Bohemia Interactive's ambitions are always set too high.

[bad benson 1733]

I understand that there is a lot of stuff going on in Arma, but players with reasonably powerful machines should be able to get 60 FPS on the lowest settings.

i think that's not going to happen and at this point an unrealistic expectation. not even arma 2 runs like that. arma 3 being arma 2 with more stuff it has to do. for example physics eventhough barely implemented. i don't even expect that from BI anymore. i think it's simply too hard for them to do. there are no clever new techniques in there. just more for your system to deal with than before. what i want is "playable". because you know it's a game. i wanna play it.

my measurements show that the latest dedi exe (no idea if it's standard in stable yet) gave me an increase of 5 FPS. it's something. but considering i'm running the game on all low (in MP) it's hopyfully the very first little step since for me personally it means ugly ass game, with 20 FPS max dipping down a lot, combined with very violent, choppy stutter (MP!). aka unplayable. if you get 30-40 FPS at best on a system that runs other games waaay beyond that, good for you. consider me jealous. it just shows what's wrong here even more. and my specs aren't even that bad.

i wouldn't repeat and elaborate all this whenever i can this, if it wasn't a real problem and, if there weren't so many people denying it. once it is fixed you won't see me posting on these forsaken forums anymore at all. my post count went up since alpha because of the game's state not because i find enjoyment in "whining".

[antoineflemming 14]

Ok, saw this comment somewhere (said I was taking a break, but I need to know if there is any truth to this), that the people who were originally instrumental in creating the RV engine have long left the company, and that there isn't enough documentation for the current developers to know everything about this engine. Can someone tell me if there's any truth to this? If this true, is this an impediment to BI's developers working on fixing the issues with the engine? (Dwarden, you don't need to respond as I already know your response)

[marcai 1]

It's unfortunate that anyone on this thread who has any criticism of BI, RV, or Arma 3 is instantly labeled a hater, a complainer, a "fixmoaner", or a whiner whose only argument is "omfg my process explorer don't report xyz GB of memory used by the game process, baaaad coding ...", by both some within the community as well as from the dev team. That is what is unfortunate here.

I don't think anyone hasn't complained in some capacity. In my post I even mentioned that my friends can't even play the game due to their framerates being unacceptably low. The forums are full of people who have posted complaints without being chased off and called names, so I respectfully disagree with your statement. Dwarden posted once in what was clearly exasperation for people trying to explain what they know better, the old phrase "teaching your grandma to suck eggs" comes to mind. I'm sure they're quite aware of their limitations and are working on how to improve it in a way that fits their budget and time constraints. Nobody's going to blame you for being upset about poor performance, as we're all suffering from it, but I can tell you now that it's far easier to talk to and make progress with reasoned and respectible discussion than taking personal offence to posts and slinging them back. Be the bigger man, make valid and reasoned responses as opposed to knee-jerk posts whilst angry.

Posts like the quoted one (and admittedly my own right now) are totally unproductive for the thread and really don't serve much purpose, so I'll try to refrain in future. Just seems a shame to waste good forum space on back-and-forth trades between individual members is all.

...people who were originally instrumental in creating the RV engine have long left the company, and that there isn't enough documentation for the current developers to know everything about this engine.

Interesting if true. I wonder if there's much chance of BIS letting the community moders have a gander at it? I mean whilst letting anyone edit the currently used code would be daft, I'm sure there are a lot of people who'd happily see if they can notice any bits of unoptimised code which has been overlooked prior to this? A few eyes doing the work for free (I'm sure some would do a small bit of code-reading for the love of the game as opposed to profit) pointing out things that could change might even get around the legal implications, no?

[antoineflemming 14]

I don't think anyone hasn't complained in some capacity. In my post I even mentioned that my friends can't even play the game due to their framerates being unacceptably low. The forums are full of people who have posted complaints without being chased off and called names, so I respectfully disagree with your statement. Dwarden posted once in what was clearly exasperation for people trying to explain what they know better, the old phrase "teaching your grandma to suck eggs" comes to mind. I'm sure they're quite aware of their limitations and are working on how to improve it in a way that fits their budget and time constraints. Nobody's going to blame you for being upset about poor performance, as we're all suffering from it, but I can tell you now that it's far easier to talk to and make progress with reasoned and respectible discussion than taking personal offence to posts and slinging them back. Be the bigger man, make valid and reasoned responses as opposed to knee-jerk posts whilst angry.

Posts like the quoted one (and admittedly my own right now) are totally unproductive for the thread and really don't serve much purpose, so I'll try to refrain in future. Just seems a shame to waste good forum space on back-and-forth trades between individual members is all.

Well my question still stands, and still has not been answered. Are you guys trying to fix your engine limitations or are you simply trying to work around them (the recent OPREP mentioned new technologies needed to fully realize proper soldier protection and such)? Yes, there are limitations in the engine. There are issues with it. So there is no need to question "limitation" as if it's odd to fathom RV having limitations. I just wanted, and still want, to know whether the current approach to these issues is the address the underlying issue, or if it's to try to work around any limitations as much as possible. That's all I was asking. Still waiting on an answer. And this time I hope that my question isn't generalized to "omfg my process explorer don't report xyz GB of memory used by the game process, baaaad coding ..." Oh, and it's posts like the underlined that are unproductive for and contribute nothing to the thread.

[pathetic_berserker 4]

First of all, he posted his specs in the fourth line of his post and it's fairly top of the line in most respects. I think his point was that if he can't run the game well at max or near max settings with that hardware, what kind of hardware would he need?

Second of all, yeah, calling him a noob is pretty much not helpful at all.

Lastly, perhaps you would like to enlighten us as to why you think a segment of the gaming industry targets 60 FPS as a benchmark for good performance?

Well if he really had been a fan since OFP days he would know that the entire series has always scaled a generation of computers ahead. So his point is in the title, a complaint pure and simple.

I have SandyE, GTX690, and have everything on SSD but I'm still not silly enough to expect to play with every thing maxed @10k view distance (for me triple head don't help much either). The range in settings made available to us is so individuals can tweak to personal tastes. I can certainly get a good playable experience, unique to ArmA, and better than any other FPS type game I can think of. And I even had the game running on an old Q9650, GTX285 though by my standards only just. No matter how much he spent on his machine its still only half the equation without settings.

As for the 60fps issue , bottom line it was self imposed, for all the reasons already stated but mostly as I saw it, because it seemed to be a great marketing ploy. It also seemed the optimal level where spikes in frame rates on console harware became mostly unnoticeable. Reality is the general population don't in fact see the difference over 25 - 30 frames. But they may notice the spikes. Hence animation and video production got away with these sort of figures for decades, because it was consistent. But when gaming the hardware has to deal with the unpredictability of human movement and the movement is often extreme. So the gaming problem is a bit different to simply running an animation. But if you can get your gaming hardware to out put at say 40-50 frames with really high consistency the issues fade away. There have been some really interesting discussion panels involving industry leaders just this last year that have touched on the subject and they seem to concede that the 60fps benchmark (which was set before it was achieved) may be becoming a bit limiting. Some implying that they would love to throw the concept away and make the games they want to make and hopefully the hardware will catch up. As DM said most games achieve the 60fps mark by reducing the unpredictability, and stripping away the substance. Hence the industry has painted itself into a corner and we still get the ever present corridors (widening slightly) broken up with cut scenes. Lucky for us BIS really do make the games they want to make.

If the OP had taken the time to find out about what the settings do and the point where his settings provide the smoothest ride, maybe even resort to a frame lock, then he probably wouldn't feel able to reasonably complain that the game has failed in some way. Sure we will always want more and I appreciate the passion people have on the subject but I think its a bit unreasonable to imply that BIS is not already approaching the issue with the same passion.

Sometimes we just have to sit back and weigh up the difference between scope of gameplay and eye candy.

Edited February 4, 2014 by Pathetic_Berserker

[calin_banc 19]

the entire series has always scaled a generation of computers ahead.

Wrong, it's designed with an old, inefficient type of thinking and doing things. THE GAME DOES NOT USE THE AVAILABLE RESOURCES!. That's the issue, the problem is not the bad performance on weak hardware, but rather about very bad performance on good hardware due to the way it's doing stuff!

[bad benson 1733]

I have SandyE, GTX690, and have everything on SSD but I'm still not silly enough to expect to play with every thing maxed @10k view distance

and therein lies the problem. the game does not run great on all low settings...especially not in MP. i posted my specs before and i get max 20 FPS (max, so you know what that means when shit goes down in a town) in MP on all low WITH new [PERF] dedi. without i get 15. yea it's easy to just say BI build games for the future. sadly it's not true. it's just a flattering way to put that the game was never optimized. i'm not saying "just do X and it will be perfect" but let's already stop with the old arguments. those where already problematic in arma 2 but now it's pretty clear that it's not that simple.

arma 2 runs "good" now because BI kept fixing it over years since they don't seem to be able to release a working game. it's not because now everyone has a super computer by standards of arma 2 release time. and we all know what "good" means in the arma context. you might be talking about 2050 when you say future but i doubt that :p

at least that implies that arma 3 might get there too some day.

i really don't get this denial crap. complain about trolls and wannabe devs all you want. i'm fine with that. but don't give me that "arma is just ahead of its time" shit. i've been with this series since the fucking ofp demo. that shit don't work on me. i know the series.

EDIT: infact after playing a bit yesterday i'm sure now that the buildings are the problem when it comes to hardcore stutter. when me and friends drove in a car through the countryside i got stutter even on small "settlements" (3 houses) next to the road in the middle of nowhere. that proves my suspicion i had in early alpha about the stutter. something is broken there.

Edited February 4, 2014 by Bad Benson

[PuFu 4600]

the entire series has always scaled a generation of computers ahead.

All generation that will be coming out are gonna be multithread CPUs (and or possible CPUs). There is no chance to push more GHz on the same silicon chip anyways. So, given the fact that quad cores have been around since A2 release, and A3 still cannot scale correctly on multithreaded rigs, how could run better of future components (which all will push for lower GHz, muti logical cores and lower power consumption)?

I've done the test before, on a dual xeon x5660 (2x 6/12 cores/threads) workstation 32gb ram, which i have fitted my own 6890 ati card in. It run worse than my i7 970 or my i7 2600k, although it has over 2x calc power than my 6core i7...You might say the cpu is workstation CPU and not made for games. While true, it does have more horsepower than your average gaming pc, although i am certain that in a couple of years consumer CPUs will come close if not surpass the 2xXeon older configurations. But technology goes that way, and not towards 10Ghz per CPU...

[roshnak 41]

Well if he really had been a fan since OFP days he would know that the entire series has always scaled a generation of computers ahead.

People have already called you out on this, but I figure I'll jump in and say this is just not true. If anything, this series is being made for an obsolete generation of hardware. You might say that it scales to a generation of computers behind.

As for the 60fps issue...

None of this is true. Please refer to my post on framerate in video games. It even has links to sources.

[chrisb 196]

Maybe its using a lot of PP that is causing headaches, I don't use PP as I don't like blur effect in games. Perhaps its worth looking at PP settings. A3 has much more in the way of tweaking these settings than A2 had (I use none in that catagory), so perhaps there may be an issue there. However some interesting articles on Wiki regards Motion Blur, FPS, PP etc. Blur being a cause of headaches, possibly moreso than rates as some games are locked to 30fps (console).

I am no expert at all in this, but its interesting reading the articles on it, well worth tweaking around a little.;)

But there again we're all going off topic, I don't think BI's ambitions were to give anyone a headache playing the game, well not a real headache, certainly they give many a virtual headache, via simple frustration.:p

[nomisum 129]

is actually anyone using PP blur? its annoying as hell since it was introduced and the first thing i turn off in every new arma game.

anyways, regarding 60 fps - i really hope the game will be optimised significantly, as oculus devs stated its recommended for their piece of awesomeness. in appr 1 year we will see, fingers crossed.

Sent from mobile

[pathetic_berserker 4]

[/color]

and therein lies the problem. the game does not run great on all low settings...especially not in MP. .................

Hmm I suppose if it appears I'm taking sides here its probably because the topic is on a problem I don't have short of MP (sometimes). I cant deny things could always be better, but I have a problem with the start of this thread and its premise. The fact that it has lead to a conversation already cloned many times over only highlights the divergent views on the issue.

And I think you and Pufu and roshnak could be correct in saying there is a flaw in my 'future' argument, though I still find by comparison to many other games that the net effect of what scaling ArmA can utilise on my hex core is a positive one. However the actual reasons for RVs scaling as a general precedent has probably changed over the decade of development of the RV engine. As BIS cant necessarily predict future hardware with absolute certainty. But as a 'fan' the OP should at least be aware of the precedent.

None of this is true. Please refer to my post on framerate in video games. It even has links to sources.

I may have been a bit of a generalist but nothing I said disputes what you linked. I'm aware of some of the science, and some of the psychology, the issue of game frames I think should stay within the realm of what we can do to effectively create the illusion required.

[bad benson 1733]

However the actual reasons for RVs scaling as a general precedent has probably changed over the decade of development of the RV engine. As BIS cant necessarily predict future hardware with absolute certainty.

sorry but to the thought that BI are trying to predict and develope for future hardware is absurd. it's just a clever excuse which "in the future" always turns out to be bullshit when the future hardware that BI supposedly is developing for is actually available.

[froggyluv 2136]

sorry but to the thought that BI are trying to predict and develope for future hardware is absurd. it's just a clever excuse which "in the future" always turns out to be bullshit when the future hardware that BI supposedly is developing for is actually available.

Not really, OFP plays at 160 FPS with my current build ;)

[pathetic_berserker 4]

sorry but to the thought that BI are trying to predict and develope for future hardware is absurd. it's just a clever excuse which "in the future" always turns out to be bullshit when the future hardware that BI supposedly is developing for is actually available.

Tat's what I said. BIS CANT do it.

But from what I've heard from the horses mouth its not uncommon for devs to try and allow for wiggle room based on the direction of current tech. Just sometimes that direction doesn't take hold.

Edited February 5, 2014 by Pathetic_Berserker

[mamasan8 11]

OK, so Arma 3 is made for future computers...why doesn't it then utilize more than 1-2 cores when Crysis 3 (older game, ironically) utilizes all my 8 cores. I'm seeing 70% usage across all cores.

Maybe BI should use Cryengine instead. Leave the engine to someone who knows their shit.

OFP and 160 FPS...you could say the same thing about Tetris. It was made for future computers. Because I bet you get many more FPS now in Tetris than when it was released.

[NeuroFunker 11]

^nobody should be bothered, replying to "arma iz baad enjine, cryrngine iz best" spam.

[calin_banc 19]

Tat's what I said. BIS CANT do it.

But from what I've heard from the horses mouth its not uncommon for devs to try and allow for wiggle room based on the direction of current tech. Just sometimes that direction doesn't take hold.

The direction for multicore, multi-threaded environment is old even in regular PC gaming rigs, like the time AMD introduced the 1st dual core. That was the direction. They don't need to look very further into the future to see that.

Why didn't they use for instance, the power of the GPU to accelerate large numbers of AI? I'll post this again -

. AMD did that demo with thousands of AIs when, before ArmA 2 hit the street, 2008? That would have been for future hardware indeed. This would have been for future hardware as well and

and

. That would have been thinking into the future, that would have been an "ok" from the community if the game actually used what we gave it, no problem the performance wasn't stellar now.

BI just patched the engine, made some nice improvements, but is still seriously lacking in performance when you go beyond CoD or BF play grounds.

[nomisum 129]

I doubt they didnt know what would have been good core enhancements of the engine. i believe it comes down to manpower and skill. even if they tried to bring fresh blood in, its not like you find good engine devs on every corner. hopefully they get their priorities right next time, search more aggressively and have the luck to find genius devs who want to pimp this monster of a lifted engine to really groundbraking levels.

Sent from mobile

[Polygon 11]

Main problems with RV is that there's only 1 expert of it active, some engine parts are undocumented (a dagger for productivity and further research into optimization, I guess), some ancient parts hidden deep and requiring drastic changes, so on.

If BIS had an independent tech department, making drastic changes to the core would be a viable solution. In this case, BIS should reconsider selling licenses of RV to game devs.

When it comes to programmers, it's not about genius or motivation only - salary is important, too. And to nobody's surprise, it's rather low in games industry.

@froggyluv: By OFP you mean Arma CWA, right?

[Varanon 892]

Tat's what I said. BIS CANT do it..

Multicore is hardly "future" these days. As such, you fail to make a point.

Sent from my Xperia Z1

[bad benson 1733]

Tat's what I said. BIS CANT do it.

But from what I've heard from the horses mouth its not uncommon for devs to try and allow for wiggle room based on the direction of current tech. Just sometimes that direction doesn't take hold.

the part you quoted doesn't say "BIS CANT do it.". so no that's not what you said.

you are saying exactly what i'm saying is bs. that they are trying to predict and THAT is the reason arma always runs badly on release. which is, again, bs.

there's a difference between anticipating tech that will be released soon and building an archtecture to utilize it and simply making a very demanding game that runs NOT perfect on new PCs 5 years later (arma 2). the fact of the matter is that arma simply demands a lot of CPU. raw CPU. always has. that has nothing to do with "tech" it simply means the game calculates a lot of things per frame in an expensive way. my PC is a future PC when it comes to arma 2 requirements. still it does not run at 60 FPS. infact it's very unstable when it comes to FPS. no i'm not saying i want or need that. it just shows that your theory is not based in reality.

the whole idea is simple based on joke that has been around in the community for ages. and now people try to recycle it as an argument. back in the day no one would have said this and been serious, now it's suddenly common knowledge and reality :rolleyes:

Edited February 5, 2014 by Bad Benson

[mamasan8 11]

Main problems with RV is that there's only 1 expert of it active, some engine parts are undocumented (a dagger for productivity and further research into optimization, I guess), some ancient parts hidden deep and requiring drastic changes, so on.

If BIS had an independent tech department, making drastic changes to the core would be a viable solution. In this case, BIS should reconsider selling licenses of RV to game devs.

When it comes to programmers, it's not about genius or motivation only - salary is important, too. And to nobody's surprise, it's rather low in games industry.

@froggyluv: By OFP you mean Arma CWA, right?

If I understand it correctly, salaries in Czech Republic are really low compared to the rest of the west (Poland etc not included, also low salaries).

I think the geniuses work in the USA or Britain, much better pay.

If you want to know the average salary for game devs, Gamasutra has the info.

http://www.gamasutra.com/view/news/189893/Industry_in_flux_What_we_learned_from_Game_Developers_2012_Salary_Survey.php

http://www.gamasutra.com/view/news/167355/

Engine devs would fall under Programming. Best paid guys, always has been, AFAIK.

My theory is that people working for BI are pretty much czechish people, thats who they can attract. While US/Britain-based companies attract everyone, the whole world.

If that is so, it is quite impressive. Sort of like SCS Software, a Polish company that has maybe 5-10 people working there and releases such a beauty as Euro Truck Sim 2.

Edited February 5, 2014 by mamasan8

[ProfTournesol 956]

@froggyluv: By OFP you mean Arma CWA, right?

There is no other genuine OFP my bad :)

[jiltedjock 10]

[/color]

Hmm I suppose if it appears I'm taking sides here its probably because the topic is on a problem I don't have short of MP (sometimes). I cant deny things could always be better, but I have a problem with the start of this thread and its premise. The fact that it has lead to a conversation already cloned many times over only highlights the divergent views on the issue.

And I think you and Pufu and roshnak could be correct in saying there is a flaw in my 'future' argument, though I still find by comparison to many other games that the net effect of what scaling ArmA can utilise on my hex core is a positive one. However the actual reasons for RVs scaling as a general precedent has probably changed over the decade of development of the RV engine. As BIS cant necessarily predict future hardware with absolute certainty. But as a 'fan' the OP should at least be aware of the precedent.

I may have been a bit of a generalist but nothing I said disputes what you linked. I'm aware of some of the science, and some of the psychology, the issue of game frames I think should stay within the realm of what we can do to effectively create the illusion required.

I am calling you out. What is the future hardware spec that Arma 2 was written to? Source please as well if you have it.