Jump to content
Bohemia Interactive Forums

invisibull

Member
 View Profile  See their activity

  • Content Count

    94

  • Joined

  • Last visited

  • Medals

Posts posted by invisibull

  3. Posted

    Though there's been some improvement with this last hotfix, the FPS is still not what it was in v1.08. These benchmarks were done with absolute rigor, i.e., every condition and setting was reproduced as closely as possible for all three runs. Thanks for the hotfix BIS, hopefully you can quickly eclipse v1.08 with your next update.

    2014-01-23 23:29:46 - arma3 40v40 unmodded v1.08

    Frames: 3390 - Time: 60000ms - Avg: 56.500 - Min: 36 - Max: 62

    2014-01-24 12:06:05 - arma3 40v40 unmodded v1.10

    Frames: 3056 - Time: 60000ms - Avg: 50.933 - Min: 34 - Max: 62

    2014-01-29 13:43:08 - arma3 40v40 unmodded v1.10 hotfix

    Frames: 3176 - Time: 60000ms - Avg: 52.933 - Min: 36 - Max: 62

  5. Posted

    J - you're looking at the wrong file. It's the EOS readme.docx you'll find immediately inside the .rar you've d/l'd. An explanation for null = [["a1","a2"],[0,0],[2,1],[0,0],[0,0,0,0],[5,1,150,EAST]] call EOS_Spawn; can be found in that document. As far as using UPSMON or any other mod alongside EOS, you should experiment yourself and see what works best.

  6. Posted

    Shag, if you look in the bcombat_config.hpp, there are three sets of skill array groups. You'll notice the top set is the Arma 3 default settings:

    aimingAccuracy[] = {0,0,1,1};

    aimingSpeed[] = {0,0.5,1,1};

    aimingShake[] = {0,0,1,1};

    endurance[] = {0,0,1,1};

    spotDistance[] = {0,0.2,1,0.4};

    spotTime[] = {0,0,1,0.7};

    courage[] = {0,0,1,1};

    reloadSpeed[] = {0,0,1,1};

    commanding[] = {0,0,1,1};

    general[] = {0,0,1,1};

    Just simply apply the above settings to the bottom group as it's that bottom group which is uncommented so the one actually read by the game.

  10. Posted

    Hi Azroul - in v1.10 stable (the one released a few days ago) i was getting all sorts of irregularities with the ai. These included FPS diving down from even the smallest of firefights, and a situation where ai would run near me without engaging as if i wasn't there. None of these issues exist when I use v1.08 stable.

    I have thanked both Kronzky and Monsada already, but thank them again just the same for their work as well. :)

  11. Posted

    Thanks, Azroul, that did the trick. However, I'm noticing all kinds of FPS issues and ai not reacting appropriately to the enemy in v1.10. I've since reverted to 1.08 and UPSMON is just fine. It also seems to be working a lot better with bcombat than ever before. Thanks again for all your work. People like you make Arma what it is.

  13. Posted

    Conceptually, updates to the stable branch should be just that, changes which are tested for stability before being released. By releasing (forcing) changes that were known to cause issues in the dev branch into the stable branch, the line between the two has become blurred. Degrading performance in a finished product by force isn't something I'd expect in a so-called stable build. The anecdotal evidence both on the forum and in conversations i've had with users all points to their having been a major blunder with this update. I am sorry for those i've upset with my apparent lapse in judgment in where i chose to post my original comments. Perhaps Shagulon was right. I just hope my points won't be lost in the shuffle as I still think they are fair and valid.

  14. Posted

    I've tried both branches, neither is stable at this point. Both have become experimental. Does that clarify?

    ---------- Post added at 14:49 ---------- Previous post was at 14:44 ----------

    Try going back to the stable branch, and if unhappy, post your comments in the general discussion thread...

    Thanks, now try being a little less obvious. My comments are here and on the stable branch because they concern the fact that the DEV branch has now become interchangeable with the stable branch with this last update.

  15. Posted

    "Playable" is a meaningless subjective distinction. This is why we use FPS, CPU/GPU usage and other objective measures to establish actual performance.

    For instance, in a benchmark mission of my own making which is basically 40v40 ai on Altis, pre-new patch, i averaged 52 FPS over 3 minutes as compared with v 1.10 over the same period of time being 38 FPS. This is with the exact same settings and completely unmodded. The long and short of it is that there's a problem, and that's not the whole story. Not only has performance degraded considerably, but now vsync is no longer working properly (I have to alt-tab to get it to work). Also, a few of the mods I've come to rely on for somewhat passable AI behavior have been broken and though this is an understandable consequence of any upgrade, it also adds insult to injury when factoring in the aforementioned performance issues. I don't understand why an update, which BIS had to know was going to degrade performance, was foisted on us. Isn't that the entire purpose of the DEV Branch? I'm now stuck using an increasingly inferior product due to compatibility issues if I want to play on others' servers.

    It's just laughable that what has long appeared to be a top complaint, namely performance issues, was not only left unaddressed, but was actually exacerbated with no option to opt out.

    And for those who insist that performance isn't that bad, my hat's off to you all. But please don't think for a moment that it matters to me personally how happy your are with your performance as I watch mine degrade empirically.

    I'd suggest that BIS roll back the performance degrading AI tweaks and vsync debacle immediately while preserving any benign upgrades such as the campaign, etc. having no effect on performance. It's simply not fair to force DEV type experiments onto those who want a Stable version.

  17. Posted

    I think you're right, Invisibull. Are you using SLI by any chance? Just wondering whether it's a driver issue? If not, I guess we'll just have to wait for a hotfix etc. I've sent in a support ticket on this issue and will let you know what comes back.

    I've got a single Nvidia 780ti, and have actually rolled back the drivers by two increments (331.82 and 331.93) with no change whatsoever in performance or screen tearing. Judging by benchmark missions I use, I've probably lost about 10% of my FPS. Definitely a step back for me, though I have noticed some claiming noticeable increases.

  24. Posted

    Hi all, I'm hoping someone can help with what seems to be update related problems.

    After updating this morning and then running Arma3, I went into a mission i had been playing in the editor last night and immediately knew something was wrong. I had a band of screen tearing slightly above the halfway point on my screen and the FPS was all over the place, sometimes dropping down into the 20s and 30s. That same mission last night was at a rock solid 60 FPS every single time I ran it from start to finish and the screen tearing's been something I've seen on only a very few occasions, so something's changed drastically.

    I went through and disabled every single mod and script i was using and still found the same results. Other missions I've had a steady 60 FPS on are also now behaving in the same fashion. The screen tearing isn't there every single time I run a mission, but it is there a majority of the time. This is with vsync enabled or disabled. Oh, and vsync in my Nvidia settings doesn't seem to have any effect at all.

    I certainly hope there's an easy fix for this because my frustration level is rising quickly. Please find my specs in the spoiler.

    i7-4770@ 4.6Ghz - Nvidia GTX 780 Ti - Arma 3 Installed on Sandisk 256Gb Drive - 16GB Corsair DDR3 2133 Mhz

×