brightcandle
-
Content Count
236 -
Joined
-
Last visited
-
Medals
Posts posted by brightcandle
-
-
BI has had wind in the game since the beginning. But while wind is supported they didn't add all the things you need to work with it, turret adjustments on the scopes, manual estimation for wind, Kestrels for accurate wind, range tables for the scopes etc. There is a lot missing in the base game that I hoped that would have made it for marksmanship instead of it being about bipods and a few extra guns and scopes. The marksmanship DLC was a massively missed opportunity to improve the ballistics. But then we have ACE3 and so we can just use that. I don't care as much as others about the base game improving since I have my mods and a community in which we play with them. Adding all this complexity to the base game just makes it harder to get into.
-
I think Tanoa is the main reason to get Apex and it has issues. Its a pretty map statically but there is a lot of level of detail changes and other artefacts close to the user that are extremely obviously that look very poor in motion, they distract the eye from looking out into the forest for bad guys and that is very very bad. Then you have the sizeable reduction in performance the map has, somewhere up to 30% less fps when compared to Altis.
The AI doesn't play fair in the forests at all and it requires substantial modding to make it work sensibly or you just die to guys you can't see all the time. Its pretty buggy as well, we had a lot of issues launching the coop campaign reliably and one player got dropped a few times. You can see what BI is doing to try and balance the jungle combat, they are pitting the 4 of you up against 1 guy and its lots of singles and staged enemies to try and make the combat less deadly. But get a similar number of AI all at once and you are going to crumble due to how close everyone is. We have kind of fixed it with a custom mod we made a few months back which we hope to release soon and it did make the coop campaign a lot nicer once enabled but its still only really brought into a reasonable balance rather than outright fixing the issue of how the AI perceives visibility.
So while I think its worth it my community isn't requiring it yet because for a lot of people the FPS is unacceptable (15fps). I am not going to sing praises over something I (we) have known has quite a few issues that wont be fixed. Tanoa is a scenario where the AI has always behaved poorly and the forest is so thick its problematic. So while I would love a jungle scenario the reality is the engine isn't up to it, performance wise nor AI behaviour wise.
-
I'll keep buying military simulators and the better they are for building team missions the better. I don't laud BI as the amazing company others do, I have too many issues with the way the community is interacted with. So I'll follow the best military simulators and I only hope Arma gains numbers to make a real development company to see some sense to compete and make us a game that runs well. They don't deserve anyone's loyalty.
-
I seem to recall in the earlier versions of ArmA3Sync we found that mod lines (Launcher Options Run parameters) showed up in alphabetical order. But recently we have been having a problem with our modset and looking at it now some mods are appearing after what appears to be alphabetically sorted mods before them. This is an issue, like many communities we override existing mods features with updated settings and without an order being defined things wont work properly. Any clue what is going on/what changed? Any way to reset the order of the mods so they launch in a defined order again?
-
Apex is no operation arrowhead, its no where near as compelling as a package. Tanoa runs pretty poorly and being mostly jungle with non enterable buildings is a big terrain but a one trick pony in the area where Arma does the worst with its AI. The AI has a big advantage with spotting in the forests and in tanoa that advantage is amped up to 11, we have tried dynamically reducing their spotting speed and distance to as little as 0.01 skill but it still doesn't even the field they still spot us first and due to the short ranges the accuracy means they kill fast.
Combine that with some meh weapons where the animation quality isn't that fantastic and mods do more variety and better quality and futuristic vehicles we just don't care about and APEX as a release isn't going to be the big thing that every community adopts. Its definitely optional and so a light version might be a good idea if they want it played more. I for one am underwhelmed with it.
-
I have no doubt Tanoa is slower, up to 30% slower at times and its heavier on the CPU than Altis is. I did a comparison video a week back which shows when the game is CPU and GPU limited and also the difference in performance between the two terrains is quite dramatic:
-
I did some testing of tanoa v altis on very high and one of the things I note in the video is that my 2x 970 SLI is maxed out in the forests, but in the towns its very CPU limited to nearly half what the GPUs are capable of and in the open ground its somewhere in between but still CPU limited.
So your FPS in a particular scenario could be due to the CPU or the GPU and this game will hit on both depending on your circumstances. But really the way to look at this is the game has X potential based on the CPU you are using and then different scenarios use the GPU more or less up to that potential and occasionally your GPU might even be the problem. That max potential on a 6700k is around 120 fps in the best case and around 45 in the worst case in single player. In multiplayer its less than that as every player seems to add some CPU cost.
The video:
-
1
-
-
HI,
Just came back from Prague - what a wonderful city it is: hot girls and cheap beers - or was it the other way around :rolleyes:
I see there are some problem with the 1.56 update.
We'll take a look on it.
EDIT:
After some reading the errors with modules in Eden are BI side and happends in other mods as well - so just stick to the 2d editor if you planning on adding modules.
On the previous page you will see we are also having issues with JIP and respawn.
On JIP we are getting nothing but default gear and the script errors we listed are appearing in the RPTs.
On respawn we are loosing our primary and secondary weapons.
Just to confirm we are on version R18.
One thing to note is we have removed the house on start location because everyone gets stuck in it when we spawn. Shouldn't impact this change but we are letting you know we did it because its broken in the latest release and its one change we are maintaining on our end in the hope it gets fixed at some point otherwise. All we do is remove the house to solve the issue. Let me know if you want a diff to for that one.
-
This weekend we noticed that the AT4 CS version has zero backblast (when used with ace). I am assuming its a bug but maybe its a bad combination of mods, we run a lot of stuff.
-
You turned out to be right. Those 400 errors were being caused by bitdefender even when it was off. Needed to be uninstalled.
-
Try to disable Firewall and Antivirus stuff.
Try also to past the repository url on web browser and download the /.a3s/sync file.
There is no issue with the user accessing the sync file from the browser, only from Arma3sync itself.
I can't see how firewall or anti virus would be corrupting the request that goes up to the server, it doesn't sound like that could possibly be the cause with everything else working. Would not all requests from the web browser also fail? Does not make any sense why either of these two could have an impact. Turning them off showed no change and that wasn't a surprise to me.
-
I am running the web server as nginx, its a passworded repository and works for 10 people. But then an 11th came along today and he gets 400 errors returned by the server while syncing. The repository details are correct and its hitting the correct url to retrieve the .a3s/sync file but nginx is returning a 400 error as a response. We know they are using the exactly same version of Arma3sync (1.5.73) and the latest version of java (1.8.0_66) and as far as we can tell all else is equal.
Nginx is limited in what it can tell me but one thing I did determine is that the request size is different, everyone else has the same size of 224 bytes but the bad client is only 104 bytes. Now I assume the problem is that the request is being mangled somewhere. Running Arma3Sync in debug mode we see the error below and the line the error is coming from.
I am hoping a developer can tell at the very least tell us what is different on this guys computer compared to the rest of us or perhaps fix the bug.
-
In the latest Windows if you run the defrag program on an SSD it doesn't actually do it, I think it just "optimises" the drive which turns out is something to do with trim and not a lot else. Its harmless now, Windows defends you against the basic mistake nowadays and has done for quite a long period of time.
Besides its not like a defrag is going to break the SSD, its just unnecessary and wastes a reasonable amount of writes on the drive but its not going to magically break it.
-
BI is claiming it just so happens to coincidentally come along with a problem in Windows with a Font patch that is optional and a problem in Steam. Supposedly installing the latest Steam and the optional important update should fix the problem. I can't reproduce the issue after these two things have been installed but that doesn't mean its fixed because I wont be playing multiplayer until the weekend. Microsoft patch https://support.microsoft.com/en-us/kb/3079904.
-
I did a similar thing in Scala a while ago and I used ..../:playername/xml with text/xml as the mine type. So it definitely doesn't need a file it will take a reasonable URL, xml/ isn't an end point but I suspect it will accept it. As to the mime type I don't remember if it has to be text/xml, I used the right mime type but then play framework will use that automatically when you use an XML as the response. I wouldn't be surprised if BI doesn't even check it so I would try a hardcoded response and see if it works or not, I am about 90% confident it will.
-
1
-
-
Normally I don't have a performance problem on Cherno or the other Arma 2 maps (always have problems on Altis) but today it was in the high 30's. Every single person was commenting how much lower it was today. So yes we run a tonne of mods and its all custom missions but performance is definitely down in our missions, as much as -20 fps.
-
I've always been able to tell the difference between 60fps and anything lower than that, especially anything under 30fps even if it is constant and stable. Heck I can tell when a monitor is 120hz instead of 60hz, I would hope that other people can too. :rolleyes:Its extremely obvious to anyone that owns a 120hz monitor. Tests done on random people show that basically everyone can tell the difference with a short amount of exposure to it, see the linus test for example where he gets it right 100% on a test to see if he can tell the difference between 60 v 120.
60hz is far from the limit of the human eye. The current estimate based on tests against pilots and spotting silhouettes of planes showed we can do that with just a 1ms flash, lower than that and we can't tell it was there. So it looks like the point where we can't tell a screen from reality in terms of update rate would be around 1000 hz.
-
Hi everyone,I jump on this post instead to open a new one.
I didn't found any specific and clear answers to my questionment.
I have 2 GTX970 graphic cards, working in SLI. With the latest drivers.
When i activate only one, or both, i don't see any difference with the PFS...it turn around 50/60 FPS, no matter if i activate a card or two.
I wanted to know if this is normal ?
Thanks.
100% normal. The game is entirely limited by CPU performance in its simulation and the BI code for rendering on a single thread. There is no way to fix the situation, its been like this since release and it'll always be like this.
-
In Arma 2 lingor managed to run similar to the other maps despite its jungle setting. However that map had to severely reduce the amount of foilage so that it wasn't really a jungle but more a wood with some other foilage about and careful use of microterrain limited the visibility distance more than the actual foilage. I anticipate the same type of trade off in Arma 3 will be necessary otherwise performance will be unacceptable poor just as it is on Altis today around the big towns.
-
A simple leaked benchmark shows basically no single threaded benefit from Intel's upcoming Skylake in Cinebench. That probably means we are looking at zero progress on Arma performance as well.
-
This is about right for ultra settings without any tweaks on stratis in a little bird in places. Performance really is at those levels. I recommend everyone set ultra, don't tweak them at all and then use the editor and place a guy in a littlebird on stratis and on altis and go fly around and see the performance. You should also drop down and go walk around in the main towns as well. I have been saying this for a long time but you can get arma 3 under 30 fps on a blank altis map just by going into the towns. SP/MP and all the other performance impacts they can bring aren't necessary to bring this game to the unplayable point.
-
With an AMD it would need to be more like about 12Ghz based on the timings I saw, but its not just about overclocking its about literally getting 3x performance.
-
I would hold off getting a AMD for ArmA because it does not support them well. It was proven on there with evidence that ArmA 3 only makes use of four cores from the total of eight cores the FX 8350 has. But for most other things more cores = better (if the program or game supports it).The sentiment is correct but the "used core count" is just more complicated. The game can and will benefit from 12 cores, maybe more. However the performance is dominated by a single thread of execution and the little pieces that do run in parallel have only a marginal impact on performance in practice. Arma doesn't only use 4 cores, or 2 cores or whatever simplification there is out there, if anything its mostly single threaded but there is significant work done in parallel such that extra cores help but with ever diminishing returns due to Amdahl's law. Some like 20-30% of the frame can be run in parallel, which means with just a single core performance would be quite a bit worse, the second improves performance dramatically but the 4 cores help less and so on and so on. The ideal processor for Arma 3 is actually 1 super core that runs at 8Ghz like performance on a single threaded piece continuously and then also has at least 6 supporting cores that can do fast work but only in short bursts of 5 milliseconds or so before they overheat and they need to work for a total of about 20% of the time at most.
The situation is just more complicated but the problem with AMD's CPUs is just that any single core doesn't have the instruction throughput that Intel does on the Arma workload and hence the extra cores don't help it. But its not true to say the game only uses 4 cores, that isn't factually correct the situation is more complicated, not very complicated but more so.
-
There was a small improvement from the Sandbridge 2600k to the Haswell processors but hardly enough to make it worthwhile. Skylake is another big increase in GPU performance and no movement in CPU performance from Intel so I would hazard a guess its not going to bring much improvement again unless something amazing happens that we haven't yet heard about.
As to Zen based on the fact its a 12 cores (probably stretching that definition like the 8350 did with 8 cores) it may or may not help Arma 3. If they get single thread performance to be roughly the same as Intel's (unlikely considering the vast difference in resources of the two companies) then the extra cores will help the rendering process go a little faster and give the Zen an advantage. But just a 10% drop in single threaded performance will hand the advantage to the Intel CPU again.
Fact is the game just isn't written with todays or future processors in mind, thus it doesn't care much about the architecture changes coming and the general trend towards more cores. Its extremely unlikely we'll see a big change to that basic fact in Arma 3's cycle, DX12 is only going to reduce the rendering time in the API which right now is pretty minimal but it wont fix the serial thread issue over the rest of the games processing. I know people want hope but its more honest to just accept the reality that it'll always perform this way and no magic CPU is coming to save the day. The only time to show any hope is once BI themselves freely admit that the client has severe performance problems due to its design and commit to fixing that, until they do that (they haven't done yet) we aren't going to see any improvement.
Update 1.62 graphical issues
in ARMA 3 - GENERAL
Posted
This white speckle bug goes all the way back to Arma 2, its been in AMDs drivers since January 2012 at least.
The workaround is to disable ATOC in the graphical settings, its that setting which AMD cards have problems with. I raised a bug with AMD many many years ago first in Arma 2 and then again for Arma 3 but they haven't responded yet, so its not really a priority for them. At least you have a workaround, it ruins the AA in the place you need it most (trees) but it does at least get rid of the white speckles that utterly ruin the game.