|
Post by spectral on Dec 9, 2010 20:12:32 GMT -5
Hey bros, after lurking on these forums for a while I though I might post a question. Like a lot of people I've been experiencing issues with Black Ops relating to lagginess and poor hit detection (especially in Nuketown for some reason) and I'd like to try and understand more about how this aspect of the game works. I found info from Valve (links below) about lag compensation etc but I'm not too sure how relevant this is to the IW game engine. developer.valvesoftware.com/wiki/Source_Multiplayer_Networkingdeveloper.valvesoftware.com/wiki/Latency_Compensating_Methods_in_Client/Server_In-game_Protocol_Design_and_OptimizationMy queries are: 1. Does hit detection for Black Ops take place on the client or the server? 2. Does Black Ops use hit boxes for hit detection or is it based on pixel collision? 3. Is there any functional difference in how hit detection works for PC vs Console? 4. How has the hit detection changed between MW2 and Black Ops? Can anyone shed some light on this? You are my bro, bro.
|
|
|
Post by guy on Dec 9, 2010 20:46:31 GMT -5
My queries are: 1. Does hit detection for Black Ops take place on the client or the server? 2. Does Black Ops use hit boxes for hit detection or is it based on pixel collision? 3. Is there any functional difference in how hit detection works for PC vs Console? 4. How has the hit detection changed between MW2 and Black Ops? In my limited experience: 1: On the server, by that I mean the console hosting the server. Packets of data are sent to the host, who then determines wtf is actually happening. This is why "The host always wins." He has 0 ping, because he is the source. 2: Hit boxes (bad ones) 3: Yes. PC has dedicated servers, making the connection based problems much more fair, but players closer to the server bank still recieve more advantage than those far away. 4: It's gotten worse. Any improvements Infinity Ward made in going from CoD4 to MW2 are null. As this CoD uses the WaW version of the Engine, aka the CoD4 version that got loaned t Treyarch. Basically everything it does is based off unrevised two year old netcode.
|
|
|
Post by spectral on Dec 9, 2010 22:16:26 GMT -5
Thanks Guy, much appreciated!
|
|
toysrme
True Bro
"Even at normal Health, there's no other choice than the Vector" Den Kirson
Posts: 1,339
|
Post by toysrme on Dec 9, 2010 23:23:44 GMT -5
nothing to do with anything.
cod4+ on all titles by default run snaps 20 & packets 30. that's how often the server updates u & u it
1) FIRST a client tells the host that you killed someone. THEN the bullets are dicussed 2) there are not hitboxes in cod. it's hit-scan. 3) no difference in anything between DEFAULT PC cod4+ and console. All dvar/cvars in place are the same minus things dealing with aiming & game layout 4) there are no changes. you're simply not used to playing sans stopping power
|
|
brandon7s
True Bro
XBox: Rune Gladius
Posts: 363
|
Post by brandon7s on Dec 10, 2010 0:47:30 GMT -5
4) there are no changes. you're simply not used to playing sans stopping power As someone who almost never runs stopping power in mw2, there has got to be more to it. Lag on Black Ops is tremendously more noticable compared to MW2.
|
|
|
Post by guy on Dec 10, 2010 0:54:25 GMT -5
4) there are no changes. you're simply not used to playing sans stopping power As someone who almost never runs stopping power in mw2, there has got to be more to it. Lag on Black Ops is tremendously more noticable compared to MW2. www.softsailor.com/news/28469-new-call-of-duty-black-ops-screenshots-and-details-released.html"On the tech side, Call of Duty 7: Black Ops uses a highly modified version of the World at War engine. The guys at Treyarch have improved the fire and water tech and the lighting has been completely overhauled. The team also focused on physics and destructibility." "Well there's ur problem!"
|
|
|
Post by mw0swedeking on Dec 10, 2010 1:17:41 GMT -5
"destructability"? Referring to...?
The lighting is much better than WaWs though
|
|
toysrme
True Bro
"Even at normal Health, there's no other choice than the Vector" Den Kirson
Posts: 1,339
|
Post by toysrme on Dec 10, 2010 1:26:41 GMT -5
again.... black ops is MW2 with a new version of trey's custom sound engine they used in W@W....
all of which are nothing more than cod4+ next to nothing. for console, W@W has trey's sound. MW2 has texture streaming and a bandwidth race every time a host leaves a lobby
BO is mw2 + trey's sound & the banwidth race is now just a 1 time test every time you sign into multiplayer.
same shit, different games.
|
|
aequinox
True Bro
hakuna matata
Posts: 366
|
Post by aequinox on Dec 10, 2010 8:43:20 GMT -5
You're claiming the netcode is the same, with the only difference being a single bandwidth race?
|
|
|
Post by guy on Dec 10, 2010 10:58:35 GMT -5
"destructability"? Referring to...? The lighting is much better than WaWs though I dunno, Infinity had a demonstration level for MW2 (based on Favela) called Toolbox I believe. It flaunted "destructability" as well. I'm assuming rockety structures, chipping sections of wall, maybe destroying light cover. I would imagine it used too much memory online and only created unecessary amounts of lag. BO propably had the same situation. (Although it might still exist in campaign)
|
|
|
Post by countcracula on Dec 11, 2010 10:13:42 GMT -5
I rarely used SP in MW2 because the M82 is a OHK to the chest. So does this game use the same engine as WaW or not? I know that WaW was made with the borrowed COD4 engine, so I thought that naturally Black Ops would be made with the borrowed MW2 engine, especially after seeing the heavy MW2 influence like payback, buzzkills, first blood, and the way the XP notifier pops up during the game.
|
|
toysrme
True Bro
"Even at normal Health, there's no other choice than the Vector" Den Kirson
Posts: 1,339
|
Post by toysrme on Dec 11, 2010 16:27:09 GMT -5
You're claiming the netcode is the same, with the only difference being a single bandwidth race? no im stating everything is identical. For all COD4+ games on console (and in general by default on PC) snaps "20" - 20/sec received from server cl_maxpackets "30" - 30/sec sent to server cl_maxPing "800" - search for pings under sv_maxclients "18" - duh sv_maxRate "25000" - preffered banwidth rate sv_maxPing "0" - max ping allowed on server (110 default pc) sv_minPing "0" - min ping allowed on server (20 default pc) badhost_endGameIfISuck "1" - game can end if host sucks badhost_maxDoISuckFrames "300" - you suck @ 300ms lag badhost_maxHappyPingTime "400" - unhappy at 401+ms lag badhost_minPercentClientsUnhappyToSuck "0.51" - % 51% of clients =/> 300ms worth of lag. End game / Test for new server badhost_minTotalClientsForHappyTest "3" - min of 3 clients in game to drop/retest Recommended bandwidth to host: band_12players "384000" band_18players "580000" band_2players "64000" band_4players "128000" band_8players "256000" It's the same code. The game nearly always runs 60fps and you receive an update from the server every 3rd frame. 99.99999999999999999999999999999999999% of the time you only go 3 frames without new information from the host. Which is easily in spec for any quake based game. The only difference in COD4/W@W, MW2 and BO is the following: COD4/W@W simply picks a host by an unknown process on console. That host retains host until they leave. MW2 chooses hosts based on a bandwidth race. The host chosen retains host until they leave, clients are unhappy (forcing a new test), or a lobby is merged bandwidthtest_announceinterval "200" bandwidthtest_duration "500" bandwidthtest_enable "1" bandwidthtest_fudge "1" bandwidthtest_ingame_enable "1" bandwidthtest_ingame_fudge "1" bandwidthtest_maxpacketsize "1000" bandwidthtest_quota "0.5" bandwidthtest_rounds "10" bandwidthtest_scalar "2" bandwidthtest_timeout "1500"
BO differs from MW2 in WHEN the testing is done. ASA you load BO, it looks for a network connection & does several things. 1) Checks / Updates & loads any playlist / hotfix data from Activision server 2) Runs the bandwidth test & records the data (Shown in the multiplayer options when hitting BACK button). This is the same test as MW2 runs bandwidthtest_announceinterval "200" bandwidthtest_duration "500" bandwidthtest_enable "1" bandwidthtest_fudge "1" bandwidthtest_ingame_enable "1" bandwidthtest_ingame_fudge "1" bandwidthtest_maxpacketsize "1000" bandwidthtest_quota "0.5" bandwidthtest_rounds "10" bandwidthtest_scalar "2" bandwidthtest_timeout "1500"
If anything, BO is better than MW2 as you don't have random re-tests. Everyone is tested right off the bat = Done.
|
|
toysrme
True Bro
"Even at normal Health, there's no other choice than the Vector" Den Kirson
Posts: 1,339
|
Post by toysrme on Dec 11, 2010 16:32:31 GMT -5
PC dedicated servers only matter if you're wanting to run custom code / tweaks without hacking on console.
Outside of frame-rate issues on groundwar - LAUNCH. There is more than enough CPU time to keep track of clients. (Remember, CPU time in even extremely bleeding edge games is only a small fraction of time spent composing a frame. Most of it is GPU time. In actuality, data sent down an HDMI link takes longer than most CPU's take to do all processing for a single frame).
Play with people from where you're from. It's that easy. If you go willy nilly on console trying to find games, don't blame console / Dedicated PC servers if you're lagging. Try playing on someone even remotely close to your own network....
|
|
aequinox
True Bro
hakuna matata
Posts: 366
|
Post by aequinox on Dec 11, 2010 18:41:42 GMT -5
But the majority of the time, the console is on a network with more than one device. I know in my house, my bandwidth can and will change a lot over the course of a XBox session. Retests are necessary to find the connection quality at the time it is needed, not some random point in time before hand.
|
|
|
Post by Contrary on Dec 11, 2010 19:40:14 GMT -5
As someone who almost never runs stopping power in mw2, there has got to be more to it. Lag on Black Ops is tremendously more noticable compared to MW2. i concur with this statement. we all know that if you see a guy and shoot where you see him in mw2, even if there is a daly between firing and killing due to lag, he will die. thats why we get all these killcams in mw2 where it looks liek the bullets are bending around corners. because the lag compensation is so good.
|
|
|
Post by caboose on Dec 12, 2010 0:05:52 GMT -5
But the majority of the time, the console is on a network with more than one device. I know in my house, my bandwidth can and will change a lot over the course of a XBox session. Retests are necessary to find the connection quality at the time it is needed, not some random point in time before hand. Unless you've got a good Quality of Service router, you're probably not a good host either.
|
|
aequinox
True Bro
hakuna matata
Posts: 366
|
Post by aequinox on Dec 12, 2010 0:25:14 GMT -5
I never said I'd be a good host, in fact I'd wager I'm well below average.
What I am saying is that there is a good chance that at least one out of the twelve (or eighteen in ground war) players in the lobby will score abnormally high for their bandwidth test (read: well above their average) and win host which they will then prove they will not be able to handle.
Basically, I'm stating that the network usage of the average XBox connection varies quite a bit, and a single test is nowhere near adequate for properly determining the best host.
|
|
|
Post by mw0swedeking on Dec 12, 2010 2:35:28 GMT -5
Say BO runs it's test and I win host. 15 seconds later, my roommate starts downloading the inception blu ray. My ability to host is going to be severely hampered by this new development. That's the point aequinox is trying to make.
|
|
|
Post by caboose on Dec 12, 2010 3:04:13 GMT -5
I never said I'd be a good host, in fact I'd wager I'm well below average. What I am saying is that there is a good chance that at least one out of the twelve (or eighteen in ground war) players in the lobby will score abnormally high for their bandwidth test ( read: well above their average) and win host which they will then prove they will not be able to handle. Basically, I'm stating that the network usage of the average XBox connection varies quite a bit, and a single test is nowhere near adequate for properly determining the best host. Ah, OK - gotcha. Yeah, that would be unfortunate.
|
|
|
Post by mysteria9 on Dec 12, 2010 9:27:03 GMT -5
I've gotten the impression that there are way fewer "magic bullets" in BO than in MW2, but it might be because we have dedicated servers now so the lag compensation isn't that strong nowadays, because it lags much less.
|
|
aequinox
True Bro
hakuna matata
Posts: 366
|
Post by aequinox on Dec 12, 2010 10:13:15 GMT -5
Yes. Now given eighteen people in a single lobby, there are pretty good chances for you to be stuck with a host who's roommate is downloading The Inception in Blu Ray.
|
|
|
Post by chip sandwich on Dec 12, 2010 12:09:04 GMT -5
Had a match today that would occasionally drop below 30fps, dunno why, I had a 4 bar but wasn't host. I was also stubborn enough to keep using the Olympia. It was mostly just random severe drops, obviously noticeable when a Napalm Strike occurred but sometimes it was just completely random too.
Also, some guy on our team was spamming his mic, just saying "blebleblebleblebleblebleblebleblebleblebleblebleblebleblebleblebleble.." for the whole match.
|
|
|
Post by caboose on Dec 12, 2010 12:57:19 GMT -5
Heard of the mute button?
If you get fps drop, then the host likely exceeds his bandwith.
|
|
|
Post by chip sandwich on Dec 12, 2010 13:14:41 GMT -5
Heard of the mute button? I thought it was pretty funny, especially listening to everyone else telling him to shut the fuck up.
|
|
|
Post by Disgruntled Jigglypuff on Dec 12, 2010 16:27:20 GMT -5
So why is it that my bandwidth averages at around 600kps, but I still get constantly picked as host? Let me give you a common scenario. In Ground War, half of the time I get picked as host after host-migration, but host-migration reactivates before the match even starts and I proceed to get kicked. Why is that?
|
|
|
Post by spectral on Dec 12, 2010 19:07:39 GMT -5
Quote toysrme: So are you saying that the starting point that Treyarch used for BO development was the MW2 code base? As far as I can tell the from searching the web the starting point for BO was the WaW engine which was based on a fork of MW1. This would mean the MW2 and BO code bases have been developed in parallel from this point. The best confirmation that I can find for this a posting from someone who appears to be a Treyarch dev who worked on BO: forum.beyond3d.com/showthread.php?p=1492472#post1492472If this is the case, then whatever work that IW did on the netcode from MW1 -> MW2 won't necessarily be in BO.
|
|
n1gh7
True Bro
Black Market Dealer
Posts: 11,718
|
Post by n1gh7 on Dec 12, 2010 19:19:58 GMT -5
I think what he is trying to say is the between all the COD games since 4 there have been no changes to the net code. Whether this is true or not I don't know, but I think that is his position.
|
|
|
Post by spectral on Dec 12, 2010 20:05:37 GMT -5
Fair enough
|
|
|
Post by vulgar on Dec 13, 2010 5:22:42 GMT -5
If Blackops' netcode is superior, why has my experience with it been so awful? It's improving, but it still doesn't hold a candle to MW2. Purely anecdotal, but come on now, when lobbies are mostly 3 bars something is wrong. And don't tell me MW2 was more generous with 4 bars. While that may be true, you don't need to see how many "bars" you have to notice those 4 shots you fired didn't make it to the killcam.
It's a shame there's two seperate engines, but it's better than waiting 2 years. Besides, BO is getting there. The shoddy hit detection is mostly a symptom of the usual lag. Fix that and it's a solid COD.
|
|
|
Post by ray661 on Dec 13, 2010 10:16:10 GMT -5
So why is it that my bandwidth averages at around 600kps, but I still get constantly picked as host? Let me give you a common scenario. In Ground War, half of the time I get picked as host after host-migration, but host-migration reactivates before the match even starts and I proceed to get kicked. Why is that? Speed doesn't = bandwidth.... 90% of the time your internet speed doesn't matter in gaming. What matters is your ping, and how clogged up your lines are (aka bandwidth). Bandwidth does effect speed, but speed doesn't effect bandwidth at all (unless everyone around you happens to be downloading big stuff all at once at a high speed, but that still comes back to my point, YOU can't singlehandedly control the amount of bandwidth you get) If you still don't understand think of it like this Bandwidth = Lanes in a highway Speed = Your speed limit Ping = The distance from where you started to where you're going Now you can see how speed doesn't effect bandwidth, but bandwidth affects speed (if there's too many cars, the lanes get congested and everyone has to slow down)
|
|