|
Anyone knows whether Windows' game mode might have an effect on performance? I seem to have sporadic frame rate issues with Unity based games, and just now I finally came to think, that there might be issues. Say Beat Saber or H3VR getting recognized as a game, Windows doing its thing and allocating cores and resources exclusively to the game, starving the VR compositor of SteamVR out of CPU cycles, or some poo poo like that. But only sometimes depending on whatever else is going. Because I don't see why I can run either of the two games mentioned at 120hz sometimes, and other times absolutely not. And that without changes to the system.
|
#
¿
Oct 21, 2019 04:13
|
|
|
|
| # ¿ Apr 18, 2024 11:26 |
|
|
RFC2324 posted:holy poo poo I wonder if this is what has been causing my annoying stuttering --edit: Meh, must be something else then. quote:In the latest versions of Windows (1809 & 1903) Game Mode no longer does GPU/CPU prioritization. This prioritization was intended to give more resources to the running game at the cost of background processes, but we found it impacted some games and other scenarios. As an example streaming was impacted in apps like OBS because we were starving them of resources needed to encode. Combat Pretzel fucked around with this message at 13:35 on Oct 21, 2019 |
|
#
¿
Oct 21, 2019 13:25
|
|
|
I don't believe it! Custom skydomes were fixed in SteamVR Beta!
|
|
#
¿
Oct 25, 2019 22:37
|
|
|
It makes no sense, because I have had the game just break performance without changing the installation or adding songs. Any "garbage collection" is memory bullshit that would get reset with restarting the game. If you go looking for help, you'll find there's plenty of people that are complaining about framerate and performance issues past 1.2 and 1.3. With 1.4 and 1.42 poo poo was just limping along with 80-100fps and mirrors disabled, even dropping supersampling didn't help. Once I reverted to 1.3, I got my 120fps back, with mirrors enabled again and higher supersampling. Of course, that's loving fun, because Steam removed the ability to disable game updates, so I have to remember every time I want to play to go into offline mode. If you care about scores, you should try getting 1.3 to work, if you're up for reverting. They changed hit detection on that version, resulting in better scores.
|
|
#
¿
Oct 26, 2019 19:15
|
|
|
Yeah, that Palm Trees map from Rustic and poo poo like that breaks performance. But simpler maps that worked fine for months still don't manage to reach the previous performance on the latest build? As far as mods go, from what I can tell, most of the modules haven't updated since 1.3, definitely none I'm using. Yet I'm getting this huge performance difference between 1.3 and 1.4 with the same mods, so they're not it either. I don't have CameraPlus installed, either. What I see tho in 1.4x is when I actually manage to configure things in such that simpler maps reach my desired framerate, at least at the very beginning, as soon I'm hitting blocks, the frame timing graph in fpsVR is having tons of fast spikes and everything jitters. No idea what they've done, because reinstalls don't fix it here. And apparently not for other people, according to Steam forums and posts on Reddit. I even went through all the different system folders like AppData and ProgramData scrubbing settings and what not. At least you can still download the old 1.3 depot from Steam by dicking around in their dev console. It's just a pain in the rear end to actually use it, with Steam wanting to immediately update it again, unless you're "offline". I sure hope it isn't some dumbass performance hacks in common code intended for Quest optimization that's firing back.
|
|
#
¿
Oct 26, 2019 19:49
|
|
|
Figures, now I'm having the same issues with 1.3. That's what I get for opening my mouth. The first four minutes of a 6 minute song, everything works dandy, then it started doing the same poo poo. Right in the middle of it. There wasn't even anything out of the ordinary in that section, effects- or otherwise. Sigh.
|
|
#
¿
Oct 26, 2019 20:20
|
|
|
^^^ Already tried all that stuff in the past. I've been poking around with Process Explorer at it. The game spends like 75% of its CPU time in the kernel. Why? It has no business there. This seems to be some bullshit with the Unity Engine, which most VR games on my system are using. Anything else, based on Unreal, whatever else and even Steam Home, they spent not even 10% in the kernel. Unity is a loving piece of poo poo. Something in Windows 10 1909 is loving with it again. Or rather, it shits itself for god knows what reason. --edit: I'm getting frame times of up to 20-22ms on a fresh reinstall of Beat Saber. With 1.0 SS and otherwise default settings. Combat Pretzel fucked around with this message at 21:45 on Oct 26, 2019 |
|
#
¿
Oct 26, 2019 21:29
|
|
|
I said it's spending 75% of its runtime in the kernel. There's no reason for the kernel to be doing three times the work of the game. Unity is throwing the system in a tizzy in some way. I say Unity, because all Unity games are doing it here. I'm currently trying to figure out how to trace syscalls without installing half a million tools, because on Windows, things have to be complicated. Maybe then I get an answer. Also, my CPU is a 16-core Threadripper, so there's plenty of room to spare for threads running rampant. Max. core usage is 50-60% when the game runs.
|
|
#
¿
Oct 26, 2019 22:45
|
|
|
People should use more Unreal engine and less Unity. Latter is loving terrible for performance.
|
|
#
¿
Oct 31, 2019 19:02
|
|
|
gently caress me, I figured my performance issues with Unity out. Goddamn Windows telemetry data collection. It switched to full telemetry due to having had an insider build of Windows 10 installed. As soon as I forced it to basic via gpedit.msc, my frame times were cut in half.
|
|
#
¿
Nov 1, 2019 05:00
|
|
|
Zaphod42 posted:Windows telemetry is goddamn bullshit
|
|
#
¿
Nov 1, 2019 05:14
|
|
|
I ran into one of EbolaIvory's videos on Youtube. Song was called Hoohah. I thought it looked doable. I downloaded it. I concluded "I guess not".
|
|
#
¿
Nov 1, 2019 17:49
|
|
|
Pistol Whip is surprisingly exhausting. It's yet another Unity game, so it should be "easily" moddable, unless the devs went out of their way obfuscating things or their internal APIs are just a confusing mess. On the other hand, these il2cpp references in the directory tree seem to suggest that poo poo got compiled down to native code before distribution.
|
|
#
¿
Nov 10, 2019 01:24
|
|
|
A bunch of sections in Pistol Whip could certainly be less narrowly packed. Heh. Also, how can anyone play this with the pants on. Or Beat Saber for that matter. I'm sweating like a pig. Combat Pretzel fucked around with this message at 17:14 on Nov 10, 2019 |
|
#
¿
Nov 10, 2019 17:11
|
|
|
Can you tell when I start playing some more demanding VR games? Humidity graph --edit: Beat haptics is what? Controller vibrating on the beat? Combat Pretzel fucked around with this message at 18:36 on Nov 10, 2019 |
|
#
¿
Nov 10, 2019 18:32
|
|
|
Hrm, I could swear there is a slight delay between bass thumps in the sound and metronome haptics.
|
|
#
¿
Nov 11, 2019 00:50
|
|
|
eonwe posted:jesus i'm actually sore from playing pistol whip Jack Trades posted:Am I the only one who likes the Pistol Whip tracks? I hope the custom song stuff gets somewhere eventually.
|
|
#
¿
Nov 11, 2019 23:37
|
|
|
It would be nice to have a laser aim mode in Pistol Whip, just like in the menu, so that one can train a bit on the maps and get a handle on the guns to Dead Eye them eventually. Put -50% or no scoring at all on that mode.
|
|
#
¿
Nov 13, 2019 18:20
|
|
|
I read somewhere that on OC6, he seemed to complain that Oculus left a lot of low hanging fruit unclaimed, and that it seemed to irk him. Maybe there's been more going on backstage.
|
|
#
¿
Nov 14, 2019 08:51
|
|
|
Started to play the first rounds on Hard in Pistol Whip. This loving game.
|
|
#
¿
Nov 15, 2019 23:28
|
|
|
Neddy Seagoon posted:It's a fibre optic cable, it's about as optimized as it can get with near-zero latency.
|
|
#
¿
Nov 16, 2019 15:26
|
|
|
Without the risk of actual physical damage and harm, this is not fun.
|
|
#
¿
Nov 17, 2019 00:23
|
|
|
I suppose there isn't any 802.11ay hardware yet? That poo poo can supposedly do up to 44GBit/s. I'd rather see them working on wireless links instead.
|
|
#
¿
Nov 17, 2019 04:32
|
|
|
Doctor w-rw-rw- posted:Network jitter is a bigger problem than bandwidth for network. Between your PC and your access point (and possibly switch/gateway), you have a lot of room for a crappy piece of hardware to not be good enough to deliver all the frames on time.
|
|
#
¿
Nov 17, 2019 14:21
|
|
|
I'm split on this. While it might be a kick in the balls for non-VR users, the effort they finally managed to bring up to create a game again, I sure would have liked it to have gone more to finishing the goddamn Halflife trilogy. Then again, there's still the question how much of a game HL:A will be, anyway, in terms of hours of gameplay. Or gameplay itself for that matter.
|
|
#
¿
Nov 19, 2019 00:27
|
|
|
Breakfast All Day posted:A 20+ hour HL3:VR launch now would piss off everyone because only ~2% of Steam gamers have VR and most VR equipment is honestly still a generation away from not sucking. Cicero posted:Well, that's not what the benchmarks are showing. What data are you looking at?
|
|
#
¿
Nov 19, 2019 19:08
|
|
|
I have no idea how people dead-eye hard maps in Pistol Whip. Some of the maps pretty much take the wind out of me in shooting gallery mode. I'm old.
|
|
#
¿
Nov 20, 2019 19:07
|
|
|
Is there a livestream to that HL:A reveal?
|
|
#
¿
Nov 21, 2019 17:48
|
|
|
50 bux. Better be a couple of hours.
|
|
#
¿
Nov 21, 2019 19:06
|
|
|
Baron von der Loon posted:Can anyone with the Knuckles prototype confirm if they've received HL: A? It hasn't popped up in my library yet. --edit: Oh indeed. I bought it by accident because of not reading the whole product page, got it refunded and it's still listed in my library.
|
|
#
¿
Nov 22, 2019 15:25
|
|
|
There's quite some salt in the Steam discussion section of HL:A. poo poo like "Valve opening themselves to class action suits", because VR-only is apparently discrimination.
|
|
#
¿
Nov 22, 2019 17:54
|
|
|
I sure hope the release of the game doesn't result in a buy -> negative review -> refund bullshit from salty gamers.
|
|
#
¿
Nov 22, 2019 18:01
|
|
|
I certainly could use gun yaw correction in Pistol Whip. I'm already at -10° for the grip to line up with my Index controller (checking my peeking through the nose opening of the headset), but it also could use like 5° of yaw to the left, because IMO the gun doesn't sit right in that plane. When I use my second hand and apply a constant bend to compensate, I manage to hit a bit more accurately.
|
|
#
¿
Nov 22, 2019 20:23
|
|
|
--edit: ^^^ Changing the Indexes display brightness to 95% seems to do some good. I started to watch this and a few minutes in, I fail to see how that's related to any lawsuits.
|
|
#
¿
Nov 23, 2019 01:30
|
|
|
This is not true VR Until there are display headsets that can do at least 1000 Hz with almost instant pixel response time and retina resolution that close to our eyes this is sand in our eyes, VR set with display port 1.2 is a joke but i guess that's what valve considers its customers a joke. I hope other digital content distribution platforms arise from your ashes lol.
|
|
#
¿
Nov 23, 2019 03:21
|
|
|
SirViver posted:Out of curiosity, can any Index controller owners try lightly tapping them against each other (plastic to plastic contact - it needs to be a hard surface impact) and report whether that causes the tracking to freak out for a second or so? Just wanna make sure this is normal behavior. I'll try to remember to test later when I play some.
|
|
#
¿
Nov 23, 2019 14:34
|
|
|
The shooting in Pistol Whip seems to be a bit off. That, or the tracers are wrong/delayed. I noticed a bunch of instances where I dead-eye moving targets, the tracers hit them right in the torso but the bullets don't hit.
|
|
#
¿
Nov 23, 2019 16:04
|
|
|
I played Arizona Sunshine up until the mineshaft. I don't want to keep playing.
|
|
#
¿
Nov 24, 2019 23:38
|
|
|
treat posted:Index question: can anyone comment on watching movies on the Index, particularly WRT the resolution and the stock headphone quality? Is it worth it? I don't own a TV or large monitor and I'd love to be able to emulate the theater experience with a fat, hot brick strapped to my face. I watch occasional TV show episodes on it. Gizmo VR is a floating screen. Skybox VR has actual theater geometry, but it requires the motion controllers for controlling it. Bigscreen is also an alternative, it has plenty of settings and seating positions. There's also other players, but I haven't tried them yet. The three quoted can also deal with 3D movies in varying formats. --edit: Netflix doesn't work in Bigscreen, but Primevideo does. YMMV. Shine posted:I haven't watched movies on it, but the Index has loving superb headphones, better than any headset I've ever owned. Combat Pretzel fucked around with this message at 00:31 on Nov 25, 2019 |
|
#
¿
Nov 25, 2019 00:22
|
|
|
|
| # ¿ Apr 18, 2024 11:26 |
|
|
Yea I guess Bigscreen is the best option, because it mirrors the desktop (it supports multiple monitors), so you can play about anything, so long it doesn't use DRM like Netflix. That includes games. It's "Grand Cinema" environment is nice for movie watching. The scenery gets lit by whats on screen, too. Alternatively, you can pick a living room environment and place your virtual monitors all over the place and do whatever else that way. In that case, for productive stuff, you definitely need more resolution tho for longterm usability.
|
|
#
¿
Nov 25, 2019 00:46
|
|