre game, but if one is someone with following and has connections and voice and name, guess what? It will be all over the media shoving it down your throat as innovation and the greatest game ever build
its a plugin though, not an engine build
have you tried Pubg? it is a terrible example of a name making a game
the game is actually fun
I haven't - not my type of game
@mighty carbon Yup, and that sucks. But I'm much more concerned about someone making an awesome game and it being lost amongst the hundreds of daily new steam releases. That seems to not be as much of a problem
I just recently finished Deus Ex MD that bought few years ago o.O
@willow trail that's actually the most common scenario in 2017
I can't think of a single example though
lol, huh?
VR being being an exception, since the usebase is so low/innactive
@tired tree ah ok, I check that out.
@mighty carbon do you mean that good game being burried is the most common example?
I dont know... heaps of Finnish games have made it to the big league without being superbly connected, particularly on mobile. PC games have had a much tougher time
first of all, a good game is a relative definition. A lot of people went crazy for Minecraft and I hated it from the day one, before it because famous and a gold mine.
I just linked an article yesterday about that indie dev who made Spellbound and how it flopped
Maybe we are getting a bit far off topic for the VR channel though
@dusk vigil well, business side is critical unless you don't care about money ๐
hmm no business channels here true
My question was about early access and what you folks expect early access VR game to have in order to be invested in it
you might have hated minecraft, but you are by far in the minority
you might want to think about how tied in to what the general consumer actually wants....
that game had the broadest appeal of any game made, hardcore FPS and tactical gamers played it...
@mighty carbon Yup, loved minecraft since the very first alpha posted on TIGsource by Notch. Hating has become a meme because of the community, but Minecraft is special anyway. To get back on topic. I could list a couple of things that have become mostly common in regular 3D games that I think should hit the VR market. If you make an action/fighting/melee games, having good defensive/evasive mechanics that make smart use of the motion controllers, I except your game can be a great success
Also...the guy that made spellbound is from the forums
there are arguments for why it flopped..
Good article that one yeah...
@willow trail like this? https://www.youtube.com/watch?v=2rQZPix34bI
im working on it
Also f*ck dual wielding of guns in VR. It's seriously useless, but that's just my opinion ๐
Well, I was told I'd be better off making a bite-size experience for Gear VR.. Theoretically, between experiences and yet-another-wave-zombie-shooter, experience sounds like a sound bet. Yet, it's just not selling, despite people saying it's good for what it is and even Oculus noting it's interesting.
you might have undersold it @mighty carbon
So, might as well do what I like and do it in early access, than worry about what masses like and have same result (or worse) at the end
by making it 1 dollar people will come with the mentality that its a very small thing
I reckon the main thing with the launch is that you have assessed that 1) people can pick it up easily and get right to the meat of the game pretty fast. 2) they have some reason to come back to the game. It's hard to lay down any feature list of what must be there or not...
if you can't come up with a real idea that you want to do, why are you even going into it in the first place?
I think people can only go through so many experiences before "getting" VR, and needing more, games don't suffer than problem I think.
games are HARD
there have been a ton of devs doing the "make EA, have customers tell me wat they want in it later" games in VR
they all flop
im seriusly bottlenecked by the lack of artist or level designer, plenty of the prototypes i have can be made full games
@tired tree DWVR didnt flop becouse it was done on no budget, but can confirm
players are stupid
@tired tree my "real" idea is A. time consuming to make and B. requires a lot of content
when you ask the community to make choices for YOUR game, they expect it done quickly, and how they want it, and everyone wants it differently
its a terrible idea
"but how can you fail when we keep telling you what direction to go with the game?!"
One Finnish company made 'Downward Spiral', I reckon they flopped because they had a puzzle section to start, even though their 'meat' seemed to be a multiplayer battle mode, which likely nobody every made it to
I wouldn't ask people what to make in my EA game.. I just need people to engage, start spreading the word and get cash flowing
and then what?
you don't promise anything until cash?
how are you supposed to hype that
??
people are going to judge the END product, by the FIRST offering
Sairento seems to be one of the rare games that had a lackluster start but has managed to keep at it and improve sales
The game will be defined, with bullet points of what's coming.
Anyone here a PSN developer?
Promises are extremely attractive, I don't think they are necessary, but IANAGD
and it will be definitely underpromised, because then it would be easier to overdeliver
attractive to the dev that is
and if it tanks, it would be just easier to wrap up
@uneven moon ME
and also require you to make a new name if you intend to try again...
and if it succeeds, then it would be even better and more hyped when overdelivered
because otherwise that will be thrown at you forever
the reason for releasing that early
If you want to make an impact, you need to look at your competition and say 'what's the new twist I have that makes this interesting?'
is to have an excuse to release early
if it works, nice, you conitnue
if it flops, you cut the losses there and GTFO
well, if my grand VR attempt will fail, then perhaps I should not make games after that ๐
thats why survival devs have done that
the moment the game is "shippable", they ship it
even if incomplete
honestly I don't think EA should be used unless someone has a real innovative idea and a solid core done
You want to have your core gameplay, the best you have to offer in place. Maybe you dont need all the achievements and heaps of levels etcetera
its pretty much the only place it does work
but you need to have the core gameplay there
nonsense @tired tree
a core that people keep playing
lol
then you add stuff on top
how is it nonsense?
@tired tree That's why you throw theme, world building, characters and story out the window and focus 100% on game mechanics before fleshing it out ๐
name an EA game that worked that came out without something special
hint, most EA games don't work
even innovative ones
because if one works on the game behind the curtains, without any feed back, there is a good chance no one will care for the game and may find it flawed. And the game might tank big.
EA allows you to correct direction early on
That's why I recommend Itch/ Gamejolt first to gauge your level of readiness for the real launch
thats why "solid core" not "nearly complete" is what i said
Steam EA doesnt really mean anything anymore, everybody does it ( almost )
Aren't Sairento and Raw Data EA games ?
no core = anything is possible to the consumer and they will expect anything an everything, or they will just walk away, either its a heavy disappointment to them, or they just don't stay.
sairento also released with a base game
Yes, but they were pretty damn well polished on launch
Sairento nearly flopped but they stuck to it and are making do
Okay Raw Data was polished : ) I didnt actually play a lot of Sairento early days
ah, raw data was
raw data was a launch title btw
@tired tree the question is, again, how much is enough of core to be in the EA game for people to invest into it
enough to play, that depends on the gameplay and what it is
you can't go EA as a solo dev with no name with a room with some lore and voiceover talking about what will be
multiplayer IS easier, because as long as you keep people around somewhat, it provides its own content
It's a pity steamspy doesnt show history farther back
why not? (besides it's a lot of work to do that to begin with)
multiplayer is easier to make, harder to sustain
even if prices at $1, its going to be forgotton, and its not going to be hyped
(easier to make from low content perspective that is)
multiplayer is volatile
multiplayer is super risky though, if there are no other players, you are sunk again
as hell
thats why i will make the game coop
maybe a pvp mode, but not the focus. Coop it is
even big games like Star Trek have not got so great concurrent player counts
btw, indie games with multiplayer can grow, indie singleplayer games cant
players that like the game (multiplayer) will nag their friends to get it
back in 2011 when we released Steel Storm, we raked up thousands in the first few days... No PR, no Marketing.. It' snot the same anymore, even though VR is emerging market :/
I think risky is an understatement. Even on Pavlov and Onwards I've never seen more than 2/3 servers with people. That vast majority there's only a single other dude playing in EU.
for a singleplayergame that doesnt happen
@mighty carbon blame steam ๐ฆ
ill come back with stats of my PSVR launch
steam isn't only to blame
freely available easy to use engines and an indie scene just coming out of the golden years shits out a lot of drivel
we released Ep1 before Steam, in 2010, for free.. It got 300k downloads.
The way I read it, VR gamers want story driven single player games, because there are already too many MP games
story driven singleplayer is impossible to do for indies
thats why ill do a looter game with coop and rng maps
currently still prototyping the mechanics, i love that melee combat, but it has its issues
its not impossible, its just a lot of time investment, and in VR taking a long time investment means the technology passes your game by
Don't take any cream of the crop devs and get MP-only other indie games, contact devs and ask them how it is.. I bet you they don't have good news to share.
like I said, I am not MP person myself, so I doubt I can come up with something awesome for MP game (since I don't care for MP)
the only way I see it is to come up with decent story, make an interesting first level or two, release to EA and see if people bite
i do see there are no PvP VR games on PSVR, like sub 10
but damn, i dont want to deal with PSN
if they do, just keep making more content and unfold the story
better be a puzzle game if its a level or two
(and I think AI and interaction between AI and player would be the key to success in SP game)
Onward/Pavlov wont ever work on PSVR, i thought of making my own (but with my own twist of course)
well, it'd have quest/puzzles.. Not just wave combat.
I personally find exploration in VR more rewarding than combat
but tracking limitations, + no joysticks + PSN sony stuff...
honestly, why are you guys even trying games? splat some barely clothed female models into a room, release a new outfit for them a week, and call it a day >,<
well, Steam might ban that, Oculus won't even let it to the store..
it seems in this day and age if your fart doesn't smell like fart you will be condemned online
so many anime girlfriend games...so much money rolling in to them, its pretty sad
thats how you run a successful EA
heh
My Finnish friend doing VR porn : ) https://www.patreon.com/femdomination
MordenTral do you expect to add some soft bodies physics to VRExpansion?
maybe I should add anime sex bot into my VR pet game ๐
random question
i still want a final fantasy style game with turn based combat in VR
might even make a prototype
@dusk vigil why Patreon? I only want to pay once for pr0n games ๐
It seems to be working pretty well for him
I bet.. Subscription based model
He's had a long road to get to this point though
btw VR killer ap is SS13 VR
just throwing it out there, I wouldn't get out of my headset if someone made that
perfect fit for VR too
Fun talks still, I gotta head out into the rain
SS13?
SpaceStation 13. It's... hard to explain
it's like some sort of competitive/cooperative mafia/werewolf like game, with tons of tools interaction, heavily realiant on communication
wasnt it a DF like in spaaaaaace?
oh but that exists
werewolfs within
might be what you mean
btw that game is WAY too volatile
it 100% depends on enough players
extremely volatile yes
and seems niche
it has to be a good implementation
niche + vr..
I don't think werewolf within is really what I mean
SS13 is only niche because people are afraid of the dumb 2D pixel top down graphics
also everything else... It's kind of niche yeah, but I think it'd pull people around
SS13 is extremely niche but there constantly are player on, because if you get hooked, you get hooked, it's really complex
Little question, if someone has any good ideas ๐ If I'm making a procedural action descent game, I only intend to use the dominant hand to fight, non-dom hand buttons will be used to trigger some powers (slow-mo and the like), but I don't want the non dominant hand to just sit idle by your side. What's a good use for non-dom hand motion beyond sign tracing (which I think will get old very fast) and defense/shielding (which I'm afraid will trivialize the gameplay)
My biggest problem is that if you use two weapons there are only a few possibilities.
-
Same weapon on both hands. I don't like this idea, the player will only end up aiming at the same target with both hands for double DPS, I think it's a poor use of the left hand. Pretty much everyone would prefer double damage on a single hand instead of that.
-
Different (ranged) weapons. Same problem, but even more confusing, reloading will be all over the place, and you rarely use two different types of weapon (shotgun and pistol) at the same time
-
Melee and Ranged. Same problem, enemy comes up close you thwack him with your melee weapon, if he's further you shoot him. You could use a single hand for both action with almost no difference. (@wicked oak I looked at aquila and realize you also use pistols up close, but I'm not really satisfied with that idea)
sorry for the essay, I just don' t like dual wielding very much
This is all after playing a lot of Sairento.
another prototype just had a shield
ive thought of using the offhand for abilities
probably shield/ability could do
current enemies are incredibly aggresive so fairly hard to parry with only the sword
My problem with shield is that they tend to obstruct view, make it weird (you clip your weapon through it or have to reach around it) and can trivialize gameplay. I don't like shields too much. It seems like a more reasonable option but I'm hoping for something better
ability would need you to point at enemies which again I feel could be done with the dominant hand... Reloading of weapons is also an option, but I don't intend to have any reloading, just scarce ammunition
Movement is another, but I don't like teleportation too much, and my experience with Sairento is again that throwing the movement arc with your shooting hand changes practically nothing (especially since it triggers slow motion)
Well anyway my best idea right now is to use it for either crowd control or item collection (some sort of vaccum or another), but I'm still not 100% on it ๐
Heh I'll mess with the grappling hook idea maybe...
interesting talk about early access you had here ๐
@wicked oak don't say growing a singleplayer game is impossible, I need to make that work ๐
@full junco static vr::Texture_t CreateOpenVRTexture_t(UTexture * Texture)
{
vr::Texture_t VRTexture;
if (Texture)
VRTexture.handle = Texture->Resource->TextureRHI->GetNativeResource();
else
VRTexture.handle = NULL;
#if PLATFORM_LINUX
#if STEAMVR_USE_VULKAN_RHI
VRTexture.eType = vr::TextureType_Vulkan;
#else
VRTexture.eType = vr::TextureType_OpenGL;
#endif
#else
VRTexture.eType = vr::TextureType_DirectX;
#endif
VRTexture.eColorSpace = vr::ColorSpace_Auto;
return VRTexture;
}
https://80.lv/articles/mawi-united-wet-lands-landscape-in-the-works/ << I wish it worked in VR ๐
i pre-checked for valid texture prior to that, should probably do it in the function as well
also...uses the wrong name there
but thats how anyway
ha, I dont check stuff for validity, crashes are easier to debug than "oh nothing happened, why dont I see the correct texture" ๐