DUST Search - PC Lag: The Hunt for Cause and Solution

IAmDuncanIdaho II wrote:

Aeon Amadi wrote:

IAmDuncanIdaho II wrote:

Good stuff. Does this mean CCP don't know what the issues are? And are there two issues? Framerate and lag? Are they different? Is the centre of your diagram "Lag" just a generic name for adverse gameplay events that slows something down?

Not sure if you've got team / squad chat covered as a source of lag (in the form of network latency). And does the number of mercs with mics actually switched on affect this...and if so, does using push to talk reduce any lag?

Also, I would assume the teams from around the world problem manifests itself only when you're not on your "home" server. The guys who are wouldn't perceive much change?

PS. Congrats on passing your exams / coursework!

I'm sure that CCP is getting closer to finding out what it is but it's been an on-going investigation ever since Uprising was released six-seven months ago. I am considering framerate and lag to be one in the same as it's any real detriment toward the smooth performance of a match. There are a -lot- of variables to consider and anything we, the players, can do to narrow it down can do nothing but help find a solution.

I will definitely add team/squad chat concerns to the chart as it is something that needs to be tested - whether it's better or worse on team/squad chat compared to a user-made channel. Voice comms is definitely going to need several runs of testing to actually work out, I feel.

As far as world location - we just recently got out of a war with The Southern Legion who is based mostly in the Australia/New Zealand area (if I'm not mistaken) and while it did generally suck not getting the home server, we'd often hear reports of the other side experiencing lag as well. There is a difference in network and framerate lag, of course but again we're looking to have both of them ironed out.

PS. Thanks! I worked really hard on it ^_^

The Southern Legion is my old corp (great bunch), and they withdrew from PC (ages ago) due to lag issues. I'd say they've generally seen far more lag than others appear to have done. It would be great if you guys could get on board with testing stuff out, coz they are great guinea pigs for your tests and there's plenty of them ;-)

[edit]

Wow I wouldn't have thought of all the things you did in that apparently simple addition of the voice chat node to your diagram. Do you think that belongs under the network node though?


Author	Thread Statistics \| Show CCP posts - 0 post(s)
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4028	Posted - 2013.12.06 09:57:00 - [1] - Quote Now that I'm done with class (I passed, go me) I have a bit more time to refocus my efforts on testing again. That being said I decided to help tackle one of the larger issues in the game that just about everyone has experienced and would like the entire community to assist in that endeavor, if at all possible. Here's where we start. As you can see, I've categorized everything into color code with yellow and green being the general topics for each individual aspect that could be related to PC lag. Red color coding is a detrimental effect while blue is a beneficial one. The majority of these need even more extensive testing as we've only just begin to hard-core test this stuff out but I'm confident in what we have currently being at least somewhat accurate. I'll get on with it and present the first evidence toward our tests. Testing nanohive/drop-uplink spam. Recorded with potato. Unfortunately, this guy (myself) doesn't own a PVR or any method of recording PS3 gameplay - so I'm restricted to using external measures. This works out to some degree, however, as it helps reduce any interference in the recording and only shows what I was seeing on my television screen. 0:00 - 0:20 assistant ran along crosswalk, which we used as our control; facing the cluster of equipment. 0:20 - 0:33 same test but this time in the center of the cluster. Noticed a framerate decrease, though slight. 0:33 - 0:50 there was significantly better performance 0:50 - 1:30 testing looking toward the cluster and away from the cluster. Notable differences. 1:30 - 1:48 testing distance from cluster; nanohives disappear at this point 1:48 - 1:54 testing minimap disappearance; cluster disappears at this point 1:54 - 2:15 nothing of relevance 2:15 - 2:28 framerate decreases again as visual effects begin on screen 2:28 - 2:56 testing framerate using weapon fire-rate and magazine ammo readout (lower right) to gauge HUD differences when looking toward and away from cluster. So, that being said, this is just the beginning of the testing and I do intend to test more extensively when I have more resources. This test was somewhat skewed as we only had four people participating in the equipment spam and it was on a domination match (pub-match, obviously). Right now I'm trying to convince my corporation and our alliance to participate in makeshift "mass tests" in which I'm likely to pay people in ISK for participating in order to truly stress test all of this in a better situation; I.E a legitimate PC match, preferably inside an outpost (aka: 'city'). How can you help? - Post suggestions on testing criteria; things you think may be causing it or potential aid in reducing lag events. - Post suggestions or even videos of your own testing methods if you feel it's more practical. - Spread the word, let people know of what we (the community) are doing and invite them to assist. - Post your comments and feedback, we could always use the encouragement forums.dust514.com/default.aspx?g=posts&m= (Frames) 1544109 (Advertisement) 1556863 (Packs) 1570030
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4028	Posted - 2013.12.06 10:18:00 - [2] - Quote SponkSponkSponk wrote: We could help, except, you know. Haha, yeah - I considered that (hence the network category) but there'd be a loooooot of sweet talking to make that happen considering recent events >_>;;; forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4028	Posted - 2013.12.06 11:03:00 - [3] - Quote IAmDuncanIdaho II wrote: Good stuff. Does this mean CCP don't know what the issues are? And are there two issues? Framerate and lag? Are they different? Is the centre of your diagram "Lag" just a generic name for adverse gameplay events that slows something down? Not sure if you've got team / squad chat covered as a source of lag (in the form of network latency). And does the number of mercs with mics actually switched on affect this...and if so, does using push to talk reduce any lag? Also, I would assume the teams from around the world problem manifests itself only when you're not on your "home" server. The guys who are wouldn't perceive much change? PS. Congrats on passing your exams / coursework! I'm sure that CCP is getting closer to finding out what it is but it's been an on-going investigation ever since Uprising was released six-seven months ago. I am considering framerate and lag to be one in the same as it's any real detriment toward the smooth performance of a match. There are a -lot- of variables to consider and anything we, the players, can do to narrow it down can do nothing but help find a solution. I will definitely add team/squad chat concerns to the chart as it is something that needs to be tested - whether it's better or worse on team/squad chat compared to a user-made channel. Voice comms is definitely going to need several runs of testing to actually work out, I feel. As far as world location - we just recently got out of a war with The Southern Legion who is based mostly in the Australia/New Zealand area (if I'm not mistaken) and while it did generally suck not getting the home server, we'd often hear reports of the other side experiencing lag as well. There is a difference in network and framerate lag, of course but again we're looking to have both of them ironed out. PS. Thanks! I worked really hard on it ^_^ forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4028	Posted - 2013.12.06 11:15:00 - [4] - Quote Acturus Galaxy wrote: Allot of equipment placed around the supply depot seems to create lag in the new Gallente city map. I just finished a match that got laggy, sneaked into the compound and found 15-20 droplinks plus some nano hives dropped around the supply depot. Lag went away as I destroyed all the quipment and game was again playable. Yeah, definitely seems to have to do with the socket infrastructure - probably due to the mass of textures and static meshes involved. I'm arranging for a legitimate testing environment in the form of a fake PC match on a surface research lab to try and narrow all of it down but it won't be for a few days. Updated with new testing categories. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4029	Posted - 2013.12.06 12:28:00 - [5] - Quote IAmDuncanIdaho II wrote: Aeon Amadi wrote: IAmDuncanIdaho II wrote: Good stuff. Does this mean CCP don't know what the issues are? And are there two issues? Framerate and lag? Are they different? Is the centre of your diagram "Lag" just a generic name for adverse gameplay events that slows something down? Not sure if you've got team / squad chat covered as a source of lag (in the form of network latency). And does the number of mercs with mics actually switched on affect this...and if so, does using push to talk reduce any lag? Also, I would assume the teams from around the world problem manifests itself only when you're not on your "home" server. The guys who are wouldn't perceive much change? PS. Congrats on passing your exams / coursework! I'm sure that CCP is getting closer to finding out what it is but it's been an on-going investigation ever since Uprising was released six-seven months ago. I am considering framerate and lag to be one in the same as it's any real detriment toward the smooth performance of a match. There are a -lot- of variables to consider and anything we, the players, can do to narrow it down can do nothing but help find a solution. I will definitely add team/squad chat concerns to the chart as it is something that needs to be tested - whether it's better or worse on team/squad chat compared to a user-made channel. Voice comms is definitely going to need several runs of testing to actually work out, I feel. As far as world location - we just recently got out of a war with The Southern Legion who is based mostly in the Australia/New Zealand area (if I'm not mistaken) and while it did generally suck not getting the home server, we'd often hear reports of the other side experiencing lag as well. There is a difference in network and framerate lag, of course but again we're looking to have both of them ironed out. PS. Thanks! I worked really hard on it ^_^ The Southern Legion is my old corp (great bunch), and they withdrew from PC (ages ago) due to lag issues. I'd say they've generally seen far more lag than others appear to have done. It would be great if you guys could get on board with testing stuff out, coz they are great guinea pigs for your tests and there's plenty of them ;-) [edit] Wow I wouldn't have thought of all the things you did in that apparently simple addition of the voice chat node to your diagram. Do you think that belongs under the network node though? Not necessarily. Voice comms are through Vivox and logic dictates that Vivox, being a voice provider, probably wouldn't have anything to do with the Battleservers that the matches take place on. They're separate issues that symptomatic of a larger disease. While the same could be said for being listed in the sound channels category (in-game sounds and voice being separate issues as well) it's easier to categorize it that way. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4029	Posted - 2013.12.06 12:29:00 - [6] - Quote Your Absolut End wrote: Cant really say changing ingame graphics help with it, tried low and high on both and for some reason i even have the feeling high works better than low Right, I've fiddled with the settings a bit and saw literally no change in graphical quality what-so-ever; thing it might just be a feature that hasn't been implemented. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4035	Posted - 2013.12.06 12:54:00 - [7] - Quote Robert JD Niewiadomski wrote: I-Shayz-I wrote: Someone mentioned something about sound being the issue. For instance, the second you get far enough away to not hear the sound, the lag goes away. Also, try testing in a building vs out in the open. And would you mind to test if setting music/SFX levels in DUST client to 0 prevents the lag? The settings can be accessed from NeoComm menu while in-battle. Here is that "lag & sound" association thread by Jadd Hatchen: https://forums.dust514.com/default.aspx?g=posts&m=1560294#post1560294 It's all on the list - I just haven't updated the chart yet. I personally play with music set to zero but even before I made that change I rarely experienced anything hardcore in the lag department unless I'm coming up on an objective that is littered to hell and back with equipment; which usually gets cleared up after a few flux grenades or an EMP strike. Another thing we're going to be looking at is PS3 Model as that could also be a factor, as well as SSD and HDD implementation. There's a lot of stuff we're going to be covering and I've maybe a day to organize every little, minor thing before we move on to the actual testing phase. As I said in the IRC, thank god we're not having to look at the Xbox 360 as a factor as well xD forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4036	Posted - 2013.12.06 13:30:00 - [8] - Quote Spectral Clone wrote: I-Shayz-I wrote: Someone mentioned something about sound being the issue. For instance, the second you get far enough away to not hear the sound, the lag goes away. ^ This has to be tested. If this were the case than we would have experienced a continuous framerate dip in the video I presented during the third run along the crosswalk. The frame-rate improved when I looked away from the NH/DU cluster with the sound still playing through my headset. If you're far enough away for the sound to dissipate, it's likely you're far enough away that the visual effects have ceased to render. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4038	Posted - 2013.12.06 13:59:00 - [9] - Quote Mobius Wyvern wrote: On this topic, along with the automatic crash report system, could we get a manual means of sending game data back to CCP? Imagine how much easier testing like this would be if Aeon and his assistants could do all of this and then pack it all into a bug report via a button in the client. I realize the client can't log everything constantly due to hardware constraints and that just being a lot of wasted files, but what if you could turn on and then off again a log function inside the client for certain time periods? That way people having issues during a PC match could turn that on to log what's going wrong, and then send that back to you rather than having to do the trial-and-error approach we're forced to use right now. As said in Skype, CCP automatically collects this kind of data - they have to in order to backtrack on bugs that occur in pub-matches and the like. All they really need is information to help pinpoint which match it is, like what game-mode you were playing, the time it occurred and pictures/video. Once the battle is scheduled we'll be getting a hold of CCP and informing them when the match will take place so they know exactly what to look for ahead of time. I'd like CCP to be present for the test so they can see first-hand and maybe help us out with various factors but I understand that they're time is important in other areas and this isn't feasible or practical when they can simply look over the data log and go from there. Thank you for reminding me though, I need to write all of that down so I don't forget. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4038	Posted - 2013.12.06 14:05:00 - [10] - Quote Robert JD Niewiadomski wrote: Some games has "Report" option (Like in LBP upon pressing SELECT button). I bet CCP logs lot's of things happening in matches (player movements vectors, buttons pressed, etc). I wonder if that data allows them to actually "re-create" that match from logged data. All that report option would need to deliver is a "time-space" coordinates in battles logs database for devs to pinpoint the exact match and moment plus player comment field to point an issue. I think a report function would be somewhat needless and unnecessary as it would detract dev hours from other things for something that we can already do. It's rudimentary, yes, but it's no less effective. Maybe after we get PC Lag fixed xD forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201

Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4040	Posted - 2013.12.06 14:26:00 - [11] - Quote Robert JD Niewiadomski wrote: If your effort will help CCP to axe that lag bug, I wouldn't mind if you were rewarded with some in-game gear I hope CCP will appreciate your commitment Good luck Quite the contrary I'd rather an internship xD I'd rather they give me nothing, honestly. The best reward would be fair, competitive matches in PC and in general. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4041	Posted - 2013.12.06 14:57:00 - [12] - Quote IAmDuncanIdaho II wrote: Dunno. Shotty GoBang used to run at 480p I think - and he did it consistently for quite a while. So he must've thought it made some sort of difference. I keep meaning to try this myself, but I keep forgetting. If I remember I'll let you know any observations. Was talking about the texture/effects settings in the Systems Operation tab. I run 480p on PC matches occasionally and it usually helps out a lot. IgniteableAura wrote: Would be nice if CCP could designate a "bug team" and have "special contracts" for people to really test these things. Similar to how they did the squad cup. CCPlogibro was able to do some pretty neat things. I had suggested this to the CPM in the past but at current, CCP has no plans to implement it. I had actually, quite a few months ago, started a channel for a testing team to make a community-based ISD-like bug hunting team but it fell through. I've since gained some notoriety and reputation, however. IAmDuncanIdaho II wrote: Perhaps you also need to set your server to a specified region, instead of automatic, so you know which server you're playing on, and then you can ask ppl where in the world they are. At least then you can have a baseline to measure the non-network related issues. Easier just to get in touch with groups like Nyain San (Japan) or The Southern Legion (Australia/New Zealand) forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi A.N.O.N.Y.M.O.U.S. Renegade Alliance 4042	Posted - 2013.12.06 15:14:00 - [13] - Quote Roy Ventus wrote: Heard you can do away with it by having players join the match before it actually starts, when you have to wait in the war barge, then everyone hard resets and link back up. It seems like a lot but I can't think of a reason why this person who told me would lie to me. I've heard similar reports from our allied corporations. Will have a squad test this out. Traky78 wrote: We, at What The French, will gladly help you in anyway you want. We could easily make some fake matchs to test some things. Send me an Email IG man. Right right, Cazaderon told me to get a hold of you. Thanks for reminding me, will definitely do that. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi Ancient Exiles. Renegade Alliance 4211	Posted - 2013.12.24 09:02:00 - [14] - Quote [UPDATE: 12/24/2013] As my previous corp (ANONYMOUS) has fallen apart there's been some (major) delays in the actual Lag Hunt testing. I did, however, manage to re-shoot the original Equipment test with the PVR, so it's now in HD and far more effective. I'm currently waiting for it to render and once that it done I'll be posting up the video. What's different in this video? Was a bit more organized and we had a lot more equipment on the ground. This wasn't so much to test the equipment lag as much as a theory I had recently in that if I had a bunch of stuff running in the background on the PS3 it might have had some effect. The FPS drop is minimal but it is notable. There was some good recovery afterward however during the next test. A few things worth noting: - Texture warps seemed more prominent. I'll be making a post about this in the bug thread. - Small decrease in FPS when facing the cluster, same as before. - Increase in performance during second test - likely caused by switching to 480i. Requires further testing. - Disconnected both times when attempting to change sound profiles mid-game. forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi Ancient Exiles. Renegade Alliance 4211	Posted - 2013.12.24 22:05:00 - [15] - Quote So, on the theory that the sound of Nanohives/Drop-Uplinks is what is causing the framerate drops I have a bit of speculation and would like some further input from the community. In the video linked above, during the first test, I noticed that the sound of the nanohives was more or less synchronized and the audio waveforms in the video editor support that. I highly doubt that we coincidentally placed our nanohives down all at the exact same timing. The second part of the video (low resolution and all background applications turned off) however, the sound of the equipment was constant and overlapping, rather than synchronized. Curious as to whether or not this has something to do with it. Anyone more familiar with this that may know if it has an impact? forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201
Aeon Amadi Ancient Exiles. Renegade Alliance 4223	Posted - 2013.12.27 03:39:00 - [16] - Quote Leither Yiltron wrote: If you need a district to test this **** on just contact me Aeon and we'll see if we can't get it sorted. We'll see what happens. What The French offered this service up but stopped talking when I asked them about it O.o; forums.dust514.com/default.aspx?g=posts&m= (Frames)1544109 (Advertisement)1556863 (Packs)1570030 (Lag Hunt) 1570201