Yeah I'm on a 120hz monitor, and I definitely notice that my setup doesn't go over 60FPS really at all, and I'm on a 960GTX + i7 980X.
This is extremely low for your setup as DC said. Here are some instructions on how to get help with your abnormally bad performance.
Type "p_logall" without quotes in the console (~ key) whenever you want to begin recording in game (don't do this until you at least spawn in a round)
then you can p_endlog /or just wait for the round to end/ or map to change/ or just quit the game and it will end automatically.
Then navigate to your hidden folder
c:/users/YOU/appdata/roaming/natural selection 2/
And you will find your file, something *.plog
Zip it/ compress it.
Upload it to somewhere on the internet
Make a thread in the tech support forums, and put a link to your .plog file there.
Before you make your post read these instructions. http://forums.unknownworlds.com/disc...-read-me-first
This will give UWE a frame by frame analysis of what is going on, so try to pick a time when you can best reproduce your lowest performance like the 24 fps you said you get.
Don't record for longer than 30 minutes, and don't record for shorter than 5.
Include your techsupport.zip file with this compressed plog.
I'm running on Linux btw, with 3 monitors, also I think there's a new driver update for my graphics card that I haven't gotten to just yet, hope that'll help.
A huge part why modern online gaming has moved to lag compensation was precisely for consistency. You don't have to know your enemy's latency at that moment in order to determine how far ahead to lead your shots.
You can just aim at them and fire.
I think you mean "your own latency"
Here's my intepretation of latency compensation:
1. Quake came out. People enjoyed it at LAN parties but found the massive delay in input annoying for online games (because you see, there was no client prediction, the only thing that moved as you commanded was your mouse, but things like the WASD / jump keys were lagged behind by double your ping + 50msec + framerate intervals)
2. Quake 2 came out with better client prediction. Maybe some modders decide to use this concept or made their own, who knows, but they certainly put client prediction into QuakeWorld. Quake 2 still didn't have any latency compensation, only client sided client prediction which allowed your character to move (appearances) which was a bonus because you only needed to lead your aim instead of leading your aim AND compensating for your own movement lag.
3. QuakeWorld came out with perhaps the world's first form of latency compensation. I don't know what date its latency compensation code was released, but judging by how Q3A didn't have latency compensation but only client prediction, it was probably after Q3A.
3. Quake 3 Arena came out. Q3A didn't have any latency compensation but when it did, it was in the form of the g_unlagged mod and CPMA mod which from what I heard, used double prediction up to something like 80msec ping (according to various people, I don't have backing on this) instead of dealing with server-side latency compensation. Iirc, the CPMA mod was originally created by 3 people for E-Sports with specialised netcode, physics and balance modifications.
4. Games from there on used the Quake 3 Arena unlagged mod/QuakeWorld netcode concept of basically back tracing / "shifting the target(s) back" to calculate for hits.
Even games like Tremulous, World of Warcraft* and Unreal Tournement** didn't have latency compensation without modifications or waiting until later versions. I have no idea if Half Life/GoldSRC had latency compensation, but I think it does considering its engine was basically VALVe's forked off QuakeWorld. I think the GSRC engine was probably the first engine to involve advanced hit boxes.
*Around 2008 or so, the oceana server didn't exist or something. My brother basically was playing on the U.S. server but problems existed with some of his ranged attacks and melee attacks not registering despite being "alllowed to attack" with its click and hit system
**I don't know much about UT, but from what I heard, it started having latency compensation with its 2004 version. Don't quote me on this one.
I'm running on Linux btw, with 3 monitors, also I think there's a new driver update for my graphics card that I haven't gotten to just yet, hope that'll help.
Driver won't matter, linux can only run ns2 in opengl, which will detriment your performance pretty bad.
My fps on win: 120-160
My fps on nix: 50-80
Same hardware, maybe the win ssd has a bit higher writespeed, but i dont think that has anything to do with it.
Just rebooted after going to 375 (driver). I suspect it's more that I haven't rebooted in ~25 days (due to work constraints), but my FPS is like 60-80 now, a touch better than before.
You're probably right about the openGL part though, that's why I'm all for Vulkan instead of any DX version.
I'm running on Linux btw, with 3 monitors, also I think there's a new driver update for my graphics card that I haven't gotten to just yet, hope that'll help.
Driver won't matter, linux can only run ns2 in opengl, which will detriment your performance pretty bad.
My fps on win: 120-160
My fps on nix: 50-80
Same hardware, maybe the win ssd has a bit higher writespeed, but i dont think that has anything to do with it.
QuakeWorld most definitely came out before quake 2 and quake 3, not after Q3A. You can thank QuakeWorld for what seriously kicked the FPS genre into high gear. It did a lot of awesome things, one of note is making the game actually reasonably playable on dialup. "HPB" (High Ping Bitch) was a moniker coined in the QW days, referring to players who, despite their high ping, did rather well at the game.
If it wasn't for QW, I would not have been able to host an actually good server on dialup. Boy was that fun.
UWE Department totally gives the vibe "Well, some people still like to play 60fps on a 60hz screen" but no.. the biggest majority i think is currently on 144hz and kinda demands 144fps with semi-average computers these days on the lowest settings.
Currently 93% of gamers have a primary display resolution of 1920x1080 or less*. In fact only just over 44% of gamers have a 1920x1080, so 49% of gamers have a display smaller than 1920x1080. I doubt very much that most people are on 144hz monitors.
dePARAJoin Date: 2011-04-29Member: 96321Members, Squad Five Blue
4,5 years after release the "we have no time for performance optimisations" argument is a bit flimsy.
The game feels just horrible for me now. Down from rockstable 180fps 2 years ago to 130 (sometimes 90 fps in endgame) i have still to shoot on spastic moving and warping skulks.
Sometimes it feels like im playing the beta again.
Making a seeding system for 150 player (im sure we will have this ammount in 2 month during summer) is kinda a joke.
All these changes made in the past 4,5 years but the game still feel like @ release (sometimes worse).
After a 6 month break where i played Battlefield4 i tried multiple weeks to enjoy NS2 again.
But for me it was the worst time i ever had in the game. From a technical view and its just sad to see where this great game ended.
80% of the players have zero clue how to play aliens for example. Its like pain in the ass to play on alienside these days.
So your "retention program" might had a impact on the playercount to atleast hold the playercounts but if failed hard on multiple other levels.
Looks like NS2 is going to have a inbuild aimbot for rines and a wall @ the floor walking skulks head in the end to please the 100 remaining players gained by the retention trys.
4,5 years after release the "we have no time for performance optimisations" argument is a bit flimsy.
By "we" you mean me, because I am not speaking for UWE. I am not saying there isn't time, but that the time is probably best spent on things other than performance. This is my best guess at why UWE is not working entirely on performance. UWE isn't completely ignoring performance either. I just mentioned some other things they are working on.
I sincerely doubt you had a rockstable 180fps to years ago. I certainly did not and I had excellent hardware.
4,5 years after release the "we have no time for performance optimisations" argument is a bit flimsy.
By "we" you mean me, because I am not speaking for UWE. I am not saying there isn't time, but that the time is probably best spent on things other than performance. This is my best guess at why UWE is not working entirely on performance. UWE isn't completely ignoring performance either. I just mentioned some other things they are working on.
I sincerely doubt you had a rockstable 180fps to years ago. I certainly did not and I had excellent hardware.
To be honest: I used to get much, much better frames in NS2. Timeframe: 2 years ago or so (my hardware hasn't changed in that time). I play with very low settings and a 144hz monitor, and I used to consistently stay above 144 frames. Since coming back to the game in the last few months, I've noticed that performance is notably worse. It's not game-breaking, and the tearing isn't too bad because I rarely drop lower than 120, but it never used to be an issue. Not sure what the cause is, though.
some ppl used to say ns2+ caused them to have lower frameskis. I'm not sure, but i have noticed after map changes when some mods dont load properly, the game is sometimes much smoother seeming. i definitely used to notice a difference in old comp mod especially, lots of choppiness and tearing.
are any common mods known to impact performance? i have only anecdotes and personal experience to go off, but has anyone done any kind of test?
This just shows how much the UWE development team doesnt care about performance, a QoL change that could change how the game feels like...
@IronHorse did you go in hibernation mode? Or there's a good argument WHY performance isnt a priority?
You've had what.. about 5 years now worth of time with developing the game? And still "no time" like DePara said is a bit flimsy.. UWE has run out of excuses, you cant use that argument anymore.. we could've understood the 2 first years, but what's the excuse for the last 3 ones?
To be honest: I used to get much, much better frames in NS2. Timeframe: 2 years ago or so (my hardware hasn't changed in that time).
So the reduction in performance is 100% down to NS2 and there's not even the slightest chance that your components are degrading in performance after so many years?
To be honest: I used to get much, much better frames in NS2. Timeframe: 2 years ago or so (my hardware hasn't changed in that time).
So the reduction in performance is 100% down to NS2 and there's not even the slightest chance that your components are degrading in performance after so many years?
Framerate-wise, anything like that would be almost imperceptible.
Kouji_SanSr. Hινε UÏкεεÏεг - EUPT DeputyThe NetherlandsJoin Date: 2003-05-13Member: 16271Members, NS2 Playtester, Squad Five Blue
edited March 2017
You'd see artifacts from hardware failure long before any perceivable performance degradation, hardware tends to be finicky like that. Suddenly it stops working, unlike mechanical HDD's those tend to give plenty of time to tell you something's wrong
This just shows how much the UWE development team doesnt care about performance, a QoL change that could change how the game feels like...
@IronHorse did you go in hibernation mode? Or there's a good argument WHY performance isnt a priority?
You've had what.. about 5 years now worth of time with developing the game? And still "no time" like DePara said is a bit flimsy.. UWE has run out of excuses, you cant use that argument anymore.. we could've understood the 2 first years, but what's the excuse for the last 3 ones?
Can we drop the overly dramatic "UWE doesn't care!" tone? I never said we don't care about about performance.
I never said it's not a priority.
I never said we don't have time.
So allow me to reiterate what I said for clarity's sake:
"There are other priorities in the works right now. That being said, increasing the send rate has been a target and a desire for the team for a long while now... it's just a lot of work, that's competing with low hanging fruit priorities."
There's your answer, Rysk. Try to consider the priorities chosen from the point of view of UWE: they could drop everything right now and spend potentially 4-6 months rewriting swaths of the engine and continue to lose players and for whatever that happens to gain them in the end, or they could spend a few weeks and create something that makes the game more accessible, get's people into balanced games faster, and overall improves retention. It's a risk vs reward assessment. If there are improvements that can have a bigger impact and are less work, they will be chosen first. That doesn't mean server performance is not a priority, and it doesn't mean it's not something that everyone is painfully aware of. It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development. (And btw.. that answer was in regards to server performance, not client, idk how we got onto that topic as it's not even relevant to the thread. Client performance is continuous and hasn't really ceased, ever, to include upcoming patches.)
Also, I was away from the forums for 3 days. And for someone who works 3 jobs, I hardly consider that "hibernation".
If you ever have something that's time sensitive, feel free to PM me in our discord server. Otherwise, I'll get to the forums when I can.
i have still to shoot on spastic moving and warping skulks.
80% of the players have zero clue how to play aliens for example. Its like pain in the ass to play on alienside these days.
So your "retention program" might had a impact on the playercount to atleast hold the playercounts but if failed hard on multiple other levels.
Looks like NS2 is going to have a inbuild aimbot for rines and a wall @ the floor walking skulks head in the end to please the 100 remaining players gained by the retention trys.
1) I mentioned that warping solution previously in these forums, and it might get revisited at some point, but for now its shelved for the reason I stated.
2) There are multiple tutorials available, but since there's only a soft requirement to complete them (20+ hours in rookie servers OR tutorial) many still don't. There's not much we can do here.. we've provided the tools, the environment, and the education.. but outright forcing players will result in friction and possibly loss of players. Hell, we've even had many complaints that our current soft requirement is stupid and people have said they've uninstalled as a result. Short of a complete redesign/new simpler game there's not much more we can do.. but... that brings me to the next point
3) You cannot complain that players have no idea what they're doing, and then simultaneously advocate for the opposite of simplifying and making the game more accessible. Don't you see that its the intricate nature of the game that makes said players not intuitively know what to do? Retroactively making the game accessible has been a goal of UWE's for a long time now for that exact reason. You either support continued accessibility work or you need to be okay with fewer players, and players that know far less, since they are inversely related.
Lastly, I'm sensing a lot of angst in here.. and while I really want to continue to answer questions, I'm not willing to engage with people who continue to be derisively sarcastic, belittling or demanding.
I'm not a community manager and I take my own time to field your questions. I realize some of you still aren't happy with the state of things despite the progress over the years, but I'd ask you to try to remain reasonable and friendly instead of demanding and quasi disgruntled. It's the only way we can continue an open line of discussion, if that's what you're actually interested in.
So the reduction in performance is 100% down to NS2 and there's not even the slightest chance that your components are degrading in performance after so many years?
Yeah. It's not necessarily just NS2 in isolation - it could be the result of how NS2 interacts with newer drivers, for example - but it's absolutely not hardware "degradation" for the reasons already mentioned by others. Also, my performance in other games hasn't suffered similarly (it's actually improved with fresher drivers in most cases; I don't really play many new titles so I'm not exactly pushing the envelope). I'm not trying to shit on UWE here. I said it wasn't game-breaking, and I'm enjoying the game just fine. What I am saying is that I can relate to others who've also seen their framerates decrease. I can't attribute it to a specific patch or mod because I took a hiatus from the game and didn't chart performance from patch to patch.
This just shows how much the UWE development team doesnt care about performance, a QoL change that could change how the game feels like...
@IronHorse did you go in hibernation mode? Or there's a good argument WHY performance isnt a priority?
You've had what.. about 5 years now worth of time with developing the game? And still "no time" like DePara said is a bit flimsy.. UWE has run out of excuses, you cant use that argument anymore.. we could've understood the 2 first years, but what's the excuse for the last 3 ones?
Can we drop the overly dramatic "UWE doesn't care!" tone? I never said we don't care about about performance.
I never said it's not a priority.
I never said we don't have time.
So allow me to reiterate what I said for clarity's sake:
"There are other priorities in the works right now. That being said, increasing the send rate has been a target and a desire for the team for a long while now... it's just a lot of work, that's competing with low hanging fruit priorities."
There's your answer, Rysk. Try to consider the priorities chosen from the point of view of UWE: they could drop everything right now and spend potentially 4-6 months rewriting swaths of the engine and continue to lose players and for whatever that happens to gain them in the end, or they could spend a few weeks and create something that makes the game more accessible, get's people into balanced games faster, and overall improves retention. It's a risk vs reward assessment. If there are improvements that can have a bigger impact and are less work, they will be chosen first. That doesn't mean server performance is not a priority, and it doesn't mean it's not something that everyone is painfully aware of. It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development. (And btw.. that answer was in regards to server performance, not client, idk how we got onto that topic as it's not even relevant to the thread. Client performance is continuous and hasn't really ceased, ever, to include upcoming patches.)
That's the problem, the priorities compared to what the community wants, and what the developer teams of UWE is miles off from each other. It should've been quite clear at this stage of what's the problem with the current state of the game is. The performance is just too bad, it's nowhere near any other game out there. Even playing Early Access games in Unreal Engine 4 for example is 100 miles better. I know it's wrong to compare the engines, but they've seemed to atleast known that the whole network code into an engine is quite vital/crucial.. It's just important.
You almost make it sound like those 4-6 months are a "waste of time".. Well, let me give you some info.. That "waste" of time could actually bring you back a bigger audience, people would stop complain about rubberbanding perhaps.
So yeah, for atleast for a chance try and do something to see the bigger picture of the game... instead of doing small changes all the time just to maintain the last 200-300 "active" players.
I know UWE doesnt have a CM, that was the biggest issue UWE had aswell, nobody listened to the community and finally you lost all your players up to this point.
And this is both for Client and Server performance.
When you say this aswell is what concerns me, clearly you havent read the forums at all, or any of the devs over at UWE " It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development"
For several years, the people have been bashing on the performance, rubberbanding ... But somehow you manage to say it's only me? From what i've seen, it's been the biggest topic for last 5 years.
The problem is that you've had moderators on this forum who's been doing posts like "Pm sent - mod" and later on closed those kinds of threads.. Cause having general discussions about the game where people are pointing on the issues is a bad thing.
It's just in the end people have finally given up, cause it's clear that it's NOT a priority over at UWE dev team... Meanwhile the community says otherwise..
I guess this will be my final post concerning the performance.. Take it how you want, but keep on going with the current path how the game is going you're not gonna gain any kind of player retention compared to how it could be. Cause "risks" is apperently a bad thing to take...
I'm with @IronHorse , you can't expect the (part-time?) dev team to make a huge effort and revamp the engine code for +10% performance. It isn't efficient in terms of cost/value.
UWE is a business first, and then a group of people who created the game and want us to have fun. They gotta feed themselves too, you know...
Also, this is a technical issue, of which we have no insight. Have any of you seen a single line of C++ code of the Spark engine? Do you guys have any idea how to integrate a LUA JIT?
Balancing is another question, community input should be weighed more heavily in that regard.
Yes, I'm also frustrated by homing shots around corners, unregistered bites and fps drops in late-game, but I'd much rather see 64bit implemented first, so the modders have more space to operate in.
PS: Remove the HP bars please! It's been some time since anyone complained about it :P
I'd bet that while typically 50% of gamers play at less then 1920x1080.... NS2 loyalists are not typical.
My last computer that I built in 2009 ran NS2 at 45-80 FPS @ 1920x1080 (its at my in-laws and I still play on it). I saved up and bought my current computer that I built in 2015 plays 120-200 FPS at 1440p.
what else can we expect? It's amazing enough that UWE/Dev team is working on an "old" game with 200 players... If you demand better performance buy a better PC.
This kind of blew beyond the original intent I had for this thread, yikes.
I just felt like the hit reg needed some lovin and wanted to bring attention to it so UWE could start planning around addressing it. I myself don't want UWE to turn on a dime here, because there's a lot about NS2 I do enjoy.
Also, I can't really say about the "drop" in performance people have been observing, but let me tell you something I myself have observed, for the better. And that is, playability on Linux has DRASTICALLY improved compared to when NS2 "shipped" (as in release).
If memory servers me right, NS2 shipped without Linux support, and UWE said "we'll do it, please bear with us", or something to that effect. A bunch of months later they rolled out Linux support for the client and at some point the server (I think the server came later? I can't recall for sure). My experience with the Linux client was rather bad. I would get into the game, and just a black screen, and never be able to actually interact with anything because all I see is black, not even a mouse pointer.
So I let it sit, checked back a few months later, same thing. Checked a few months later, oh, I can see stuff now, get into game, my frame rate is pretty unplayable, like seriously. So I let it sit some more.
A while later, like a year or two, I try again, and my framerate is quite a bit better, now playable, but loading times sucked.
This went on for a bit, checking in every now and then...
Recently I loaded NS2 back up, and it's like... worlds better than it's ever been for me on Linux. Not only does the game start in a good amount of time, loading maps is a good pace (it would be nice to be faster, but I'm okay with how it is now). But the big thing that I noticed recently, was the way it handled texture streaming. I load into a game, get low res textures, and the higher res ones stream in as I keep playing.
This one texture streaming function alone has made my enjoyment of the game worlds better. I get into each map/game way faster, and my framerate is good before and after the textures load, so I can just start playing. There's a few weird bugs that I can't quite explain just yet that I get in erratic ways, but they are so few and far between that I just let it slide (until I can get enough info to report/reproduce them).
As it stands, for me, the hit reg I would like to see improved. But I greatly appreciate so many of the other improvements that have come along the way. Many other dev teams would have scrapped Linux support forever and a half ago. But UWE has stuck by it, and I for one, really appreciate it! I've dumped a bunch of seed money (check my icon) at various points of the dev cycle, and I'm glad it has not gone to waste. The way match making works now, is really awesome, and I'm seriously considering spinning up a server in my area to help foster bringing people back into the game.
I think what would be worthwhile, is to continue the discussion of hit reg, but maybe we should do our best to keep it objective. Clearly, hit reg is an important element to people of the FPS' genre, so having productive feedback between players and the UWE crew, I hope that can turn into positive results. But these things take time, and I feel both camps aren't quite taking each other's interests to heart. I know UWE cares, but some of the things said here has made me feel like the value in hit reg isn't quite received by the UWE peeps. But at the same time, there's some people in this thread that are taking it a bit too far. UWE didn't kill your cat, so maybe we should just change the way things are presented hey?
My performance went down after Win10, but didn't really notice it, and after upgrading to a RX480, while getting a massive improvement, it wasn't the improvement expected.
Turns out from user benchmark, while most of my system is great, my Gen1 SSD is slowing down and is actually performing in the 21st percentile compared to all other users benching with my drive. That's a lot of wear. In fact it is so much wear that I am losing performance in everything, but especially games.
I only bought the thing in 2014, so it's not even that old, but it's degredation is having a massive effect on my system.
Kouji_SanSr. Hινε UÏкεεÏεг - EUPT DeputyThe NetherlandsJoin Date: 2003-05-13Member: 16271Members, NS2 Playtester, Squad Five Blue
edited March 2017
SSD's are a special case though with a maximum write per memory cell, also if the drive is near full (~85-90%) it will have to keep moving data around the whole drive and will get a lot quicker to that "cellwritecount"
This is quite unlike RAM/VRAM and cpu's (motherboards), they only get wear and tear if you overclock them and they run at higher temperatures due to higher voltages, if your cooling isn't sufficient. If they are running within specs (or OC within acceptable temps), there wont be any issues.
Also it's come to my attention that with older (no longer supported) hardware comes the annoying bit that the newest drivers actually have the potential of being the main cause for performance degredation. Rolling back your drivers can in those cases help. For 3D cards the update program usually suggests to run an older version with more optimized support
Not mentioning dust inside the case or an older windows installation starting to bog down (win7 and higher seem to be coping a lot better than Vista and before...)
When you say this aswell is what concerns me, clearly you havent read the forums at all, or any of the devs over at UWE " It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development"
For several years, the people have been bashing on the performance, rubberbanding ... But somehow you manage to say it's only me? From what i've seen, it's been the biggest topic for last 5 years.
No, it's not only you.. clearly.
I just wanted to demonstrate that not understanding the reasons behind the priorities and goals of the dev team is the source of your frustration and misunderstanding.
Do you honestly believe that if NS2 was able to have a sendrate of 60 tomorrow, that suddenly player retention would improve? That alllll the other complaints from those negative steam reviews would vanish, like imbalanced rounds or steep learning curve? Do you think that newer players would even recognize the improvement after months of our efforts? Would UWE practically see increased revenue from it?
This is the issue: It's a priority, but it's not the top priority, and that upsets you because you feel it's what is best for the product - and while I obviously fullheartedly agree that it's needed, it's an epic amount of work to implement, with potentially not much reward in a business sense, at least compared to the other top priorities. So this is the impasse that we are in, that stems from a lack of a development point of view. It's easy to say "fix it" without understanding the risks involved for a part time team - which btw the risk that you feel is so easy to dismiss is the risk of not having NS2 be developed further - so the risks need to be weighed very carefully by the team. But please don't assume the forums aren't read or the issues aren't thoroughly understood... If there's any failure involved on the development side it's clearly communicating the reasons behind the priorities, which I hope I've at least made a dent in, with my comments?
As for client side performance, it's continual.
The differences from launch to today are vast, with big steps like VMs to LuaJIT, to dozens of small wins with 10% here, 20% there throughout updates over the years.
I am going to receive flak for this, but the fact that Depara is complaining about only getting 150-200 fps makes me smile a bit... since the bar has obviously been raised over the years. It means that while clientside has improved, the expectations have shifted with it. Which is fine. I much rather that be the case than the dire post launch complaints of only getting 40-60 fps. His very familiar complaint of "i have 140 fps but it feels like 80" is being looked into as I write this.
So like I said, clientside performance is continual.. we're even testing out a 10% improvement on CPU costs this week internally, with even much more exciting stuff coming down the line later that I wish I could talk about. (I won't until its at least tested, so I don't create false hopes)
It's the server performance side that is far far more involved, sadly.
Hope that at least provides some answers and transparency for you, even if you disagree with the decisions. @BloodyIron
Again.. I don't think you mean Hitreg.. I think you mean the scenario of dying around a corner, which is due to sendrate + interpolation server settings.
IF you did mean hitreg, that is the upmost importance to us, so please collect some video of it using @Nordic 's thorough How To Objectively Test Hitreg method.
In regards to hit reg, I am revisiting through continued play to determine what the issue I have is. Because, I actually had forgotten about the physical dimensions of a skulk, compared to that of a "terrorist" in CSGO, or a "soldier" in TF2, which are the dimensions I am historically more accustomed too. As such, I am not entirely certain if my "issue" is real or not just yet.
One thing I think I can tell you is sometimes it feels like incoming damage from rifles to a skulk "bursts", in such a way that it makes it nigh impossible to react to. What I mean by this is, say I'm chowing down on a stupid marine power node (stupid marines), my ass sticks out and marine shoots me. I try to get away, but by the time my position in the world has moved, I have already received more than 50% of my life in damage, and a "split second" later I receive the rest of the damage. This is kind of hard to refine as an explanation, but the issue in this example I have is the bursty-ness nature of the incoming damage makes it seem almost impossible to react to and retreat from. I've tried to pay attention to the pings, and this does happen at times when I have 60-80ms ping to the server, and the stinky marine that killed me has a similar ping too. It does indeed happen with higher pings too, but it does seem to with lower pings.
Now, I am testing more and more, and it _MAY_ be related to server performance too, but the servers I play on report as "happy face" performance. I don't have enough samples to be confident of this theory just yet, but it may be that I see this more on some servers than others...
On another, kinda related note. I do think I am getting "frame latency" of sorts. While I'm getting 60-80FPS or so, it doesn't feel smooth at times, sometimes I get jittering, I think. I'm on a 120hz monitor, so I can tell the difference between 60fps and 120fps, but that's not quite what I'm experiencing here. I'm on a 960 GTX, so I don't think this is related to the AMD frame latency of yester years. Is there any chance we could see Vulkan for NS2?
What SSD do you have? How big is it and how much free space do you normally keep on it? What OS are you running on? If you converted from an HDD to an SSD previously, a possibility is that TRIM may not be on, check that out.
My performance went down after Win10, but didn't really notice it, and after upgrading to a RX480, while getting a massive improvement, it wasn't the improvement expected.
Turns out from user benchmark, while most of my system is great, my Gen1 SSD is slowing down and is actually performing in the 21st percentile compared to all other users benching with my drive. That's a lot of wear. In fact it is so much wear that I am losing performance in everything, but especially games.
I only bought the thing in 2014, so it's not even that old, but it's degredation is having a massive effect on my system.
...also if the drive is near full (~85-90%) it will have to keep moving data around the whole drive and will get a lot quicker to that "cellwritecount"
One thing I think I can tell you is sometimes it feels like incoming damage from rifles to a skulk "bursts", in such a way that it makes it nigh impossible to react to.
Now, I am testing more and more, and it _MAY_ be related to server performance too, but the servers I play on report as "happy face" performance.
On another, kinda related note. I do think I am getting "frame latency" of sorts.
That description aptly falls into the category I mentioned, sendrate and interpolation. If updates are sent more frequently between clients then there would be less of a burst of information, not to mention the interpolation delay that comes with said lower sendrate compounding the issue.
Server performance ratings just speak to how well the server is running, not whether it is running faster rates with lower interp (most don't, and those that do cannot do enough due to the performance strain and bugs associated).
If you join a server and see a popup window telling you that network rates are different from default, *that's* a server that is running modified rates. FYI, you can use the console (~ key) and type perfmon three times to see what rates are being used.
You will most definitely experience some input delay / frame latency with linux unfortunately.
I know it's not a popular thing to say to a linux user, but the best performance regarding input delay will be from using Windows, DX9, Fullscreen, raw input on, AO off.
NS2 would benefit greatly from Vulkan, especially for linux users .. but I don't know of any plans to implement it as of right now.
Comments
And I hit averages of over 100fps just fine. You sir, have a issue somewhere in your setup.
This is extremely low for your setup as DC said. Here are some instructions on how to get help with your abnormally bad performance.
Also guys, speaking of hitreg. We have a way to test that. Here is the thread with instructions on how to objectively test hitreg.
I think you mean "your own latency"
Here's my intepretation of latency compensation:
2. Quake 2 came out with better client prediction. Maybe some modders decide to use this concept or made their own, who knows, but they certainly put client prediction into QuakeWorld. Quake 2 still didn't have any latency compensation, only client sided client prediction which allowed your character to move (appearances) which was a bonus because you only needed to lead your aim instead of leading your aim AND compensating for your own movement lag.
3. QuakeWorld came out with perhaps the world's first form of latency compensation. I don't know what date its latency compensation code was released, but judging by how Q3A didn't have latency compensation but only client prediction, it was probably after Q3A.
3. Quake 3 Arena came out. Q3A didn't have any latency compensation but when it did, it was in the form of the g_unlagged mod and CPMA mod which from what I heard, used double prediction up to something like 80msec ping (according to various people, I don't have backing on this) instead of dealing with server-side latency compensation. Iirc, the CPMA mod was originally created by 3 people for E-Sports with specialised netcode, physics and balance modifications.
4. Games from there on used the Quake 3 Arena unlagged mod/QuakeWorld netcode concept of basically back tracing / "shifting the target(s) back" to calculate for hits.
Even games like Tremulous, World of Warcraft* and Unreal Tournement** didn't have latency compensation without modifications or waiting until later versions. I have no idea if Half Life/GoldSRC had latency compensation, but I think it does considering its engine was basically VALVe's forked off QuakeWorld. I think the GSRC engine was probably the first engine to involve advanced hit boxes.
*Around 2008 or so, the oceana server didn't exist or something. My brother basically was playing on the U.S. server but problems existed with some of his ranged attacks and melee attacks not registering despite being "alllowed to attack" with its click and hit system
**I don't know much about UT, but from what I heard, it started having latency compensation with its 2004 version. Don't quote me on this one.
This is what playing without Latency Compensation Looks Like:
https://coub.com/view/t0225
Driver won't matter, linux can only run ns2 in opengl, which will detriment your performance pretty bad.
My fps on win: 120-160
My fps on nix: 50-80
Same hardware, maybe the win ssd has a bit higher writespeed, but i dont think that has anything to do with it.
You're probably right about the openGL part though, that's why I'm all for Vulkan instead of any DX version.
If it wasn't for QW, I would not have been able to host an actually good server on dialup. Boy was that fun.
Currently 93% of gamers have a primary display resolution of 1920x1080 or less*. In fact only just over 44% of gamers have a 1920x1080, so 49% of gamers have a display smaller than 1920x1080. I doubt very much that most people are on 144hz monitors.
*As of most recent Steam Hardware Survey - http://store.steampowered.com/hwsurvey
The game feels just horrible for me now. Down from rockstable 180fps 2 years ago to 130 (sometimes 90 fps in endgame) i have still to shoot on spastic moving and warping skulks.
Sometimes it feels like im playing the beta again.
Making a seeding system for 150 player (im sure we will have this ammount in 2 month during summer) is kinda a joke.
All these changes made in the past 4,5 years but the game still feel like @ release (sometimes worse).
After a 6 month break where i played Battlefield4 i tried multiple weeks to enjoy NS2 again.
But for me it was the worst time i ever had in the game. From a technical view and its just sad to see where this great game ended.
80% of the players have zero clue how to play aliens for example. Its like pain in the ass to play on alienside these days.
So your "retention program" might had a impact on the playercount to atleast hold the playercounts but if failed hard on multiple other levels.
Looks like NS2 is going to have a inbuild aimbot for rines and a wall @ the floor walking skulks head in the end to please the 100 remaining players gained by the retention trys.
By "we" you mean me, because I am not speaking for UWE. I am not saying there isn't time, but that the time is probably best spent on things other than performance. This is my best guess at why UWE is not working entirely on performance. UWE isn't completely ignoring performance either. I just mentioned some other things they are working on.
I sincerely doubt you had a rockstable 180fps to years ago. I certainly did not and I had excellent hardware.
are any common mods known to impact performance? i have only anecdotes and personal experience to go off, but has anyone done any kind of test?
https://forums.unknownworlds.com/discussion/140821/performance/p1
@IronHorse did you go in hibernation mode? Or there's a good argument WHY performance isnt a priority?
You've had what.. about 5 years now worth of time with developing the game? And still "no time" like DePara said is a bit flimsy.. UWE has run out of excuses, you cant use that argument anymore.. we could've understood the 2 first years, but what's the excuse for the last 3 ones?
So the reduction in performance is 100% down to NS2 and there's not even the slightest chance that your components are degrading in performance after so many years?
Framerate-wise, anything like that would be almost imperceptible.
Can we drop the overly dramatic "UWE doesn't care!" tone?
I never said we don't care about about performance.
I never said it's not a priority.
I never said we don't have time.
So allow me to reiterate what I said for clarity's sake:
"There are other priorities in the works right now. That being said, increasing the send rate has been a target and a desire for the team for a long while now... it's just a lot of work, that's competing with low hanging fruit priorities."
There's your answer, Rysk. Try to consider the priorities chosen from the point of view of UWE: they could drop everything right now and spend potentially 4-6 months rewriting swaths of the engine and continue to lose players and for whatever that happens to gain them in the end, or they could spend a few weeks and create something that makes the game more accessible, get's people into balanced games faster, and overall improves retention. It's a risk vs reward assessment. If there are improvements that can have a bigger impact and are less work, they will be chosen first. That doesn't mean server performance is not a priority, and it doesn't mean it's not something that everyone is painfully aware of. It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development. (And btw.. that answer was in regards to server performance, not client, idk how we got onto that topic as it's not even relevant to the thread. Client performance is continuous and hasn't really ceased, ever, to include upcoming patches.)
Also, I was away from the forums for 3 days. And for someone who works 3 jobs, I hardly consider that "hibernation".
If you ever have something that's time sensitive, feel free to PM me in our discord server. Otherwise, I'll get to the forums when I can.
1) I mentioned that warping solution previously in these forums, and it might get revisited at some point, but for now its shelved for the reason I stated.
2) There are multiple tutorials available, but since there's only a soft requirement to complete them (20+ hours in rookie servers OR tutorial) many still don't. There's not much we can do here.. we've provided the tools, the environment, and the education.. but outright forcing players will result in friction and possibly loss of players. Hell, we've even had many complaints that our current soft requirement is stupid and people have said they've uninstalled as a result. Short of a complete redesign/new simpler game there's not much more we can do.. but... that brings me to the next point
3) You cannot complain that players have no idea what they're doing, and then simultaneously advocate for the opposite of simplifying and making the game more accessible. Don't you see that its the intricate nature of the game that makes said players not intuitively know what to do? Retroactively making the game accessible has been a goal of UWE's for a long time now for that exact reason. You either support continued accessibility work or you need to be okay with fewer players, and players that know far less, since they are inversely related.
Lastly, I'm sensing a lot of angst in here.. and while I really want to continue to answer questions, I'm not willing to engage with people who continue to be derisively sarcastic, belittling or demanding.
I'm not a community manager and I take my own time to field your questions. I realize some of you still aren't happy with the state of things despite the progress over the years, but I'd ask you to try to remain reasonable and friendly instead of demanding and quasi disgruntled. It's the only way we can continue an open line of discussion, if that's what you're actually interested in.
Yeah. It's not necessarily just NS2 in isolation - it could be the result of how NS2 interacts with newer drivers, for example - but it's absolutely not hardware "degradation" for the reasons already mentioned by others. Also, my performance in other games hasn't suffered similarly (it's actually improved with fresher drivers in most cases; I don't really play many new titles so I'm not exactly pushing the envelope). I'm not trying to shit on UWE here. I said it wasn't game-breaking, and I'm enjoying the game just fine. What I am saying is that I can relate to others who've also seen their framerates decrease. I can't attribute it to a specific patch or mod because I took a hiatus from the game and didn't chart performance from patch to patch.
That's the problem, the priorities compared to what the community wants, and what the developer teams of UWE is miles off from each other. It should've been quite clear at this stage of what's the problem with the current state of the game is. The performance is just too bad, it's nowhere near any other game out there. Even playing Early Access games in Unreal Engine 4 for example is 100 miles better. I know it's wrong to compare the engines, but they've seemed to atleast known that the whole network code into an engine is quite vital/crucial.. It's just important.
You almost make it sound like those 4-6 months are a "waste of time".. Well, let me give you some info.. That "waste" of time could actually bring you back a bigger audience, people would stop complain about rubberbanding perhaps.
So yeah, for atleast for a chance try and do something to see the bigger picture of the game... instead of doing small changes all the time just to maintain the last 200-300 "active" players.
I know UWE doesnt have a CM, that was the biggest issue UWE had aswell, nobody listened to the community and finally you lost all your players up to this point.
And this is both for Client and Server performance.
When you say this aswell is what concerns me, clearly you havent read the forums at all, or any of the devs over at UWE " It means your personal opinion of what should be the top priority right now is not in line with what UWE needs to justify continued development"
For several years, the people have been bashing on the performance, rubberbanding ... But somehow you manage to say it's only me? From what i've seen, it's been the biggest topic for last 5 years.
The problem is that you've had moderators on this forum who's been doing posts like "Pm sent - mod" and later on closed those kinds of threads.. Cause having general discussions about the game where people are pointing on the issues is a bad thing.
It's just in the end people have finally given up, cause it's clear that it's NOT a priority over at UWE dev team... Meanwhile the community says otherwise..
I guess this will be my final post concerning the performance.. Take it how you want, but keep on going with the current path how the game is going you're not gonna gain any kind of player retention compared to how it could be. Cause "risks" is apperently a bad thing to take...
UWE is a business first, and then a group of people who created the game and want us to have fun. They gotta feed themselves too, you know...
Also, this is a technical issue, of which we have no insight. Have any of you seen a single line of C++ code of the Spark engine? Do you guys have any idea how to integrate a LUA JIT?
Balancing is another question, community input should be weighed more heavily in that regard.
Yes, I'm also frustrated by homing shots around corners, unregistered bites and fps drops in late-game, but I'd much rather see 64bit implemented first, so the modders have more space to operate in.
PS: Remove the HP bars please! It's been some time since anyone complained about it :P
My last computer that I built in 2009 ran NS2 at 45-80 FPS @ 1920x1080 (its at my in-laws and I still play on it). I saved up and bought my current computer that I built in 2015 plays 120-200 FPS at 1440p.
what else can we expect? It's amazing enough that UWE/Dev team is working on an "old" game with 200 players... If you demand better performance buy a better PC.
I just felt like the hit reg needed some lovin and wanted to bring attention to it so UWE could start planning around addressing it. I myself don't want UWE to turn on a dime here, because there's a lot about NS2 I do enjoy.
Also, I can't really say about the "drop" in performance people have been observing, but let me tell you something I myself have observed, for the better. And that is, playability on Linux has DRASTICALLY improved compared to when NS2 "shipped" (as in release).
If memory servers me right, NS2 shipped without Linux support, and UWE said "we'll do it, please bear with us", or something to that effect. A bunch of months later they rolled out Linux support for the client and at some point the server (I think the server came later? I can't recall for sure). My experience with the Linux client was rather bad. I would get into the game, and just a black screen, and never be able to actually interact with anything because all I see is black, not even a mouse pointer.
So I let it sit, checked back a few months later, same thing. Checked a few months later, oh, I can see stuff now, get into game, my frame rate is pretty unplayable, like seriously. So I let it sit some more.
A while later, like a year or two, I try again, and my framerate is quite a bit better, now playable, but loading times sucked.
This went on for a bit, checking in every now and then...
Recently I loaded NS2 back up, and it's like... worlds better than it's ever been for me on Linux. Not only does the game start in a good amount of time, loading maps is a good pace (it would be nice to be faster, but I'm okay with how it is now). But the big thing that I noticed recently, was the way it handled texture streaming. I load into a game, get low res textures, and the higher res ones stream in as I keep playing.
This one texture streaming function alone has made my enjoyment of the game worlds better. I get into each map/game way faster, and my framerate is good before and after the textures load, so I can just start playing. There's a few weird bugs that I can't quite explain just yet that I get in erratic ways, but they are so few and far between that I just let it slide (until I can get enough info to report/reproduce them).
As it stands, for me, the hit reg I would like to see improved. But I greatly appreciate so many of the other improvements that have come along the way. Many other dev teams would have scrapped Linux support forever and a half ago. But UWE has stuck by it, and I for one, really appreciate it! I've dumped a bunch of seed money (check my icon) at various points of the dev cycle, and I'm glad it has not gone to waste. The way match making works now, is really awesome, and I'm seriously considering spinning up a server in my area to help foster bringing people back into the game.
I think what would be worthwhile, is to continue the discussion of hit reg, but maybe we should do our best to keep it objective. Clearly, hit reg is an important element to people of the FPS' genre, so having productive feedback between players and the UWE crew, I hope that can turn into positive results. But these things take time, and I feel both camps aren't quite taking each other's interests to heart. I know UWE cares, but some of the things said here has made me feel like the value in hit reg isn't quite received by the UWE peeps. But at the same time, there's some people in this thread that are taking it a bit too far. UWE didn't kill your cat, so maybe we should just change the way things are presented hey?
Now I feel like I'm rambling, sorry about that.
My performance went down after Win10, but didn't really notice it, and after upgrading to a RX480, while getting a massive improvement, it wasn't the improvement expected.
Turns out from user benchmark, while most of my system is great, my Gen1 SSD is slowing down and is actually performing in the 21st percentile compared to all other users benching with my drive. That's a lot of wear. In fact it is so much wear that I am losing performance in everything, but especially games.
I only bought the thing in 2014, so it's not even that old, but it's degredation is having a massive effect on my system.
This is quite unlike RAM/VRAM and cpu's (motherboards), they only get wear and tear if you overclock them and they run at higher temperatures due to higher voltages, if your cooling isn't sufficient. If they are running within specs (or OC within acceptable temps), there wont be any issues.
Also it's come to my attention that with older (no longer supported) hardware comes the annoying bit that the newest drivers actually have the potential of being the main cause for performance degredation. Rolling back your drivers can in those cases help. For 3D cards the update program usually suggests to run an older version with more optimized support
Not mentioning dust inside the case or an older windows installation starting to bog down (win7 and higher seem to be coping a lot better than Vista and before...)
I just wanted to demonstrate that not understanding the reasons behind the priorities and goals of the dev team is the source of your frustration and misunderstanding.
Do you honestly believe that if NS2 was able to have a sendrate of 60 tomorrow, that suddenly player retention would improve? That alllll the other complaints from those negative steam reviews would vanish, like imbalanced rounds or steep learning curve? Do you think that newer players would even recognize the improvement after months of our efforts? Would UWE practically see increased revenue from it?
This is the issue: It's a priority, but it's not the top priority, and that upsets you because you feel it's what is best for the product - and while I obviously fullheartedly agree that it's needed, it's an epic amount of work to implement, with potentially not much reward in a business sense, at least compared to the other top priorities. So this is the impasse that we are in, that stems from a lack of a development point of view. It's easy to say "fix it" without understanding the risks involved for a part time team - which btw the risk that you feel is so easy to dismiss is the risk of not having NS2 be developed further - so the risks need to be weighed very carefully by the team. But please don't assume the forums aren't read or the issues aren't thoroughly understood... If there's any failure involved on the development side it's clearly communicating the reasons behind the priorities, which I hope I've at least made a dent in, with my comments?
As for client side performance, it's continual.
The differences from launch to today are vast, with big steps like VMs to LuaJIT, to dozens of small wins with 10% here, 20% there throughout updates over the years.
I am going to receive flak for this, but the fact that Depara is complaining about only getting 150-200 fps makes me smile a bit... since the bar has obviously been raised over the years. It means that while clientside has improved, the expectations have shifted with it. Which is fine. I much rather that be the case than the dire post launch complaints of only getting 40-60 fps. His very familiar complaint of "i have 140 fps but it feels like 80" is being looked into as I write this.
So like I said, clientside performance is continual.. we're even testing out a 10% improvement on CPU costs this week internally, with even much more exciting stuff coming down the line later that I wish I could talk about. (I won't until its at least tested, so I don't create false hopes)
It's the server performance side that is far far more involved, sadly.
Hope that at least provides some answers and transparency for you, even if you disagree with the decisions.
@BloodyIron
Again.. I don't think you mean Hitreg.. I think you mean the scenario of dying around a corner, which is due to sendrate + interpolation server settings.
IF you did mean hitreg, that is the upmost importance to us, so please collect some video of it using @Nordic 's thorough How To Objectively Test Hitreg method.
One thing I think I can tell you is sometimes it feels like incoming damage from rifles to a skulk "bursts", in such a way that it makes it nigh impossible to react to. What I mean by this is, say I'm chowing down on a stupid marine power node (stupid marines), my ass sticks out and marine shoots me. I try to get away, but by the time my position in the world has moved, I have already received more than 50% of my life in damage, and a "split second" later I receive the rest of the damage. This is kind of hard to refine as an explanation, but the issue in this example I have is the bursty-ness nature of the incoming damage makes it seem almost impossible to react to and retreat from. I've tried to pay attention to the pings, and this does happen at times when I have 60-80ms ping to the server, and the stinky marine that killed me has a similar ping too. It does indeed happen with higher pings too, but it does seem to with lower pings.
Now, I am testing more and more, and it _MAY_ be related to server performance too, but the servers I play on report as "happy face" performance. I don't have enough samples to be confident of this theory just yet, but it may be that I see this more on some servers than others...
On another, kinda related note. I do think I am getting "frame latency" of sorts. While I'm getting 60-80FPS or so, it doesn't feel smooth at times, sometimes I get jittering, I think. I'm on a 120hz monitor, so I can tell the difference between 60fps and 120fps, but that's not quite what I'm experiencing here. I'm on a 960 GTX, so I don't think this is related to the AMD frame latency of yester years. Is there any chance we could see Vulkan for NS2?
I'm sorry, what? That's my drive right now...
PM me a technical explanation plz?
If you join a server and see a popup window telling you that network rates are different from default, *that's* a server that is running modified rates. FYI, you can use the console (~ key) and type perfmon three times to see what rates are being used.
I know it's not a popular thing to say to a linux user, but the best performance regarding input delay will be from using Windows, DX9, Fullscreen, raw input on, AO off.
NS2 would benefit greatly from Vulkan, especially for linux users