Jump to content
The Dark Mod Forums
Sign in to follow this  
woah

Thoughts on VR 4 years later

Recommended Posts

So it's been about 4 years since VR's consumer launch and with Valve finally showing 4 HL:Alyx gameplay videos today (one, two, three, and four), I figured it would be a good time to share my current thoughts about the state of VR.

I like VR quite a bit myself (it's largely the only thing I play) but I also think it's an immature technology.  The hardware, software, and mechanics are very early.  Just anecdotally, the userbase appears to be separated into two groups: a core group of enthusiasts that are pretty regular, and a larger more casual and high turnover (low retention) group whose usage follows exponential decay.  Overall it reminds me of the early phase of other major technologies/new mediums (e.g. smartphones) but with abnormal degree of hardware subsidization (abnormal for this stage of the tech) and an "improper" amount of hype.  VR is also abnormal in the sense that due to the intense spectacle it typically has an extended (between a month and a year) "honeymoon phase", but this phase doesn't seem to be representative of a stable/regular user over the long term.  


Benefits of gen 1 VR - These are the good aspects of the current technology, but basically all of them have significant downsides as well (noted below).

  • Immersion - This is the obvious one and what most (unfortunately) only focus on--I mean the pure spectacle of VR.  It simply gives you a better sense of being "in" the virtual world and amplifies the intensity of many aspects of the experience.  For example, combat intensity (Onward), horror (Walking Dead: S&S), heights and speed (Windlands 2 and Jet Island), social connection (VRChat or really anything else with multiplayer), etc etc.  Each hardware iteration improves this (e.g. Vive/Rift -> Index is very nice) and playing a flat game after using VR for a while is kind of like "watching yourself" play a game in that it feels "disconnected".

    As an example, I have a friend that lives approximately two hours away and I would typically visit him in person several times a year.  However a couple of months ago we both realized that we hadn't actually seen each other in person in almost 4 years, and yet it "felt like" we'd been seeing each other all along.  I can only attribute that to the social connection afforded by VR--it's not like we weren't playing games together before that.
     
  • Perceptual enhancements - Similar to immersion but with respect to things that allow one to engage with the virtual world in useful or consequential ways specific to VR.  For example, a head-mapped visual perspective allowing one to look around independently, the ability to better ascertain the depth of things with stereoscopic vision, a correct perspective from which hand interactions are viable/intuitive, proper head-relative 3d audio, and so on.  E.g. I find it much more rewarding and natural to communicate with nearby teammates in VR FPSs because the experience is as if they're right there in the same room as me.
     
  • Interaction and new mechanics - 6 DOF inputs for both your head and hands allows one to interact with the virtual world in some interesting ways.  Over time I've personally come to find that this is the most interesting aspect of VR and appreciate today's (rather limited) VR visuals more for how they provide the correct perspective for this kind of interaction rather than for the pure spectacle.  A basic example of this is the well tuned firearm interaction model of Onward which can be extremely rewarding to use and changes up the dynamics of first person shooters dramatically.

    More complex examples would be the physics based melee combat, climbing, and general interaction models of Blade & Sorcery and Boneworks.  Rather than relying on simple QTEs and a small set of mapped inputs that play out largely the same way every time they are triggered, VR opens up the possibility of accomodating an enormous number of other interaction possibilities--if the simulation allows for them anyway.  Over time I've felt like flat gaming was being limited by the interface through with you engage with the virtual world and VR appears to be a way of overcoming that.


Issues with gen 1 VR - It might seem odd but the attributes of VR that I find the most compelling also happen to have aspects about them that I think are the most problematic.  I think the biggest issues are in the areas of comfort, "perceptual limitations", and clunky interactions.  

  • Discomforts
    Ergonomic Discomforts - The headsets are hot, heavy, strapped tightly to your face, and tethered.

    Visual Discomforts - You can't change focus so there's basically only one depth plane that is perceived correctly at about 2 meters away from the user (closer objects appear out of focus and "medium distance" objects look decent but still "off"). And there are many issues with the visuals that people are sensitive to e.g. pupil swim, distortion, god rays, glares, chromatic aberration, etc etc.

    Physical Discomforts - Too many games require you to stand which is ultimately a losing proposition during the critical end-of-day gaming timeslot.  How many gamers are actually going to stand up to play games after work/school?  Not many I think.  I personally suspect that the default mode of play will eventually settle on a seated mode with smooth vertical translation on the dominant hand's vertical joystick axis, accomodated by a seat that swivels.

    Simulator sickness - Simulator sickness is a problem and may always be a problem.  I know most can overcome it with careful exposure (and thus getting your so called "VR Legs") but getting more casual users to that point is difficult.  And yet not building up one's VR legs will leave them very limited in how they can experience VR (so limited that it understandably may not even be worthwhile).  I think there's a chance that this problem will always constrain VR to the more "hardcore" end of gaming (a noteable exception being for a more niche group that uses VR for "active gaming" / exercise)
     
  • Perceptual limitations - Pixel density, SDE, FOV, clarity, the fixed focus, poor black levels, etc etc impose frustrating limits on how you're able to interact with the environment.  E.g. devs tend to avoid near field interactions like reading something in your hand because you can't actually change focus to that depth right now.
     
  • Clunky interactions
    In Games - In many games, the input is at least as clunky as it is compelling because current controllers are lacking in feedback, i.e. think trying to navigate through everyday life with unfeeling hands.  We take for granted how complex and informative the senses of feedback through the hands are and how critical they are to even the most basic of interactions.  Object reorientation, grabbing objects, throwing objects, swinging melee objects, opening doors, etc etc are all quite hindered right now.  Proprioception hardly compensates and visual feedback must be largely independent of this for viable input mechanisms that aren't frustrating (imagine trying to carefully watch your hands perform even the most inconsequential interaction).

    Another problem is with the software itself and how well their interaction systems are being implemented.  The same general interaction concepts are being implemented in a multitude of different ways, with some being rather gimmicky/pointless (i.e. you may as well just press a button to execute them) while others truly offer a degree of depth that warrants the use of motion controllers.

    In Desktop interaction - Outside of and in between games you really feel crippled.  There's no good way (yet) for key input and we need integrated eyetracking to move past this silly laser pointer/Minority Report UI phase.
     

Right now all of these fatigues and frustrations lead to many people not feeling particularly motivated to use the headsets very much after the honeymoon phase.  Until the hardware better replicates human vision and until it is somewhat comparable to the clarity and comfort of a desktop monitor, I think VR will continue to have major retention issues.  And yes, to me it's currently worth it--especially with content that does something interesting with VR beyond spectacle, but I'm just a nutty enthusiast.

And, sure, when it comes to what's missing from current gen VR, "content is king" and "cost" are the reflexive platitudes you hear from most people.  However, at this point VR already has a good deal of engaging content (especially given the age of the medium) and it's also about as cheap as it will get for a decent experience and for the foreseeable future (what most people don't realize is that the Facebook, WMR, and Sony headsets are subsidized), but retention remains to be an issue.  The problem I see is that even an extremely polished and grand experience (like HLA)  is difficult to keep coming back to month after month if what you're experiencing it through is fatiguing/uncomfortable and frustrating, especially when it's competing with the wildly successful medium of flat gaming that has been refined over many decades and that has none of VR's comfort problems.  There is a very high threshold that must be overcome here (there's a big difference between "tolerable" and "I actually want to use this every evening past the honeymoon phase") and getting to that point will take some incredible technology.  I don't think it will be any different for Half-Life: Alyx.

 

Future advancements - I'm trying to be really candid about all of the issues VR has but, to be clear, I'm really interested in the technology and it's what I invest much of my free time playing and hacking around with.  I could see the following things dramatically improving the experience and gradually (with each incremental improvement) making the technology something more and more PC gamers will actually use regularly.

  • Variable focus - This is the feature that I think everyone wants without actually realizing it (though not quite the same, I think if you disabled one's ability to change focus in real life the value of this would immediately become apparent).  It will dramatically improve comfort, interactivity for the near field, and immersion.  Facebook detailed their approach at the last Oculus conference and others have patents for similar solutions.  Unfortunately I think we'll be very lucky if we see this in the next major headset refresh around 2024 or so.
     
  • General visual improvements and foveated rendering - This is with respect to things like pixel density, screen door effect, field of view, lens quality, pixel persistence, refresh rate, etc etc.  Having used the Valve Index for a while, I actually think we're getting pretty close to being "good enough" here.  With another decent incremental improvement beyond today's specs (for which eyetracking and foveated rendering will be required), I think the primary concerns most users have with VR technology will be focused elsewhere.
     
  • Smaller, lighter and wireless headsets - The use of things like pancake lenses will allow headsets to become much smaller and lighter (there are already some examples shown at this year's CES).  60GHz 802.11ay and foveated compression will allow us to transmit the data wirelessly and with negligible latency.  This should dramatically improve comfort and make the technology less distracting to the experience.  The less conscious you are of the thing strapped to your head, the better.
     
  • Interaction (hardware) - Haptic feedback is a major area that I really hope the industry leaders (Valve, Facebook, Sony) are focusing on internally.  The single dimension of vibration in each controller is insufficient for the reasons noted above and I think the lack of feedback is the major source of motion controller clunkiness.  Right now I think you need something that gives users a sense of positional and rotational forces relative to at least a single point within the motion controller (e.g. something like this https://www.roadtovr.com/miraisens-3dhaptics-directional-haptic-feedback/ ).  This would give the user a vision-independent spatial sense of how their hand is interacting with the environment, how wielded objects are behaving, and how to solve "violations" of the simulation (e.g. when one's virtual hand intersects with a solid object).  Like any other input/feedback system used in gaming, I don't think the feedback manifested in one's hand needs to be 1-to-1 with the simulation--the brain just needs the information in an intuitive form over a sufficient number of dimensions.

    However, I don't think things like full hand simulation / finger level simulation (e.g. per-finger force-feedback and touch sensation) will be viable for gaming until such an input device can adequetly simulate joysticks, triggers, and buttons itself--and something like that is so far in the future that I don't think it's even worth thinking about in the consumer space at this point.  I've seen the early implementations of such a thing but I don't think they're anywhere near being viable.
     
  • Interaction (software) - When the hardware interface through which one interacts with a virtual world gives the user a higher capacity for control, the software simulation must also expand in tandem to respond to that capacity (actually, not doing so can make a VR experience feel rather lifeless).  Over the past few years developers have been fleshing out VR interaction systems across a wide variety of contexts to show how motion controllers can be used (with each game typically focusing on one or two core mechanics e.g. firearm simulation, flying, swimming, climbing, swinging, etc etc).  More recently you see developers like SLZ (Boneworks) and WarpFrog (Blade & Sorcery) trying to generalize the interaction systems through the physics system so all aspects of the simulation behave consistently and to allow for a much larger space of potential interactions.  As these mechanics improve (and there is a lot of room for improvement), I can really see VR interaction being something that is highly coveted.

 

Anyways, I generally write things like this up just to help organize my own thoughts but hopefully it's of use to someone else.  Right now I'm enjoying what VR has to offer and I'm optimistic about future tech, but I'm also a bit nervous about hardware and software developers sticking this out for another 5 years or so due to the expectations they came in with.  There are only about 1.3 million VR headsets connected for the Steam Hardware Survey, and I suspect many more have been purchased but just aren't being used.  I think that back in 2016 the expectations about the speed of progress were completely unrealistic.  When you look at the consumer launch of other mediums and how long it takes for them to get established, 10+ years isn't that unusual.  The impression I got from the VR marketing and tech journalists was that this was being treated like the launch of a new console but this was obviously something quite different.

Edited by woah

Share this post


Link to post
Share on other sites

The key to the mass market sadly is the price. VR becomes mainstream, when consoles come with a headset by default and the majority of console games supports it properly.

Share this post


Link to post
Share on other sites

Well... for me it's the key as well, because, I wouldn't pay half the price I paid for my PC for such a device.

I think they will become affordable in the near future. Another problem is the system requirements though. You need a powerful PC to project to 2 HD or more screens.

Also, the early headsets suffered from some childhood diseases, like bad resolution, bad acuity, bad tracking.

Share this post


Link to post
Share on other sites
4 hours ago, Abusimplea said:

The key to the mass market sadly is the price. VR becomes mainstream, when consoles come with a headset by default and the majority of console games supports it properly.

Not too long ago an all-in-one VR console (processing, battery, headset, and controllers) could be had for $400, Facebook's Oculus Quest (which can also connect to a PC).  On the PC end, the Samsung Odyssey was going for $250 to $300 a few months ago.  There are lower end WMR headsets that go for $150.  Right now all prices are inflated due to Corona virus related supply issues.

The problem I see is that even if they literally gave the best VR systems on the market away for free very few people would use them much past the honeymoon phase for the reasons I noted above, i.e. "free isn't cheap enough".  The hardware still has a ton of issues and addressing them is going to take a lot of costly investment and time.

Now, I do think you're correct in that mass market appeal requires low prices and a steady stream of high budget content, but I don't think that's the bottleneck at this point and there are viable markets that can be tapped prior to mass market appeal.  Most new technologies and mediums are initially very expensive and only gradually come to mass market as they are refined in the high end (both software and hardware).  There are plenty of people that will pay a lot of money for a good experience but I don't think VR is offering that for most people.

A few years ago Gabe Newell said something like "you make something good first, and only then worry about major cost reduction" and I think that makes a lot of sense.  When they see that VR users only play for short periods because the headset makes them uncomfortable, that they do not play frequently because they are too physically exhausted at the end of the day, that long term there are retention issues, etc etc, that must tell them that the quality of the experience isn't there yet.

Edited by woah

Share this post


Link to post
Share on other sites
7 hours ago, chakkman said:

I think they will become affordable in the near future. Another problem is the system requirements though. You need a powerful PC to project to 2 HD or more screens.

Also, the early headsets suffered from some childhood diseases, like bad resolution, bad acuity, bad tracking.

The powerful PC thing is why i expect the mainstreaming of VR to come from the console side. Powerful PCs aren't something, the mainstream wants to need to have just to be able to enjoy some VR dating sim or game.

The casual gamer seems to be fine with paying a premium for games but the hardware has to be cheap to the point of being crappy because of its cheapness. We also will see ingame advertising in "freemium" and non-"freemium" VR, because that is just, how the advertising market cancer works.

6 hours ago, woah said:

The problem I see is that even if they literally gave the best VR systems on the market away for free very few people would use them much past the honeymoon phase for the reasons I noted above, i.e. "free isn't cheap enough".  The hardware still has a ton of issues and addressing them is going to take a lot of costly investment and time.

I think it is more about game support and low entry barrier than anything else. People even play shooters on their phone because theat is the hardware they got. And the ones who get a console will probably use the VR headset if it came with one and the games support it.

6 hours ago, woah said:

A few years ago Gabe Newell said something like "you make something good first, and only then worry about major cost reduction" and I think that makes a lot of sense.  When they see that VR users only play for short periods because the headset makes them uncomfortable, that they do not play frequently because they are too physically exhausted at the end of the day, that long term there are retention issues, etc etc, that must tell them that the quality of the experience isn't there yet.

Current headsets are surely bearable for an hour or two - wich would already be enough for the mainstream.
The seated experience is must though. If it can't be played while sitting on the couch, they just won't play it.

Another thought about that honeymoon phase:
If it already is a month for a gamer, i would expect it to last way longer for the casuals. They spend way less time per month with games. So they probably have way more time to tell all their friends about the awesomeness of VR until the excitement wears off. ;)

Share this post


Link to post
Share on other sites
Quote

I think it is more about game support and low entry barrier than anything else. People even play shooters on their phone because theat is the hardware they got. And the ones who get a console will probably use the VR headset if it came with one and the games support it.

That's certainly not what I am personally seeing among the VR users I've interacted with and it's not what I'm reading from those familiar with the actual usage numbers.  E.g. Palmer himself (the vr poster boy) wrote an article on this very issue where he points to large scale real world market testing that demonstrates very poor retention outside of the hardcore users (of course you're not going to hear this first hand from Valve, Facebook, Sony, etc etc directly because that would not instill confidence in the medium they've invested so heavily in).  The Steam Hardware survey shows that after 4 years there are only about 1.3 million connected VR headsets on Steam (and that's just connected, it doesn't mean they're being used), despite many more headsets actually having been sold.  The total number of concurrent VR users across all games that are actually using their headsets on Steam is roughly only 10k (up from ~5k in 2017), which is something that a single game in the top 50 on Steam is easily capable of.  You have for instance Pavlov VR which is essentially the CS:GO counterpart to VR (the gameplay, the behavior of the weapons, the game modes, literally people have ported all of the maps) but it typically gets 500 to 1K users.

Anecdotally I basically see the same thing: probably about 80% of the people I've gotten into VR (with cheap Rifts off of eBay) and those I've met online hardly use their headsets anymore despite the content improving year after year--they would rather just play flat games.  There is definitely a hardcore kind of user that sticks around but they are far from the majority.  I predict you're going to see a boost from HLA but the same general result.  During the honeymoon phase many people will claim that they can't possibly go back to flat games (you hear this over and over again) but a few months to a year later most will.  As the collective experience improves, less and less will drop off.

Quote

Current headsets are surely bearable for an hour or two - wich would already be enough for the mainstream.
The seated experience is must though. If it can't be played while sitting on the couch, they just won't play it.

In the early days of VR that's something I actually thought myself--that people would be willing to tolerate it for what VR adds to the experience, but what I've come to learn is that there's a big difference between something being just being "tolerable" and actually wanting to use it every day when you have much more comfortable/less clunky and "good enough" alternatives (the alternative in this case being flat gaming).  E.g. I can tolerate an annoying glare from a nearby window on a high quality desktop monitor, but if we assume that glare is for whatever reason unavoidable, eventually I'd almost certainly tire of it and even opt for a much worse monitor that doesn't suffer from such a glare.

VR has tons of issues like this that are constantly nagging you--there are not just glares but distortion, large parts of the scene that are out of focus, clarity problems etc etc (many of the things I listed above), and it takes a certain kind of user to actually put up with them at this point.  I actually see the opposite effect with respect to casuals: the more casual the user, the less likely they are to put up with all of this.  It's hard enough to just get them to point of gaining their "VR legs" (resilience to sim sickness).  It's similar to how in the early days of the smartphone you had a small market of individuals interested in e.g. PalmPilots and while the average person might genuinely think it was cool device, they wouldn't actually use it due to the collective problems resulting from immature hardware and software.  It was appropriate for a certain kind of user but the more casual consumer needed a user experience that was improved on basically all fronts.

Edited by woah

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...