kinda wondering if i should go for implementing player characters with variable height, or if i should even it out so every1 has the same height, then do some 'scaling' (on z axis, to the recieved positioning data) to reflect their ingame height?
#virtual-reality
real needle
reason im thinking about giving all players the same height is because it seems more fair towards tall and short players 😛
raven halo
@mighty carbon what kind of problems are you having?
mighty carbon
raven halo
oh shiet!
mighty carbon
@raven halo I am getting CPU stalls, a lot of them
real needle
well it is a mobile vr solution, with all due respect, you shouldn't really be expecting performance
for starters, try if its any different when the phone is plugged in the charger
raven halo
@mighty carbon when are those stalls happening?
real needle
im sure theres some very rigorous throttling going on to keep the thing from overheating/sucking dry the battery
raven halo
there isn't
mighty carbon
I don't overheat
raven halo
well, by default epic sets the CPU to 75%
I have it set all the way down to 25%
and the game runs fine
need to be more careful with draw calls, though.
but cpu is what causes the phone to overheat the most
mighty carbon
@raven halo this is my scene
this what profiling results https://www.youtube.com/watch?v=P-Ylh6OfPwk
wicked oak
motorsep. the WaitForEvent dont count
it means that current thread is not getting executed
raven halo
what ms times are you getting on GPU and CPU?
raven halo
there is not much going on on screen, it should run fine. So there must be something wrong
raven halo
:S
mighty carbon
GC ?
and what do I do about World Tick Time ?
@raven halo is it stat Unit that gives GPU and CPU timing ?
mighty carbon
didn't take a screenie with UnitGraph because all it was showing the same as Unity + graph at the bottom.
yeah, 4.14.3
raven halo
gpu doesn't show up in 4.14
only 4.15
you can't do any sort of gpu profiling with 4.14
mighty carbon
can't do 4.15 until FMOD releases the update
mighty carbon
too many tris or drawcalls or both ?
raven halo
could be pixel shader as well
draw time = probably too many draw calls
gpu, could be anything. Too many triangles on screen, expensive shaders, post process?
tired tree
is your landscape material opaque?
mighty carbon
yep
tired tree
noticed a few VR devs using marketplace terrain materials that were really expensive and masked so they also didn't cull geometry
raven halo
you should take a frame capture with renderdoc and take a deeper look
wicked oak
no, thas a GC error
probably due to the memory of the landscape
thats not on the graphics side
wicked oak
didnt thought of it with 130 ms on GC mark and sweep?
tired tree
they were talking about GPU time weren't they?
raven halo
I would never recommend using landscape, just in general, it has sucked for 8 years, and I'm sure it will continue to suck
tired tree
how large...is the total terrain...
raven halo
@wicked oak I was looking at the ms readings and got tunnel vision
tired tree
it won't LOD it if its huge and close to you
mighty carbon
it will
raven halo
how do you handle collision?
mighty carbon
each chunk has 3 LODs
wicked oak
the meshes shouldnt be any kind of problem
mighty carbon
the chunk you stand on won't LOD of course
raven halo
polycount could be an issue though
raven halo
I've had, just last week, issues where 50K triangles on screen were causing frame drop because they were pretty concentrated on screen
mighty carbon
everything is UVed
texture is tiling and is 2048^2
eeh, I take it back - it's 1024^2
normal map is 2048^2
is there a stat that shows how much memory textures and models take up ?
@raven halo 100k tris is what S6 with Gear VR can take easily
my scene is ~60k when I stand at the point where performance drops
raven halo
(not always in my experience)
mighty carbon
so, is there a stat to see how much memory assets take up (preferably with breakdown for textures and models) ?
sharp swan
stat memory
real needle
stat rhi
noble crater
how would you go about making a vive controller mesh match user input e.g. virtual trigger depression = physical trigger depression? skeletal controller mesh?
real needle
@noble crater If you can rig it then a skeletal mesh running an animation blueprint would be the way to go, but you can also just rotate/move each individual mesh with some math. Have a look within the folder where SteamVR is installed and you can find the vive controller with each individual piece as it's own mesh
noble crater
@real needle thx
mighty carbon
alright, I did another pass (added culling distances for terrain chunks which brought drawcalls down). Here is what I got (performance still shitty):
ha, wtf... in PIE I did bring number of drawcalls down. On the device - it remained the same 😦
Curse you mobile hardware 😦
that's all I gathered
@wicked oak @raven halo @tired tree any suggestions guys ?
raven halo
are these screenshots taken from gearvr?
why is the ms so high and in the video so low?
mighty carbon
@raven halo yeah, running on the device in dev mode (so I wouldn't have to use Gear VR - wouldn't be able to bring console and type cmds)
so what do I do with tick ?
tired tree
figure out whats causing the issue, you profiled it earlier and it was GC right?
tribal citrus
Sup guys, have you've gotten Error unknown when trying to build your game for VR?
mighty carbon
@tired tree vblanco said it was GC.. I don't even know what it stands for :/
tired tree
Garbage collection
mighty carbon
ah, I see
the thing is that I don't do anything special on tick.. I have 2 traces happening and that's about
and when player moves away from the robot, there is nothing else happening
restive blade
arghh... I don't have enuf space to set up oculus touch
tired tree
you aren't spawning objects and deleting them?
mighty carbon
nope
tired tree
have you tested in a shipping build?
I know that verifying GC in test builds with mobile is supposed to hitch a lot
mighty carbon
no, haven't tried shipping build yet
never had this issue in 4.13
I guess I'll post on Oculus forums... This is just bizarre 😦
mighty carbon
looks like S6 owners are in luck with new Gear VR controller 😃
101 deg. FOV, better lenses and some distortion correction tech? (maybe to replace old code that Zeni is trying to use as legal reason for their injunction)
wintry escarp
real needle
@tired tree It's not, and if they continue to be a thing it's important that they are named for what they are
mighty carbon
it's a motion controller alright - you move your hand and your virtual hand moves too
thus - motion controller
there is no depth though - it moves in a plane and can rotate
3 DoF
and I am sure 98% of Gear VR users will be satisfied
when there are like 50M - 10M users, then maybe it would make sense to make something with positional tracking and proper motion controllers
real needle
Just saying that if we call 6dof and 3dof for "motioncontrollers", there can be quite a bit of misunderstanding
mighty carbon
but by then Santa Cruz will be out
real needle
We have a motioncontroller devkit for mobile VR, it doesn't work very well but it's def better than none
mighty carbon
@real needle you already have Gear VR controller? o.O
real needle
No, Ximmerse
mighty carbon
oh, that one
I already forgot about it 😃
IMO Nolo VR already beat all of the 3rd party solutions for mobile VR input
real needle
Yeah I've seen better, but haven't tried anything else
We didn't focus on the Ximmerse because it's not as good as it needs to be
And generally speaking mobile VR is not as much fun, so if I can choose I'd go PC every time. I don't feel present without pos tracking
you guys seen this yet? https://www.kickstarter.com/projects/morgansinko/hardlight-vr-suit-dont-just-play-the-game-feel-it
'feel vr' 
prime iris
definitely don't see chest subwoofers as a particularly interesting technology, in the same way they were pretty crappy as seat peripherals for racing games
that said, more tracking points would be nice
hands > elbows > chest > feet > knees > waist
feet maybe being a little higher priority if a faked alking locomotion turns out to be compelling
and by hands i mean some degree of finger interaction in place of controllers 😄
real needle
ye we just need gloves and feet trackers, dont really need the rest
prime iris
i think body/arm tracking might be slightly better than feet tracking. being able to have a virtual gun holster or something is good, but doing that well without extra tacking points for your torso/waist/thighs gets akward
real needle
its not that hard tbh
just get the angle of the 2 motion controllers (get angle between motion controllers, find 'forward') and get angle of hmd, then body sits between those 2
prime iris
that isn't where you know your body is though
real needle
tru
prime iris
which means you often have to look down
real needle
well, if it rotates along with the hmd and controllers predictably you wont have to look
where most ppl seem to mess up the implementation is when they start adding movement direction too the mix
wich really shouldnt be relevant, since instead you should look at letting your characters walk in all directions
prime iris
well it's a not enough information problem given the current number of tracking points
prime iris
how do you tell if a player is looking left and have both arms pointing left shooting a pair of pistols!?
real needle
it is when u add a little logic, humans cant rotate around their own backs, etc
you use the roomscale data for that
prime iris
pfft! show me your 3 point skeletal tracking system and i'll show you how to break it! 😄
real needle
no ty, im still to unsure about it myself xD
prime iris
you can definitely do it well using just 3 points
but i think there's enough value to add more tracking points before feet (unless VR walking works as a good locomotion method)
real needle
seems to be holdin up tho, and yea, doin just that, calculating the angle between the motion controller positions, then use the hmd angle to help decide on a 'forward' position for them controllers, and then also use hmd angle, and then i just go sit in the middle between those 2 angles
ye i do wanna see better tracking too
but its not as impossible as some ppl/some games make it seem
@real needle @prime iris They're not subwoofers, they're pretty powerful rumble motors with a wide range of haptics
The sdk allows you to play them in sequences, and do interesting things like heartbeats
It's a great experience, and it allows me to communicate in which direction players get shot from
prime iris
@real needle i'm sure it's a neat addition to the VR experience but i think body tracking needs to come before it
real needle
It does support "audio feature" to make it playable with any game, but that's only a bonus feature. Best execution comes from direct integration. Otherwise one can just use a subpac
prime iris
it's $500, body tracking needs to be cheaper than that to see the adoption that will make it worthwhile
similar issue with the gloves that have been showing up frequently. They're charging 1k euro. Should have sent a poet because all i have is this sad face :'<
real needle
I agree it's a bit hefty, but if all of us who said we will be supporting the suit at launch there will be a good range of games supporting it. It's niche, like most of VR right now 😃
prime iris
if there's a market for those vests it will be at commercial vr centers
vr lazer tag places
real needle
I can atleast tell that these guys are gonna pull it off, they've already manufactured different sets of units with just their seed funds
I've also helped their plugin dev with the integration for unreal and it's pretty solid
@real needle i never said there was subwoofers, i just said 'look at dis <ks link>'
prime iris
meh. if $500 is them selling at cost to try and get investors then the product is destined to fail. the dk1 was cheaper lol and provided signficiant more value in terms of what it did
real needle
@real needle bushido did, sry for the confusion
prime iris
i made the comparison to the old subwoofer cushions 😄
real needle
^^
If you guys can try a subpac I would recommend that
i do like the idea tho
a suit that has little piezo speakers, them things the size of a coin
and then like a 100 across the suit
also, some1 please make a version with some sort of solenoids/linear actuators, so u can simulate bullet impact xD
(like, actual weight moving into your chest lol)
would be insanelly heavy to wear tho, but if somehow it could be done it would be the best thing ever
prime iris
pfft, just hire a kid from a 3rd world country to shoot you with a paintball gun when you take damage!
real needle
lol
real needle
@real needle one of the hardlight prototypes had linear actuators in the arms, it worked alright but they would need more powerful ones to actually simulate recoil and that's why I think they decided to not do it
so hey, question, should i have scale-able player character height, or should i scale playerspace z so i can have all players same height without their cams/controllers beeing in wrong places
@real needle yea realisticly it would require some heavy weights wich would make it shitty
prime iris
@mighty carbon what do you expect it to do?
mighty carbon
@prime iris what do you mean?
I had horrible unexplained performance issues earlier. Posted profiling results and screenshots of various stat cmds taken on the device.
had something to do with tick and GC
I posted here https://forums.oculus.com/developer/discussion/50098/gear-vr-odd-performance-issues-with-4-14-3
so then I remembered I had all these inventory items stashed under the terrain mesh. I move them around and performance issues were gone.
I don't get why, since those were suppose to be culled anyway.
real needle
A question for those that have put a game up on steam - how many Oculus users do you get?
is it worth the effort to make a game work for them? (assuming it is a standing-friendly game)
its always worth it to try reach more customers
but no experience releasing to steam myself so lets see what awnsers u get ^^
but do oculus users actually use steam? or is "no home, no go"
i think they do
any1 care to pitch in on my player height question?
prime iris
@mighty carbon ah, i didnt see the earlier stuff. only saw the video you just linked where you moved around some primitives under a landscape
real needle
player character scaling vs 'scaling' the tracking z data so all players can b same height
digital marlin
Anyone, aside from Zap, at GDC?
prime iris
@real needle assuming you're using ue4's VR abstractions in your game i can't see how it would hurt to make an oculus home version
assuming the rumours are true oculus do have 1/3-1/2 of the desktop vr market, so if your motivation is making sales then it's probably not a terrible investment
real needle
i would assume oculus has at least half the market, if not more
theres just alot more people that have heard of oculus
prime iris
stats and rumours say no
real needle
vive/valve/htc dont do nearly as much marketing afaik
prime iris
tim sweeney hinted that it's 2:1 in the vives favour
real needle
hm
real needle
hmm, i see, i wish he/the article would have something to back up those statements tho
prime iris
real needle
i mean i have a vive too and tell every1 to get one instead of oculus lol, but i doubt im making a dent in facebooks marketing budget
prime iris
it's probably not an unfair assumption to say that if you have a pc that's good enough to run a VR headset you'd probably have steam installed as well
real needle
any1 remotelly interested in vr will see ocolus ads all day
provided theyre in the social media circlejerk, ie; 70% of the world population
prime iris
advertising doesn't really matter in the current market. the market is enthusiasts
real needle
hmm
prime iris
they see adverts then research what they're buying
real needle
also that graph is nice but (obviously) doesnt include stats from oculus home, so not really representive
i guess you have a point tho with the enthousiasts thing
prime iris
@real needle how many owners of an enthusiast pc will own a rift but not have steam installed for regular games!? 😄
real needle
also a fair point, but isnt oculus also goin for the non gamers already?
prime iris
i'll gladly accept that it will be the case in some fringe cases, but 90%+ are likely to have steam installed on their VR machine 😄
real needle
fine 😛
prime iris
oculus will probably take back some of the market as consumer desktop VR starts becoming a thing. when it's console priced and everybody has a gpu that can handle the min specs
so starting with the next generation of nvidia gpus probably, and a few years after
real needle
i see
prime iris
because that's when the facebook marketing will start having bigger influence on purchasing decisions
but still not completely clear cut since steam is also a huge name in the game industry
either way, i see valve and oculus still releasing headsets in a decade 😄
real needle
yep true
and honestly i see oculus introducing 'steam support' as some feature, with alot of hype
in maybe a year from now
prime iris
nah. oculus are in it to trap people
they'll support other platforms supporting them
but they'll never support other platforms/headsets 😄
real needle
idk
if what we discussed/you say is true, then they will have to give in at some point
i guess it kinda depends on future headsets, what platform will have more/the coolest
that one will get most attention and games developed for it
prime iris
they don't care about being the most adopted product so much as they want to be percieved by the public as providing the better product (even if they are just average or perhaps even worse than other products)
the apple marketing branch of technology
real needle
hehe
prime iris
thinner, elegant, silky smooth, and perfect
the kind of tech company that markets for the professional but doesn't have a discrete gpu and thermal throttles their i5s
real needle
i was gonna say 'next up, white oculus!' but then i remembered gear vr
anyway ye lets just see how it unfolds ^^
i think (even tho id like it to be different) that oculus is gonna be the big player, just cuz they can already reach so many people trough facebook for practicly no fee
prime iris
yeah, too many variables in a fledgling market to make a real call on what will happen
real needle
and steam will really have to have some nice hmd partners to compete with that, even if the experiences may be better
prime iris
my bet is that desktop vr will be a tough sell
but it's possible that it's benefits will outweigh the mobile options for a considerable time
real needle
dunno, the few ppl ive shown it that have also seen mobile vr all go 'i want your pc'
prime iris
yup. console/desktop vr is going to be the pinacle of the experience for the next 10-20 years
real needle
mobile vr is cool but imo its just limited, mobile devices arent really ready yet, despite the whole marketing bs going on atm
so yea, i agree, pc will have better vr then mobile for the forsee able future
and console will prolly follow suit, psvr isnt even that bad lol
tried it and was pleasantly surprized
prime iris
psvr sucks
because i can't play rhombus of ruin on my vive when it's exclusive to psvr!
real needle
the tech, taking price into concideration, is pretty good
that sucks tho, i agree 😛
also all the games they hyped and then all turned out to be crappy short 'experiences'
mighty carbon
PSVR sold 900k units
prime iris
and double fine sold out :'<
yeah, but i think that all the short experiences are what we need right now
make games on budgets that reflect the number of people who will buy your game/experience/app
not on projected figures that might not happen in the timeframe your studios bank will want them to be turned around in ;P
mighty carbon
900k is actual figure, not a forecast
digital marlin
It's not nothing
And I'd wager creating a great VR game has a better chance to penetrate than something on greenlight (RIP)
Then again I'm a guy on the internet, what do I know
prime iris
@mighty carbon yeah, i know. i meant about taking exclusive paychecks that create a bubble for developers
mighty carbon
it's what necessary at this time
look at the Oculus/Epic line up for GDC
I'd buy Rift only because of those titles
prime iris
people said the same thing about the throw away experiences that got released 6 months ago 😛
i'll be more supportive of making an industry bubble when i see something created that has the substance to carry the medium forward
mighty carbon
are there non-VR games with substance? Give me a break 😛
prime iris
in the meantime i'd be happier that it was a hipster medium that gamedevs and creatives use and aspire to make consumer friendly
mighty carbon
you go to see a movie that is less than 90 min and a 3 hr movie. You pay the same price. You don't complain that short movie as expensive as the longer one. You also don't ask for refunds.
prime iris
no
mighty carbon
So, why should it be any different with VR ?
prime iris
i think movies are terrible value for money
mighty carbon
lol, well, you are the minority
prime iris
it's not even the value thing. i guess it's just that movies are a bad medium for story telling full stop. 90 min isn't enough time to tell a meaningful story start to finish
anyway, i'm just a realist. VR isn't ready for the consumer. Throwing money at it will not make it ready. It will just make it so that some studios expect extra sugar money
mighty carbon
how is it not ready ?
maybe on PC side, it's not (too expensive, still glitchy )
PSVR is very ready
Gear VR has been ready
prime iris
psvr and mobile VR is gimmick VR though
the virtual boy shipped 770k units!
while the psvr sales are promising, i think seated vr has the potential to cause disinterest in VR because it's just an expensive pair of 3d glasses
pearl tangle
have you tried it?
prime iris
i have a vive and have played seated games
and wouldn't buy it if it was just for that kind of experience
as fun as elite and racing games are
pearl tangle
I think there are plenty of genres that work fine for it you dont really need room scale to play the racing games
i personally am the same, i only do room scale games. but im not the majority.
raven halo
": it's not even the value thing. i guess it's just that movies are a bad medium for story telling full stop. 90 min isn't enough time to tell a meaningful story start to finish" <--- wtf
steep hamlet
hey guys, anyone knows how vive controller haptics work? I am calling PlayHapticEffect in both code and BP, and controllers are vibrating, but only thumb pads are vibrating. the wands don't have haptics, where they suppose to have
real needle
@steep hamlet You need to use "Client Play Force Feedback"
It's a function within the PlayerController so it doesn't show up unless you uncheck "context sensitive"
pearl tangle
http://www.roadtovr.com/new-samsung-gear-vr-with-touch-controller-announced-mwc-2017/ you saw that im assuming @mighty carbon
steep hamlet
PlayHapticEffect is a function in PlayerController as well
I wonder the difference though
mighty carbon
@pearl tangle yep. Also, it will work with S6 😃
real needle
Haptic works on the rift
pearl tangle
well not the motion parts of the controller no
pearl tangle
nobody has made the games to support it. everything needs an update to handle motion controllers. it would be like the oculus games magically working with touch
steep hamlet
@real needle so it's a vive specific issue where PlayerHapticEffect doesn't fully work? and all I have to do is use ClientPlayForceFeedback?
mighty carbon
what are you talking about, @pearl tangle ?
pearl tangle
the controller has motion input stuff like daydream, the previous games can't do anything with that
real needle
@steep hamlet Yeah
There's a dynamic one for force feedback aswell, you don't have to make a feedback curve
mighty carbon
@pearl tangle I was saying that I can fully utilize motion controller and new Gear VR with S6, making new game that supports controller. Of course old games won't support it unless devs add support for it/
steep hamlet
what if I do want to use a feedback curve? @real needle
mighty carbon
I don't even care for old games since I don't like any of them :/
real needle
Then use it 😃
steep hamlet
PlayHapticEffect does fully work on oculus right? I can't tell anything wrong with it
real needle
Yeah
steep hamlet
thank you very much!
prime iris
@raven halo 90 minutes to encapsulate an entire story just doesn't do it for me. by the time they're done introducing the characters the film is almost over! i prefer tv series
and don't even get me started on action movies where that 90 minutes has to include 30 minutes of explosions and other pauses in story telling!
raven halo
My jaw is on the floor I don't even know what to say
prime iris
or that fantastic method of story telling technique called "fast cut" so that you can't tell what's going on
raven halo
I think any medium whether it's movies, tv shows, VR, mobile VR is capable of telling compelling stories
mighty carbon
someone please give @prime iris a blue pill and put him back into Matrix 😛
prime iris
that was actually one of the last action films i enjoyed
raven halo
I get the impression you have only watched a handful of blockbuster movies and that's it xD
prime iris
john wick was okay too for an almost ironic story to sew together action scenes
but the vast majority of cinema is pretty bad
actually, i used to work at a movie store
mighty carbon
Game of Thrones is a good one, but it's TV series
prime iris
and if you look at the context of the quote it's not that i don't think that all cinema is bad. just that i dont pay to go to the cinema because i'm disappointed too often to care
mighty carbon
LotR was good too, and had plenty of length to develop characters
prime iris
it was also 3 hours per movie
and 3 movies written in to a single story
lotr was a miniseries that was shown at the cinema 😄
mighty carbon
I've learned to enjoy a lot of mediums (as long as it's a genre I like).. I don't think too deep when it comes to entertainment.
prime iris
i just find myself wishing i wasn't sat in a cinema when i go to the cinema
mighty carbon
lol
mighty carbon
I enjoy almost all of those "dumb" Hollywood blockbusters
raven halo
if you can't stand the dumb hollywood blockbusters, you know, there is more out there in the cinema to go watch
anyway, let's stay ontopic
someone was criticizing VR as not being a valid medium
real needle
i have to say, I have been disappointed by the number of units being sold (rift, vive)
i definitely think this will not be a fad, i just think the graphs and projections will be pushed further out than it felt like back in the devkit days
prime iris
I'm not. I think too many people have them at the moment!
people who should be getting on the VR train at the moment are developers and creatives
silly consumers are buying it and getting disappointed that there's not enough content for them ;P
real needle
chicken/egg problem methinks - not enough numbers to get every dev on board, but enough numbers to expect decent software selection
prime iris
i think letting the numbers grow organically is better than creating a bubble with unsustainable content!
with outside funding at some point there will be too much money invested in VR than can be made from the market
real needle
well, when you put it that way i think i have to agree
i guess what i really feel is i'd like to fast forward 10 years 😉
prime iris
definitely 😄
prime iris
there is more money going in to VR than can be made from it. at some point there will be less/no money going in to VR and it will have to stand on it's own feet. i guess what i'm trying to say is be mindful that that transition will come
raven halo
again, I don't think that is true
95% of publishers are not even looking at VR right now
there just waiting
they are curious, but it's still too risky
if there was a bubble, they would all be handing out checks
prime iris
we'll see! just be careful with your dependencies!
raven halo
you should never put all your eggs in the same basket basically 😛
mighty carbon
what's easier / more practical - make BP to move an actor (not player) along a spline or use Sequencer ?
Sign me up for PSVR + this https://uploadvr.com/vr-aim-ps-vr-secret-weapon-finally-release-date/ and a good FPS (with teleport locomotion)
(assuming tracking is better than PS Move)
prime iris
@mighty carbon go with the method you'd find easiest. there's probably not a wrong way to do it
sequencer is probably the engine idiomatic way of doing it, but then i guess there could be reasons you'd want to control it programatically
mighty carbon
well, I don't know either method. Haven't had a need yet to use either 😃
I guess Ill try spline
I just don't know how to make actor always look down the spline, so to speak (remain tangential to it when moving)
prime iris
splines have functions for retrieving the tangent at a given point (on the spline or near it)
mighty carbon
I use BP only
wise thunder
USplineComponent is exposed to BP? @mighty carbon
mighty carbon
dunno, but I'll check
wise thunder
And I see that the tangent functions seem to be exposed, at least most of them
Yes, I'm telling you it is, sorry haha 😛
mighty carbon
I still wonder if I should make BP system to govern BP actors following spline (tripping triggers, calling events on time, etc.) or if it's possible to make actor follow spline and do the rest in Sequencer
prime iris
yeah, splinecomps are available to bps
if you just want to do lerp along a spline i'd say sequencer is probably better
mighty carbon
the idea is that I can have actor to follow different splines at random, play different anims when required, change speed, etc.
I am sure I can do it in BP, but I don't know anything about Sequencer besides it's a tool to make cinematics
would hate to write something custom if there is already a system ready out of the box
prime iris
i haven't touched sequencer yet, it lets you keyframe actor properties and manipulate their curves like you would in an animation program right?
not sure that you'd have a spline if you used sequencer to create your animation, but in a way you would because you'd access to the animation spline in the form of a curve editor
most 3d tools have the ability to see your animation curves in 3d too as the path of the object you're animating
so if sequencer does that there'd be no real difference at all
the reason to not use sequencer is probably if you want to dynamically change the spline path
mighty carbon
aye, thanks
glossy agate
Sequencer is super easy once you get the hang of it. If you are going to use it for part of the feature, may as well use it for the whole thing.
marble trail
I saw a youtube talk a while ago about how to design interaction in VR with motion controllers. It was by two guys making a surgery simulator in VR. It was Unity based, but lots of the ideas were applicable. Anyone know the video I'm talking about? I can't find it now
Never mind - found it (https://www.youtube.com/watch?v=XYdEO-zIjlw)
marble trail
Are there guidelines on implementing things like drawers, hinged objects etc. with VR motion controllers?
Not sure if I should use physics constraints, manually do some math in blueprint, re-parent objects or what.
real needle
@marble trail Depends of the rest of your interactions. If you have collisions on your controllers for example, physics constraints can be tricky and jittery as you will be able to push for example a door, past it's constraint.
You can find a shared project with examples of drawers and doors on the forum in the VR section.
spiral zephyr
i think this is the thing Deams is referencing, and it seems sweet. Now to find a 4.15 branch. https://forums.unrealengine.com/showthread.php?116782-VR-(OpenVR)-Expansion-Plugin
prime iris
Kind of a reasonable price taking the included headphones in to account
They just need to pull a sub $100 miracle with the tracking pucks 😄
If they can get them out for $60ish then it's a no brainer to have multiple pucks, and the concept will thrive for body/feet/custom peripheral tracking
oh. nevermind. they announced it at $99 too
hmm, might be a little on the expensive time for what the community will want. at that price you're looking at a 1-of or maybe 2-of max
wicked oak
tracking pucks are a bit too expensive
at least normal people will just need 1
for the props
but come on, 100 dollars + the plastic prop wich is also $$$
meanwhile, PSVR has the AIM controller wich will sell for 50 dolalrs (seems so)
prime iris
yeah, but that's the least interesting application!
2 tracking pucks on rubber domed overshoes = footstep prescence!
3rd party glove peripherals need a trackign solution too so pucks were meant to fill that gap
want to know where your chair is in vr game? another $100
prime iris
yeah. but 1k euro :<
wicked oak
wait what
prime iris
it's all falling apart
wicked oak
meanwhile PSVR is selling all the stock in some places
nowhere near the same with pc vr
prime iris
the one that looks like a fingerless sports glove is 1k euro
wicked oak
that + oversaturation of games, im jumping ship
wow, really?
1k euro for the sports gloves one?
jesus christ
prime iris
it has a lot of snsors and haptics in each finger
but yeah, the 1k euro number was thrown around in discussion
tired tree
1k for the dev kit
its not for sale yet
Also I would take 2x tracking pucks on a belt before 2x on feet
tired tree
more stable center of mass
prime iris
that's some shady design going on http://hi5vrglove.com/
several gloves all using the same color scheme 😄
@tired tree yeah, hands > forearms/elbow > torso
but feet have some potential too
walking on the spot locomotion and such
tired tree
Well walking on the spot isn't really feasible
its still jogging really
and waist tracker gets that fine
tired tree
for animation the feet are great, and running step mechanics
prime iris
regardless. $100 is putting me at the slightly disappointed spectrum of the trackign puck announcement pricing :<
prime iris
but i want to track all the things!
tired tree
its bulky and really good accessories are going to use their own sensor array
prime iris
i can't turn my cat in to a pet dragon without more pucks!
prime iris
like that's cheaper! you need to attend a $3k training course still don't oyu? 😄
prime iris
cool, i must have missed out on the fnews
did they put up some online training instead?
tired tree
dunno, but their link to buy the sensors is 404 atm
prime iris
$600 for something that might end up on my shelf unused...
it's seriously tempting 😄
but i need to focus on making games
not catdragons
spiral zephyr
I mean, if someone sells a 500$ catsuit(tracked and preferably with software incl) we would have to very seriously consider building the budget for next few months around that...
but I would use it on the baby 50% of the time
tired tree
it wouldn't cost 500, dev kits are far more expensive than actual product
spiral zephyr
if its cheaper we would get several
prime iris
just to be the worlds first crazy dragon guy
house filled with dragon poop, smell of dragon urine on their clothes
spiral zephyr
we dont need our smell yet in VR, can be temp disabled with superglue
immersion
remeber to trim nosehairs first, really
tired tree
additional trackers are an interesting problem though in the end
people that buy them will want to use them in everything
whether they use them for feet/waist/elbow or external objects
but its another subset of a subset
mighty carbon
this is getting interesting
interest rate is too high IMO.. I could probably get a loan in my bank for lower interest rate
tired tree
they aren't a bank
tired tree
and small loans have high interest
wicked oak
well, suck it up and save that money
tired tree
more so for rent to own
wicked oak
then get gen 2
mighty carbon
regardless, good to see they are offering payment plan. Most people are not going to go to their bank for VR loan 😉
wicked oak
i have a deep hatred for bank loans
you just pay more for everything, while making others rich
mighty carbon
well, if you need something and you don't have deep pockets, you have no other alternatives
clever sky
You take out a bank loan if the bank thinks you can generate money faster than their interest rate.
mighty carbon
if not for my bank, I wouldn't be able to get laptop for my kid for college
clever sky
I don't think bank loans money to consumers looking to buy luxuries.
mighty carbon
they loan for everything
clever sky
Well, as long as you can pay 😃
mighty carbon
indeed 😃
clever sky
Fair enough. Ok, I guess I kinda just tipped my hat on how I manage my finances 😛
mighty carbon
oh, don't go with banks.. go with credit unions
clever sky
Aren't they the worst?
clever sky
ok
mighty carbon
also depends on credit union. same goes for anything really.. Some banks suck, some don't.
clever sky
Which shouldn't be read as 'Valve is working with SMI to bring eye tracking to the next gen of Vive'.
tired tree
oh it would be glorious though
clever sky
Yes, indeed, indeed it would be.
When do you guys estimate we'll get a gen 2 premium PCVR?
mighty carbon
what do you need eye tracking for ? Don't visuals get progressively blurry at the edges of the lenses ?
clever sky
Research seems to be rolling fast... lot's of really compelling tech just becoming real... can see them been in a dilemma
if we wait another 6 months, we'll have this really important feature... but we said that 6 months ago about this other really important feature.
spiral zephyr
hw q4 2018, decent sw using it q2 2019 😉
clever sky
That's a reasonable guess.
But then the cut-off point for feature inclusion is way before that.
spiral zephyr
@mighty carbon You only need high LOD where you looking. Way less to realtime render
clever sky
Because of manufacturing, testing lag.
mighty carbon
@spiral zephyr so you look to the side, where image is blurry because of lenses. And you render it at higher quality. You won't see that quality because of the lenses. What's the point?
spiral zephyr
and of course sweet eye control over ui
we cant be using fresnel lenses for long, curved displays or something with eye tracking im "guessing"
clever sky
@mighty carbon It's most useful when you get higher resolution displays.
mighty carbon
the only use, at this stage, for eye tracking I see is to have avatar showing (to other players in MP projects) where you looking at
clever sky
And also the current premium VR headsets aren't that blurry at the edges - they merely reduce the quality compared to the center.
It's not like it goes from perceptual 1080p to perceptual 480p just because it's at the edge.
spiral zephyr
vive isnt premium anymore, damn
cause its blurry as shit with its lenses
unless you center your line of sight
clever sky
Well, you need it in the sweet spot.
mighty carbon
well, on mobile VR that's exactly how it is - edges are so blurry that can't really have anything there (I wanted UI on the sides - nope, doesn't work)
tired tree
motor that is flat wrong
the peripheral vision is far stronger than you assume, and foveated rendering is very useful even on low res displays
looks at VRworks, which kind of does the same thing to a lesser extent
the more pixels you can render lower res or throw out entirely the better
clever sky
@spiral zephyr @mighty carbon doing some first hand testing - I can confirm that I can read text in steam VR menu (smallest text) with relative ease from the corner of my eye. Image quality is reduced at corner, but not significantly.
So foveated rendering would definetly help. Especially at potentially higher gen 2 resolutions.
tired tree
foveated rendering gets you that 300% supersampling that everyone wants, pretty much for free
clever sky
That too.
spiral zephyr
problem with vive at least is lens artifacts, no matter the res it will be pretty bad at edges
tired tree
hell VRWorks manages you get you to 300% at around the cost of 200% without it
clever sky
Who cares? The point is it still benefits from foveated rendering.
@tired tree I'm hard pressed to notice significant quality improvements above 150% TBH... 300% seems quite overkill.
tired tree
there isn't much
but it is noticable
eye tracking also allows you to do better focusing / non infinity focusing
clever sky
Even above 130% I find that improvements diminishes quickly.
But yeah, eye tracking would be sweet.
And not just for foveated rendering - for eye track UI, for fake focus, for social VR gestures, for AI response.
tired tree
Zap it isn't the texture quality that improves with more supersampling, its the geometry edges
you don't need AA as much
obviously resolution limits how fine an edge can be
clever sky
@tired tree Yeah, I'm looking at the edges. But my point of reference is FXAA with deferred.
tired tree
TAA benefits a lot more, that is true
clever sky
Personally, can't stand TXAA myself 😛
it just smooshes fine details at lower resolutions.
spiral zephyr
i thought FR and msaa was the way to go, are you talking about something else?
tired tree
MSAA is the way to go for solid geometry
clever sky
Kills much of the benefits of FR.
clever sky
You might as well go TXAA and up the res.
tired tree
and the visual fidelity for non foliage scenes is worth it
clever sky
That way you get better AA across textures too.
spiral zephyr
I'm not doing geometry heavy stuff anyway, small spaces etc. also on 1080. msaa for now on projects, but realize it wont necvessarily stand if we go for a release
tired tree
MSAA is intended to be their goto VR AA
thats the whole point they implemented it for Robo Recall
its not suited to everything
but its still the overall best
clever sky
Not from the way I hear them tell it.
It's more of a - we implemented it because... there are some advantages in some use cases, and Robo Recall is one of them.
tired tree
the stuff currently the most made in VR
tired tree
Even have GodCarmacks word itself saying that MSAA is the goto AA solution
clever sky
@mighty carbon So what's your thoughts on rotationally tracked controllers now? Wasn't here to see your reaction with the Gear VR announcement 😛
spiral zephyr
wait what
tired tree
they aren't 3axis tracked
clever sky
@tired tree oops. got words mixed up
tired tree
they aren't positional-ally depth tracked afaik
clever sky
Nope. Just rotationally like the daydream controllers
tired tree
yeah, can't beat the drift....
mighty carbon
@clever sky It's 3 DoF controller. Pretty hyped as it works with S6 too. Hopefully UE4 support will be coming soon.
clever sky
You planning on making it compatible with your game?
If nothing else, would be much more fun for yourself to play 😛
mighty carbon
I am working on a semi-interactive cinematic experience, so it doesn't really need motion controller.
spiral zephyr
put a fisheye low latency webcam on the back of the phone and track their depth shittily™
clever sky
And for me... means I'm potentially willing to build some gear VR content.
mighty carbon
I might fully utilize it in another project I will be working on (a game)
it would be nice if UE4 use same components it uses for Rift/Vive for Gear VR controller (accounting for limitations of the hardware of course)
this way porting it to Rift/Vive would be a breeze
mighty carbon
Daydream controller has a different components in UE4, afaik
clever sky
Daydream stuff is google stuff.
mighty carbon
I am not 100% sure.. For all I know people have had hard time getting it to work
clever sky
in the sense you need that google plugin to build for daydream 😛
mighty carbon
Zeni's injunction overshadows MWC and GDC for me 😦
but judging how Oculus moves forward on various fronts, it seems like they are not bothered
raven halo
you shouldn't either.
Samsung and apple are suing one another all the time
wouldn't be the first time either that one requests the other to halt production of their phones
it will never happen
prime iris
Ooo, theres already eyetracking vive prototypes in the wild?
foveated rendering confirmed for next generation?
mighty carbon
wicked oak
well, not anymore
you could call Multires/LMS a kind of foveated rendering
it renders the center at better res than the sides
they can be modified to render another part instead of the straight center
prime iris
and being able to do 4k-8k per eye on current hardware would be worth it!
wicked oak
he said that becouse Nvidia GPUS with stuff like multires or LMS wich have hardware support werent ready
but now it is
prime iris
even if there was a bit of latency
prime iris
how long until gdc opens up?
wicked oak
with relatively little changes, you could even implement the moveable multires in ue4 probably
prime iris
i want to hear what the VR related people have on display 😄
prime iris
well this floorplan image has me intrigued what might be there
that gallery ballroom beneath hall B has 3 exhibitor id's but only valves 104 is on the list of exhibitors
so there's a chance they've booked an entire sealed off from everybody area to themselves
prime iris
they're the only exhibitor on that list that has anything in the gateway ballroom!
so there's a chance they're not just 104, but 102-104
tired tree
Pretty sure Carmack was also primarily talking about mobile hardware for Foveated in that quote, that being said Foveated rendering on mobile hardware is the huge leap forward that it needs to get the weaker hardware providing desktop level graphics.
prime iris
yeah, carmack is right about mobile vr being the end goal. I just think it's going to take a while
desktop and high end consoles will be the only platforms with the chops to do the ideal 4k/8k resolutions in the near future
even with foveated tech
raven halo
well, supposedly snapdragon 835 is the key to foveated rendering
we will know soon enough
mighty carbon
March 29th
wicked oak
foveated rendering is not a panacea
it just covers the pixel crunching part
CPU power and drawcalls and the amount of vertices rendered still exist
but with foveated you can go ultra overkill on the screen reslution
prime iris
which is why i want it!
i suspect headsets will have a release cycle that's synchronized with gpu release cycles
just because of their dependency on gpus to raise the bar. so i suspect that with foveated rendering tech we'll definitely have 4k per eye in the next gen, and maybe even have an 8k option
wicked oak
Nvidia is on the best side right now
with their hardware acceleration for this stuff
from MultiRes to foveated rendering is a very small step
instead of having the center at more res, have it variable for the location of that high res window
they also have actual foveated rendering research
they have a great demo around
they found that you had to apply sharpening filters
and "oversharpen" it
sharp swan
is a real nvidia fanboy
wicked oak
if you just did a blur, the eye notices
sharp swan
i dont normally fanboy for companies, but Nvidia are the shit.
wicked oak
i dont famboy at all, i give no craps. I just buy the one with the features i need
prime iris
i dont really care if it's red or it's green, but nvidia have been doing a lot more developer friendly technology pushing things
wicked oak
sadly, AMD gpus dont have a cool VR implementation in ue4
prime iris
amd just got "it's fast and it's cheap, look at me!"
wicked oak
AMd just follows the MOAR CORES phylosofy
thats why AMD gets such a massive boost with DX12 and Vulkan
prime iris
we're so pro-developer, our sources are open, aka. go do it yourselves we can't be arsed
wicked oak
amd is real bad writing drivers
but on DX12 and vulkan, its better due to much more slim drivers
sharp swan
yeah they are certainly making efforts
wicked oak
interesting enough, have you seen how amd cards get more power over the years?
a Nvidia card releases, and its similar performance for all its life
maybe a bit better
you can see amd cards getting 20% more power than they had at release
becouse they manage to improve the drivers XD
prime iris
yeah, they deserve credit for mantle/vulkan. i'll probably never write anything in it but it sounds like the right direction things need to go
sharp swan
I hope AMD do come from behind and introduce real competative development. I have no issue with that as it benefits everyone
wicked oak
they compete quite well
a big issue with amd right now is that they get lower profit
Nvidia has a "more optimized" chip
for gaming
their chips are smaller
and more energy efficient
AMD makes parallel machines
prime iris
so they can fit more in the salsa?
wicked oak
with a fuckton of cores
thats why they are normally more power hungry
the problem is that the chip itself is bigger
so its actually more expensive to make
nvidia gets MAD profit from each sale
and then they have the AI market, where they sell the Tesla accelerators
those ARE "moar cores"
just a fuckton of cuda cores
and then they sell those at 10k dollars per card
anyway, for VR im on Nvidia 100%
for now
Multires, LMS, singlepassStereo, holy shit
if you use the nvidia branch of ue4 and enable the stuff, you get a lot extra power
ive been doing that when i demo my game at trade shows
i use a laptop with a 1070, wich is already quite powerful
but then i go and build the "demo" version with the nvidia multires branch if i can
and enable singlepass stereo and LMS or whatever
and go from 140 screen percentage to like 180 screen percentage
on a laptop
without any drop in fps
prime iris
how nice does the nvidia branch play with amd devices?
prime iris
i meant more are they configurable extensions that can be turned on/off as needed or do they replace some systems in the process meaning you'd have to keep an unpatched version for intel/amd
prime iris
neat
wicked oak
enabled and disabled at runtime as you wish, but shaders got to be compiled
tired tree
don't know about that
they had a lot of AMD compatibility issues
they are made to be toggled on/off but their hardware testing isn't that stringent with amd
when 4.13 came out, they ran a NVIDIA only API call regardless of card that crashed AMD cards right out before screen init
thats something that any prelim test on AMD at all would have found
I wouldn't use it for production builds without multiple AMD cards and rigs to run tests on
thats with ALL VRWorks features off btw, also when I fixed that had shadow ghosting and duplicated objects between eyes still because they were still running some of the VRWorks shader code on AMD since it wasn't checking on the amd side for if the variables were enabled or not.
tribal citrus
Anyone here with any knowledge for building a packaged build for VR? Mine's keep failing
prime iris
@tribal citrus I probably can't help, but are you getting any error message inparticular?
tired tree
yeah post log errors, there are a ton of reasons it could be failing
sharp swan
im guessing a plugin/module that isn't packaged. /waits for the answer
tribal citrus
Everything or just the errors/warnings?
wicked oak
Nvidia just pushed some Vulkan extensions for multires stuff
tired tree
UATHelper: Packaging (Windows (64-bit)): UnrealBuildTool: ERROR: No 32-bit compiler toolchain found in C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\bin\cl.exe
wat
wicked oak
the what
tired tree
well thats the error blocking you
prime iris
where on the doll did you touch visual studio 2015?
tired tree
oh
you forgot to install C++ package when installing visual studio
thats one of the required steps for community
I didn't know it reported like that though
tired tree
"Before Setting-up your UE4-to-VS Workflow
It's important to note that Visual Studio 2015 (VS2015) doesn't install C++ tooling support by default. When installing VS2015, choose Custom installation and then choose the C++ components that you'll need for your workflow. If you've already installed Visual Studio 2015, choose File > New > Project > C++ to install C++.
The following table lists which versions of Visual Studio are integrated with the binary version of UE4."
mighty carbon
raven halo
@wicked oak nvidia pushed vulkan extension? whaat? 😮
mighty carbon
I wonder if that extension is Nvidia-only 😃
(sounds like OpenGL story all over again)
tired tree
Kronos just announced openxr as well
though there is no real details
about expected features
prime iris
where are you folks keeping up to date with gdc news?
mighty carbon
what I don't like about AMD is they naive approach to game dev - they would rather adhere to standard and lack feature and have crappy drivers. Nvidia on the other hand doesn't care much for standards and provides awesome features and stable drivers.
prime iris
mantle kind of makes that not entirely true
sharp swan
didn't mantle die?
sharp swan
so the mantle point is irrelevent :p
prime iris
but they could do more!
sharp swan
they could, thats what annoys me. They choose not to
they could make a crapload more money if they supported cuda stuff
prime iris
not really. i mean mantle was worked on by many other parties, but it was the blueprint for how it would work to this day
sharp swan
and directcompute versions of gameworks are a start but not enough
its all lagging behind
prime iris
still no physx-like lib either holding back nvidia
devs wont go full physx support because they get a backlash for not writing their own version for amd
but still! dont' make amd out to be complete grinches! they do do some cool stuff! 😄
sharp swan
they are starting too and mantle was exciting when it was announced, but just like a 20 dollar hooker, it's always exciting until it's done then it's disappointing.
prime iris
well the thing with graphics api's is you can't really strong arm adoption when you're only 25% of the marketshare
wicked oak
its the same as with opengl extensions
nvidia is really REAAAAAALLY active with them
they have super cool extensions for lots of things
amd doesnt
when both amd and nvidia do a extension that does the same, that extension merges as a "official" extension
then it becomes part of the official spec
sharp swan
if I was nvidia, I would undercut AMD until they dropped out of the market. Then I would continue my domination of the 3d world from my Babylonian tower wearing a 1870's ruffian hat !
wicked oak
and get sued
they cant let that happen
why do you think intel hasnt tried to oblierate AMD?
if they were a real monopoly, they would be targeted with massive anti monopoly lawsuits
sharp swan
monopoly laws?
wicked oak
becouse governments themselves need the competition to happen so they can get the hardware
cpus are ridiculously important worldwide
at the very least
sharp swan
meh capitalism is so hypocritical at times. :p
tired tree
Intel was forced to pay AMD to keep them alive back in the day
wicked oak
governments will invalidate that patent stuff
so 3rd parties can get into desktop market
its better to just be better than AMD and charge 4 times as much
sharp swan
Maybe we should start a campaign to oust AMD then. Refuse to make things work well on their hardware. :p
tribal citrus
@tired tree @prime iris Installed the C++ package and it came ALOT futher but still failed ( probably kept on building for about 10 min before failing ) The entire pastebin file is to large to upload via a free account for here's the warnings / errors, I might be able to give you the beginning and end for the normal output if that might help. Thank you for helping http://pastebin.com/kaE8kGRS
sharp swan
yeah thats broken :p
mighty carbon
4.16 will have Gear VR controller support. Also coming to github early March.
tribal citrus
@tired tree @prime iris Here's the beginning & end log aswell http://pastebin.com/BZus7ce8
tribal citrus
14.3 , I'll double check
tired tree
UATHelper: Packaging (Windows (64-bit)): UnrealBuildTool: ERROR: No 32-bit compiler toolchain found in C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\bin\cl.exe
UATHelper: Packaging (Windows (64-bit)): CommandUtils.Run: Run: Took 3.3699607s to run UnrealBuildTool.exe, ExitCode=5
UATHelper: Packaging (Windows (64-bit)): Program.Main: ERROR: AutomationTool terminated with exception:
still getting that
did you regenerate your project files after re-installing?
its specific to the fmod oculus plugin though, I don't know the exact installation steps for that, last I heard it had binaries that needed copying. Think Motorsep has it working in his project
prime iris
yeah, i'm happy with htc's contribution and will continue to buy their headsets if they keep up with other devices in recognition of that
raven halo
holy crap
prime iris
but daaaaamn, compeition is gunna be good!
raven halo
is this GEN 2?!??!?!
prime iris
unless foveated rendering is announced along with it, probably not gen 2
tired tree
could be a price drop though
prime iris
if they make their screens higher res they'll be more expensive and more niche than desktop vr already is
yeah
tired tree
if they want to play their cards well
prime iris
i definitely get the sense that htc are quite comfortable with the lack of price competition at the moment 😄
mighty carbon
Rift + Touch is $2 cheaper 😛
prime iris
oculus weren't competition until they started shipping touch ;P
tired tree
but a third party will ruin that
rift + touch is 2$ cheaper assuming you don't buy a third camera
which you kind of need to get equivalent performance.
prime iris
especially LG. they're pretty healthy businesswise right? a pretty big name in display tech
mighty carbon
yeah
prime iris
how deep in bed will samsung be with oculus i wonder
noble crater
anyone know if the bug with vive left/right controller detection was fixed in 4.15? I remember seeing it got merged but not sure it happened in time
prime iris
i wonder if they'll start doing lighthouse/daydream in the same way they support android
@noble crater you mean the trackpad being glitchy?
noble crater
no - openvr has a feature to autodetect which controller is left hand vs righthand
and ue4 wasnt using it
prime iris
oh, cool
mighty carbon
I hope Oculus drops price a bit, when LG releases their HMD and it turns out to be cheaper than Rift
prime iris
i was projecting my hopes for my buggy trackpad being fixed in an update 😄
@mighty carbon oculus needs to do more than drop their price. they need to drop constellation
mighty carbon
1.12 beta provides almost flawless tracking
mighty carbon
it's all solvable in software (constellation tracking that is)
tired tree
its still kind of a terrible solution compared to lighthouse though
mighty carbon
Lighthouse is not a scalable solution
prime iris
oculus's idea of roomscale is that you take a room, you scale down the space to 0.9 of it's dimensions!
tired tree
yes it is....
mighty carbon
Engineer who worked on Lighthouse wrote about it
tired tree
alan yates?
mighty carbon
yeah
tired tree
yeah he wrote the first gen doesn't scale well, thats why single pillar lighthouses are coming out
mighty carbon
we shall see.. I am not happy with Constellation requiring a ton of USB ports and cables, but Lighthouse is not a perfect solution either
tired tree
that same Alan Yates also wrote a bunch of heavy handed text about how constellation is terrible
mighty carbon
we'll see what future holds
tired tree
he is a little over the top
prime iris
there are also ways to make it scalable that aren't being applied right now because first gen tech. using different frequencies of light would be one way to have mulitple lighthouses without making the tracking math overly difficult
tired tree
the new lighthouses will sync better in quantities over 2
prime iris
mordentral confirmed mole
tired tree
they confirmed it already
its not just the power draw and vibration reduction that they switched over
prime iris
valve has never confirmed anything relating to numbers greater than 2
time to get some icecream
wicked oak
single rotor lighthouses also become smaller, more FOV, and should be cheaper to make
1 motor instead of 2
with added reliability
tired tree
yeah
its all +++
don't think three lighthouses or more can come to first gen headsets though
its second gen that is supposed to be able to handle more with the new design
mighty carbon
that being said, we don't know what Oculus has in store for Gen 2
could be inside out tracking
prime iris
nah, that would be a productive solution. they'll have nicer marketing
"Here at oculus tech we understand what is important to VR. That's why we're releasing a designer satin headstrap making Oculus once, again the most comfortable headset on the market! drops mic triumphantly"
mighty carbon
Apple does this already.. 😛
One of the points of getting Oculus is that you can release in Oculus store and on Steam. If you only have Vive, you are bound to Steam only.
mighty carbon
well, if you don't value your business (or you are into enterprise where you don't need to support multiple platforms to maximize your target audience), then be silly like that
prime iris
I guess you're right, you should design your software for the lowest common denominator
raven halo
I can't wait to see LGs HMD, I wonder what kind of screen they will have. LG screens are awesome!
mighty carbon
I don't see how Rift is the lowest common denominator. PSVR or Gear VR maybe are.
wicked oak
psvr is the common denominator
mighty carbon
bizarre - no one seems to know at #animation and lounge and on the forums
are there any good animation tutorials floating around youtube? (working with anims in UE4 that is, using IK, blending, etc.) I am skimming through official docs and it's a massive system which requires more actual production examples I guess.
prime iris
i kinda learned the little i know piecemeal
there was an official video that covered animbp's pretty well. i think it was a 3rd person template tutorial
don't think it covered IK but that's relatively simple to do if you're familiar with IK from 3d packages already, just a bp node that solves for you
mighty carbon
basically I have an actor moving on a spline and I need to play main animation for it, and randomly blend it into a secondary animation, then blend back into main animation
prime iris
wicked oak
its also 8 cores vs 4
prime iris
i thought they already announced it at $400-500?
mighty carbon
I doubt I will need more than 4 cores / 8 threads
mighty carbon
newegg
tired tree
for what vs what
mighty carbon
$314 Kaby Lake vs $399 cheapest Ryzen
tired tree
lol
mighty carbon
4Ghz vs 3.4Ghz
mighty carbon
tired tree
and that is 4 ghtz turbo btw
prime iris
did anywhere do realworld benchmarks yet or are they still going off the launch event numbers?
tired tree
a few, its on lockdown till 2nd
so no official official
all leaks so far looking good though
fairly competitive on single thread, blows everything out of the water on highly multi threaded operations, thats to be expected though
mighty carbon
this one is still cheaper than Ryzen
I don't like 91W though
I am aiming for this (eventually) https://www.newegg.com/Product/Product.aspx?Item=N82E16819117727
wicked oak
do not get the nonK version for that thing
that is a massively overclockable chip[
get the overclockable version
its basically a intel 6000 but a bit overclocked by default
mighty carbon
I don't overclock, ever
wicked oak
and the 6000 is a slightly better 5000
you can get a 5000, overclock it, and get more than the 7700 performance
easily
tired tree
if you want to compare quad cores than wait for the Ryzen quad cores, why are you price checking different tiers of CPU's?
you are literally saying a 4 core is barely cheaper than an 8 core
mighty carbon
to go with Ryzen I'd have to change mobo and I am not going to
wicked oak
and not with the intel?