Does anyone know the exact numbers it takes for a game like that to occur with no lag in ping, download, and upload?
Thanks.
Forum Index > Tech Support |
IncubusStarCraft2
Canada18 Posts
Does anyone know the exact numbers it takes for a game like that to occur with no lag in ping, download, and upload? Thanks. | ||
Cyro
United Kingdom20150 Posts
| ||
WonnaPlay
Netherlands912 Posts
For example you had a SC1:BW which would use 200mb over a whole month, where a browser game as HABBO Hotel (my little sisters played this) would use 10 GB. Now this had more to do with the insufficient way the other program functioned, but it's a fair example. RTS games also don't rely that much on latency compared to FPS for example. You can use 'Wireshark' to see how much data you use. (you'll need a bit of understanding of computer programs to quickly understand how it works, but it's very good to check, since it will differ from network to network how much you use). | ||
Cyro
United Kingdom20150 Posts
100mbit/s for example is 12.5 megaBYTES per second, AKA 12,800KB/s. An internet load of 128KB/s would show as 1% of that. 64KB/s would show as 0.5%, etc. That's just a basic display and it doesn't tell you utilization per program, but you have a pretty good idea if you're on desktop and it says a consistent 0.1%, and then you open sc2 and it goes to 0.5% during a 1v1 game for example. The entire graph on my screen goes from 0% load to 1% load so it's pretty easy to see exactly. The more random stuff on your PC using internet bandwidth (especially inconsistently) the harder it is to be able to quickly glance at that tab and see load from a specific program. I do it quite often though, just hit ctrl+shift+escape and click across tabs --- RTS games also don't rely that much on latency compared to FPS + Show Spoiler + To add to this, sc2 processes commands in a tick about once every 40 to 50 milliseconds (~20-25x per second at fastest game speed). Since there's some latency involved and it would be weird to be dancing between almost zero latency and 50ms (depending randomly if you clicked right before the tick or right after) i think they buffer commands by 1 tick or so, so there's a lag of about 100ms (0.1 seconds) before internet latency is taken into account. Since FPS games run differently with different methods to sync every game client*, having a 30ms ping can be twice as responsive as a 60ms ping. Having a ping of 30-50 instead of 100 can make it feel like a very different game, where you get a lot less "but i totally shot him first!" moments when you hit stuff on your screen but his shot in your face reached the server first. With the same difference in an rts though, 50 ping instead of 100 - since your commands are being delayed by 100ms anyway, it'd be a 150ms delay vs a 200ms delay - only 1.33x faster instead of twice as fast - and nowhere near lagless. It does vary by engine, but a lot of RTS's use that model i assume. Heroes of the Storm also uses it - which is disappointing - it naturally can't be as responsive as League of Legends for example due to that. There are some other negative effects - while it uses less bandwidth, it has higher CPU load - massively higher in some cases - and since a lot of simulation has to be done on ticks, there are slight spikes in latency on your system (not internet latency) and loss of visual smoothness whenever those ticks happen, because CPU load spikes momentarily and one frame gets delayed. You need a system more powerful than you would expect to mitigate those spikes - you might need as much as 120-180fps for it to be as good as a constant 60fps on a 60hz screen, and being at 80fps for example would show visual stutters over 20 times per second coinciding with hardware lag spikes. Not only that, those high framerates are simply not sustainable - much of the processing is done on one thread and it becomes a very high load when lots of units are involved, even on heroes of the storm there are issues there on the best processors, so you can't simply overpower the problem and get your responsiveness and smoothness back. Overall i feel that blizzard took a compromise here that was bad for playability - they didn't have a MOBA engine, they took the sc2 engine and repurposed it instead of building one - and as a direct result, both performance and responsiveness was compromised on both low and ultra high end systems - even if many of the positive features are intact. This style of engine is good for an RTS with two players and hundreds of units - it has some serious tradeoffs though many of them might even be considered necessary - but it's just not suited for a MOBA. --- *FPS games with low player counts can just have a server send packets to everyone saying "Player 3 is at XYZ coordinates" in a continuous stream AFAIK, many many times per second SC2 doesn't do that - it sends the clients a list of commands like in a replay file. Something like "Player 2 told marine X to attack move to XYZ coordinates" and then it expects everybody who's playing in the game to have their computer simulate it, work out a path, see if any unit moves into attack range of the unit etc. There's not even any communication sent saying that it did encounter another unit, or even that it killed another unit - they're just simulated simultaneously and identically from the same command given earlier. It's pretty surprising how much you can assume happened for every player in the game as long as you have a matching game state, a command and a completely deterministic engine (the same command will do the same thing, even on another computer 5000 miles away) This post is like 5x longer than i intended tl;dr sc2/heroes engine = very low internet bandwidth usage, not particularly ping sensitive since input processing is laggy anyway. High CPU load. | ||
| ||
Replay Cast
SW Korea Closed Qualifier Day1
Maru vs StatsLIVE!
Cure vs Classic
Solar vs GuMiho
ByuN vs herO
[ Submit Event ] |
StarCraft 2 StarCraft: Brood War Bisu 3382 Dota 2Mini 1563 ggaemo 511 ZerO 434 Snow 397 BeSt 369 Soulkey 328 Light 290 Hyuk 283 Rush 204 [ Show more ] Counter-Strike Other Games B2W.Neo2786 hiko1816 DeMusliM915 crisheroes626 Mew2King521 Hui .493 ArmadaUGS114 KnowMe104 QueenE95 NuckleDu92 Trikslyr45 Crank 35 Liquid`VortiX23 Organizations Counter-Strike StarCraft 2 Other Games StarCraft: Brood War StarCraft 2 StarCraft: Brood War
StarCraft 2 • Adnapsc2 2 StarCraft: Brood War• aXEnki • intothetv • Gussbus • Kozan • IndyKCrew • LaughNgamez Trovo • Laughngamez YouTube • Migwel • Poblha Dota 2 League of Legends |
Big Brain Bouts
BSL
TerrOr vs XuanXuan
Dark vs JDConan
Korean StarCraft League
StarsWar
Maru vs Classic
GuMiho vs ByuN
Stats vs Cure
Solar vs herO
WardiTV Invitational
CSO Cup
ForJumy Cup
BSL
Zhanhun vs WolFix
Dienmax vs Cross
Sparkling Tuna Cup
StarsWar
[ Show More ] WardiTV Invitational
Online Event
ESL Open Cup
Afreeca Starleague
StarsWar
ESL Open Cup
ESL Open Cup
Afreeca Starleague
StarsWar
Club NV x Duckling Show…
GSL Code S
Stats vs SHIN
Cure vs GuMiho
ESL Pro Tour
|
|