Discussion in 'Gaming' started by RedFlames, 18 Nov 2019.
Anyone with any experience playing it non-VR ?
What...Alyx? You can't. Can you? I didn't think you could.
Re: development, I worry about the long-term economic viability of VR games development, given what you've described, edzieba. As you say, not many people have gaming VR setups, it's a little niche. And I don't see it growing massively - there are too many barriers to entry besides cost. Even if cost dropped massively (as it did with the release of PSVR, for example) most gamers actually just don't want it or can't get on with it.
Will devs ever take a break from making a lot of money, and voluntarily invest more time to make less money? I can't see why they would.
Valve, we can infer from their interviews, were simply trying to do something new and cool, and trying to infuse the VR gaming scene with some much-needed momentum. But if other devs are focused on profit, first and foremost...VR gaming's never going to explode, is it?
"It's cool, but needs a really expensive bit of proprietary hardware to run, and is it really all that much better? Nobody will make games for it if it's so expensive and so few people have one! It'll never catch on!" also applied to: graphics cards. And discrete sound devices.
It's not fundamentally harder to develop for VR. The existing engines still work and have integrated the tools needed for HMDs to run properly. The code is the same in terms of what the commands are, how syntax works, etc. It's new, not harder, which means all those tricks you've learned for successfully shipping non-VR games have to be adapted or changed or simply don't apply.
The major differences are in optimization and gameplay mechanics. VR plays differently and needs to be paced differently. You have to bake the lighting in for everything that won't move in a scene in order to save GPU for keeping the frame rate up. It also means the number of dynamic light sources has to be pushed down as well. You can help the issue by limiting the number of items in the scene that can be moved as well, but this runs counter to the idea that, in VR, players want to be able to interact with everything. That means those decisions need to be much more deliberate in terms of what's important to be interacted with, but also in terms of art direction so the world feels lived in while not feeling empty. It's a balancing act. And as I've said before, performance and optimization is important for all games, but VR must hit the target frame rate essentially all the time or people will get sick. There's different "reprojection" schemes developed for all the headsets, but if there aren't enough frames being generated in the first place, you won't have enough stuff there to interpolate a dropped frame when necessary. I'm currently running a 2080Ti at home (thanks again, work) and find that I drop the most frames in Beat Saber of all things. HL:Alyx looked amazing with almost no dropped frames at max settings and 120hz. It's been optimized to the very last inch of its existence and there's no way that was inexpensive.
On the topic of VR mechanics, HL:Alyx' weakest component may have been the shooting. It was just really easy to die and every enemy felt like it had a trick for how to defeat it. 'This enemy will shoot six rounds then pause. I need to wait behind cover and then shoot during that break. This enemy will release manhacks if I maintain cover too long, so I need to keep moving and try to flank them. Antlions just advance slowly like a weird (but fun) Pirates of the Caribbean ride.' Another notable weakness for me were the Combine puzzles to get into their ammo crates and upgrade stations. HL games have always tried to vary the gameplay with action and puzzles, but these felt like filler. Having the puzzle tied to alien technology works (almost) in terms of metaphor. I would have been more irritated if I was playing some weird mini-game in order to hack a human computer, but I still wasn't convinced by it. It sounds weird, but my biggest complaints about the game all revolve around the things that make VR "necessary" to play it. But if they had just not made those design decisions, the game could have been a standard WASD affair and still been great, just not innovative.
Back on topic, the reason we don't see more AAA games (other than bad ports) in VR is the fact that it's very very expensive. I recently watched a talk (which I can't seem to find right now) of a professional in the industry talking about just how expensive game development it. He broke down a simple street scene for a AAA game: "This bench, $3000. This light pole, $1800. This building, $8000 . . ." By the time he was done adding up the cost to make a street that probably didn't play a significant role in the game's progression, it was over $100k.
AAA games cost hundreds of millions of dollars to produce. They also make more than anything else; more than movies, music, TV. This is possible because these games get released for PC and all major consoles simultaneously. Total player base across all platforms is in the 100s of millions, so capturing a small portion of that at $60 each will earn your money back. This just isn't the case for VR.
PC compatible HMD adoption may be in the range 14-20 million units total. Even if you get 100% of the players buying a game at $60 (which you won't), you're still only starting to get near the space that most AAA developers live in for their major releases.
Valve's motivation in this seems to have a few facets. First, Gabe Newell has stated that he's been jealous of Nintendo, who can manufacture hardware and software concurrently, which allows for innovations and allowances for both simultaneously. With the Index and HL:Alyx, they were able to deliver on that for the first time.
Second, I think Valve needed to show what could be done. Despite my criticisms, HL:A is amazing. The Index has been sold out since HL:A's announcement and for good reason. The current wait time is 8 weeks from time of order. Hopefully other developers notice the success of both the game and hardware adoption and start dedicating some talent for AAA VR games, even if they're similarly small 6-12hr games.
Third, I think Valve knew they needed to return to the Half-life universe but didn't know how until VR was a way to innovate in the FPS space. The whole industry has iterated year after year, but it's been pretty much stagnant in terms of mechanics and game play for over a decade. The Half-life series has always done something to break new ground and VR was that thing.
I saw the other day that someone had produced a hack for it so can use keyboard and mouse to play it.
Yesterday, I finished the game for the second time. I played on easy this time and was bit more bold for my gun battles and for a few scripted sequences where I think it's somewhat expected that you take a little damage.
There were definitely parts I liked more the second time than the first. I spent less time hunting for resin because I remembered where most of it was (and wasn't) and I didn't take time to stare at every little detail. I also noticed some new things I didn't during my first play through. The combine puzzles were less annoying. Essentially, I wasted less time and the game was better for it.
Despite excellent advice I was given. I made an impulse purchase on Oculus Quest last week when it was briefly in stock.
While waiting for delivery. My i7 2700k is 10% slower in single thread than i5 7500, so I overclocked it by 30%. My 970 is 10% slower than 1060, I overclocked 10% factory OC by another 8%. Shame I can’t download more VRAM
The recommended link cable isn’t recognised, might be because the motherboard isn’t charging the headset for some strange reason. Also might be dodgy cable, might be Intel 2000 chipset is too old.... too many unknowns.
Fortunately I was able to play HLAlyx wirelessly using ALVR, within LOS to the router. Playing at 72 Hz low setting, game looks immense. So far, only noticing a few dropped frames with hologram in Russell’s lab. Everywhere else it’s butter smooth and very immersive.
Controller tracking is spot on, the touch controller feels natural. Flick and grab works great. I’ll get a strap to allow Index-like grip release.
Now that it looks like headsets are coming back in stock, I'm getting tempted again,
Is a Strix GTX980 going to be able to run Alyx on a Rift S?
I completed HLAlyx twice using 970. 980 should be no problem.
Your i5 4690 might be problematic though. Min requirement is i5 7500, which my overclocked i7 2600k is just slightly faster. Give it some beans (overclock) and give it a try.
My brother was curious about it and he has my old 980Ti. Looking around on te intertubes, it does well for up to medium settings. A 980, with a bit of overclocking, should let you play it comfortably at low settings (backed up by @wyx087).
Doesn't look too different to me on the various settings, though perhaps someone else who actually has the game can comment.
Going to be OC'ing my CPU once I decide on a new cooler.
Cheers guys, have been pondering an upgrade, need more monies though as can't really afford VR and gfx card and new case and CPU cooler (pref water), .
Might just stick with my current backlog of games for a bit longer.
Separate names with a comma.