Support NeoGAF

[kodt]

This is why we should all be gaming on PC's with 100hz CRT monitors.

edit: I think someone who is very good at a fast game, and is used to playing on PC with a CRT monitor (Say Quake 3 or CS) can notice input lag once it gets above 50ms or so. I could notice it easily on early LCD monitors, my current LCD monitors have 15ms lag which is less than a frame so there is no way I would notice it. 2D fighters as well, SF3 3rd Strike for example would be greatly affected.

I realize this is software lag, but the effect is the same pretty much.

 

[DDayton]

The difference between 360/PS3 wired vs 360/PS3 wireless is so minute I don't think anyone could tell the difference. It would be logical that wired controls are more responsive.. but I would guess that it wouldn't be anything more than ~2ms difference.. so it doesn't matter.

Well, yes, that's what I assume, but there always seems to be some whining about wireless controllers around these parts...

 

[kodt]

Well, yes, that's what I assume, but there always seems to be some whining about wireless controllers around these parts...

I think it is because when wireless mice first came out you could notice the lag, so many PC gamers wrote them off right then and there and ignored them while they improved dramatically.

 

[Minamu]

What was Killzone 2 pre-patch?

You don't want to know Personally, I think that placing so much of the KZ2 blame on the deadzone is just making it easy for yourself. It does play a part but I believe that, in general, players that complain about the lag mean just that, not the deadzone. But that's just me

I think what partially masks it in GTA is that it's not a precision game. You have a tendenacy to notice it in games that require more precise timing for the most part. A God of War game or Street Fighter for instance generally needs very low lag or you start noticing it quickly becuase you aren't getting the results you want. Same with some fps games. But if the genre or game is a little less tight about how much of a window of time you have to do things in to be successful, you probably tend to notice it less.

Yeah that makes sense. I watched the video of it in the article & it looks horrible when in slowmo like that :lol But I'm pretty sure that once you're used to it, you'd take the lag into consideration, albeit subconsciously, when you plan your attacks, no matter the title. Anyone who has played a shooter must've noticed that lag works both ways, be it controller or server related. Countless times have I won a firefight simply because I was lagging, you know

The article also mentions that human perception can be deceptive & that it's an imprecise measurement tool. But shouldn't game design then be designed around that very fact so that the human perception is perhaps tricked into believing it's right even if it's not? After all, it's the players who are supposed to enjoy the product. That's how I'd design my game anyhow Perhaps they are, I dunno.

The difference between 360/PS3 wired vs 360/PS3 wireless is so minute I don't think anyone could tell the difference.

Dude, tell that to Guitar Hero fans. Around GH2 & 3, lag caused quite a shitstorm imho

 

[kodt]

Yeah that makes sense. I watched the video of it in the article & it looks horrible when in slowmo like that :lol But I'm pretty sure that once you're used to it, you'd take the lag into consideration, albeit subconsciously, when you plan your attacks, no matter the title. Anyone who has played a shooter must've noticed that lag works both ways, be it controller or server related. Countless times have I won a firefight simply because I was lagging, you know

yeah, you know I think for single player games it really isn't a big deal, these games have been playtested and obviously you can adjust for small amounts of input lag and still play the game just fine. Where the issue comes in is multi-player. If you can notice the input lag it is going to be very frustrating, not to mention many players have screens with different input lag times so the playing field is not perfectly even.

 

[DoomGyver]

Dude, tell that to Guitar Hero fans. Around GH2 & 3, lag caused quite a shitstorm imho

That's a third party peripheral, I was talking strictly first party hardware.

I doubt GH or RB peripherals use the same quality of bluetooth/2.4 tech.

 

[Inkwell]

The whole latency issue bothers me so much. I mainly play games on my PC, but also have access to a PS3. I try to avoid the PS3 as much as possible because of lag. We have a decent TV. It's a 52 inch Samsung LCD. It still has horrible input lag in game mode. I tried the Killzone 2 demo and hated it because of the sluggish controls + input lag.

I have tried so many different things to lower the lag. I have tried changing all of the special display settings on the TV, as well as multiple video input modes (PC input). I have tried using the controllers plugged in, hoping some of the lag was due to the controller being wireless. Nothing seems to help at all.

Do you think developers take this into account when you have games with QT sequences? The only game I've played that gave a large enough window was Prince of Persia (sometimes even too much).

As a side note, the most horrible experience with any kind of lag was with Mega Man 9. In full 3D games it seems less noticeable. After a while you seem to adjust to the delay, but it still sucks when you need to have a split second reaction time.

 

[Chinner]

pc gaming: 0 ms

 

[Talon]

definitely this is the worst generation of video games period.

I was never exposed to crap like this in any of the previous generations with fame rate problems, enormous amount of screen tearing and now this BS.

Okay, the framerate problems from Perfect Dark definitely deserve a spot in the pantheon of performance issues.

 

[Kittonwy]

So much goodness:



holy fucking shit



No, but that was in there on purpose. right? right? :lol

It's been fixed for a LONG time, everything from input lag to sensitivity to joining friends have been improved.

 

[gregor7777]

I'd like to see this done for Madden 10. I have it on the PS3 and I'm constantly frustrated by input lag. Even hitting the start button on the Press Start screen seems to be laggy.

It's probably not the same issue but it's annoying as hell.

 

[MultiCore]

pc gaming: 0 ms

:lol :lol :lol :lol :lol

As much as I love my PC's, this is not the case.

 

[FoxSpirit]

:lol :lol :lol :lol :lol

As much as I love my PC's, this is not the case.

Okay, around 20ms with 60 fps and 500 MHz USB polling.

 

[brain_stew]

The difference between 360/PS3 wired vs 360/PS3 wireless is so minute I don't think anyone could tell the difference. It would be logical that wired controls are more responsive.. but I would guess that it wouldn't be anything more than ~2ms difference.. so it doesn't matter.

tl;dr Don't worry about the controllers, worry about your display.

Yep, that's been tested and as expected the difference in simply negligible and within margin of error. The reason the tests are restricted to a wireless 360 controller is because that's the only controller the DF guys have one of Ben Heck's latency monitors for, so all other tests would have some degree of unreliability to them. Quite smartly the results are restricted to those that are reliable, repeatable and quantifiable, which is how stuff like this should always be.

I've got to say my respect for the Infinity Ward guys has shot up after reading this, they obviously realise the responsive controls of the COD series are large reason for its success. COD is really the poster boy success story for this generation, so surely the fact that IW emphasise the importance of input lag and framerate as a way to improve the gameplay in their games is evidence enough that consumers actually do care about this stuff, even if its on a subconscious level.

 

[Tain]

Why is 50ms the minimum, exactly? Is that just PS3's case?

 

[Stoney Mason]

I'd like to see this done for Madden 10. I have it on the PS3 and I'm constantly frustrated by input lag. Even hitting the start button on the Press Start screen seems to be laggy.

It's probably not the same issue but it's annoying as hell.

I haven't played the most recent maddens but it may or may not be input lag. I've noticed that EA games use a lot of transitioning animated menus in the past which can make them feel laggy. And I think sports games have in recent years introduced an increased perceived lag thing going on becasuse the animation cycles have become longer. I notice the same thing in some platformers where it feels like you constantly are waiting for animations to finish because the animations are so.... animated to pick a word.

I think a lot of things can contribute to input lag or perceived input lag.

animations
actual real input lag
frame rate issues
occasional/rare by design lag in certain specific cases (not near as common as claimed)
hdtv
latency lag in a multiplayer game.

etc,

Get a worse case scenario of many of these instances and in some cases a game can feel really bad.

 

[kodt]

pc gaming: 0 ms

Okay, around 20ms with 60 fps and 500 MHz USB polling.

yeah even the fastest LCD monitors have around 15ms input lag. Unless you are playing with a CRT running at same refresh rate as your fps.

And you can put some mice up to 1000hz USB polling Which brings it down to 1ms + almost 0ms input lag on CRT = you can get pretty damn close. But honestly anything under 40ms you won't notice.

Polling rate | Response time
125 Hz | 8 ms
250 Hz | 4 ms
500 Hz | 2 ms
1000 Hz | 1 ms

 

[LabouredSubterfuge]

I have nothing entirely inciteful to add to this thread other than the fact that this is one of the best tech articles I've read in a while.

:bow Digital Foundry (once again)

 

[mavs]

pc gaming: 0 ms

Educate yourself.

 

[StealthGoblin]

Killzone 2's controls were perfect before the patch. Screw the arcade shooter fanboys who whinged. I wish more developers would look at the weight and overall sweetness of Killzone 2's controls. Yay for not feeling like a camera on wheels.

 

[tanod]

I'd be interested to see how the NXE does. They tested the XMB. They should have tested the others.

 

[DeadGzuz]

EG and DF, no surprise they highlight bad examples on the PS3 and ignore 360 ones.

 

[LabouredSubterfuge]

EG and DF, no surprise they highlight bad examples on the PS3 and ignore 360 ones.

Whilst resisting the urge to quote your tag, didn't they say the XMB had low latency? Also, having a go at Killzone 2 was fair since they aren't the only ones displeased with the input lag present. Like me. (Weight my arse)

I'm also pretty sure that, had he the tech present, he would've tested the Dualshock 3. The guy simply puts the numbers together at the end of the day. Sorry if you feel victiimised by the fact that the PS3 doesn't always meet expectations (in a specifically quantifiable way).

(Disclaimer: Yes I own a PS3 and it is my console of choice a lot of the time)

 

[Stoney Mason]

EG and DF, no surprise they highlight bad examples on the PS3 and ignore 360 ones.

Wow. It's amazing that you could read that article and this is what you take away from it. Just wow.

Didn't they say the XMB had low latency?

Yes.

 

[TheExodu5]

Input lag is one thing I'm quite sensitive to.

Warhawk has to have the worst input lag I've ever experienced in a game.

LOL at everyone who thinks Killzone 2's input lag is "on purpose".

 

[Slavik81]

Is the cause of high input lag the use of an events/messaging system? I could see event chains causing input lag if you didn't process the event loop enough times per frame.

But even then... You'd think player-driven events would be pretty much directly mapped...

 

[The Faceless Master]

The difference between 360/PS3 wired vs 360/PS3 wireless is so minute I don't think anyone could tell the difference. It would be logical that wired controls are more responsive.. but I would guess that it wouldn't be anything more than ~2ms difference.. so it doesn't matter.

tl;dr Don't worry about the controllers, worry about your display.

yeah, iirc bluetooth latency is 2-3ms and the 360 wireless controllers are 2-3ms also.

 

[Beer Monkey]

Objective tests have shown that it is true that 360/PS3 wireless controllers do not cause a single frame of lag over their wired counterparts:

http://forums.shoryuken.com/showpost.php?p=5771841&postcount=40
http://forums.shoryuken.com/showpost.php?p=5773657&postcount=48

Regarding Killzone 2, I rented it and played it on release and found it completely unplayable at that time compared to CoD4 (the best game to compare it to IMO). I didn't post any such strong opinion because I knew the defense force would go apeshit and cry bias. It's nice to see some independent confirmation of the situation.

In my personal experience I find with shooters once cumulative lag (game lag plus TV lag) hits around 200ms it goes from becoming an annoyance to a total gamebreaker. I don't know where pre-patch K2 lag stood, but after adding in 60ms of lag from my plasma (I have measured it with both optical Rock Band 2 guitar tests and actual camera tests) K2 sure felt like it was over that threshold.

 

[Afrikan]

* 30FPS games have a minimum potential lag of 100ms, but many exceed this.

hmmm.... F1CE for the PS3 was 30fps... I wonder what it would score.

 

[CultureClearance]

The difference between 360/PS3 wired vs 360/PS3 wireless is so minute I don't think anyone could tell the difference. It would be logical that wired controls are more responsive.. but I would guess that it wouldn't be anything more than ~2ms difference.. so it doesn't matter.

tl;dr Don't worry about the controllers, worry about your display.

I don't know about this. I tried playing PacMan:CE on my wireless and actually had issue making proper turns with it after playing with my wired all the time. It was actually very frustrating.

I would bet you're adding another 40 or so ms to your lag with a wireless controller. I would love an official study on it though. That would put it all to rest.

I found this study quite interesting. Good stuff to know.

 

[poppabk]

I don't know about this. I tried playing PacMan:CE on my wireless and actually had issue making proper turns with it after playing with my wired all the time. It was actually very frustrating.

I would bet you're adding another 40 or so ms to your lag with a wireless controller. I would love an official study on it though. That would put it all to rest.

I found this study quite interesting. Good stuff to know.

Its more likely you were suffering from interference from another wireless device. 90% of stuff from wireless speakers, to most wireless networks is broadcasting in the same tight freq range. My wireless speakers interfere with my Wii such that I had to separate my HTIB from my Wii and place foil over the wireless speaker transmitter to prevent the interference. There is no proper indication that you are suffering from interference.

 

[Afrikan]

I'd like to see this done for Madden 10. I have it on the PS3 and I'm constantly frustrated by input lag. Even hitting the start button on the Press Start screen seems to be laggy.

It's probably not the same issue but it's annoying as hell.

this, This, THIS!!!

my god Madden has to be the worse of all games...mainly the option screen, in game.....and it gets worse when your editing the depth chart....and it gets worse when your editing the depth chart ONLINE....

and to think Madden 07- 09 was even worser....yes worser.

 

[Worm_Buffet]

You must have one hell of a strong stomach if you could handle the tearing in that mode. I seriously thought it was some sort of sick joke, trying to show off the most illegible and messed up image possible. That mode is just unplayable to me so it was never an option.

On the PC side, running at 60fps, tweaking some config lines, reducing the driver's render ahead setting and using triple buffering and a 1:1 mapped display for minimal lag, made any input lag a non issue.

Wouldn't enabling triple buffering increase input lag by 17ms at 60fps, when compared to double buffering? Do PCs default to quad buffering now? Or am I missing something here?

 

[SuperSonic1305]

I wish they had tested both Resistance games, they are surprisingly reactive for 30fps games.

That would mean something positive for PS3 which is a big no no for these comparisons.

 

[brain_stew]

I don't know about this. I tried playing PacMan:CE on my wireless and actually had issue making proper turns with it after playing with my wired all the time. It was actually very frustrating.

I would bet you're adding another 40 or so ms to your lag with a wireless controller. I would love an official study on it though. That would put it all to rest.

I found this study quite interesting. Good stuff to know.

Properly implemented triple buffering does not increase latency, well you "could" argue it does, but only compared to the part of the screen above the tear in standard double buffering. Double buffer v-sync does however increase latency, this is why triple buffering is so great compared to double buffer v-sync, better framerates and less latency, at the cost of a small amount of memory.

It sounds like you're getting triple buffering confused with "render ahead" which is something totally different and does indeed have the potential to introduce extra latency. As a PC gamer, you do ofcourse have full control over this through your drivers anyway.

I don't know about this. I tried playing PacMan:CE on my wireless and actually had issue making proper turns with it after playing with my wired all the time. It was actually very frustrating.

I would bet you're adding another 40 or so ms to your lag with a wireless controller. I would love an official study on it though. That would put it all to rest.

I found this study quite interesting. Good stuff to know.

Its all in your head. Its been measured and all controllers wired or wireless only add around 2ms of latency.

 

[brain_stew]

That would mean something positive for PS3 which is a big no no for these comparisons.

Oh just fuck off back to your cave why don't you. It really pisses me off seeing clueless little fan kiddies rubbishing an article that I personally know has involved a heck of a lot of work, time and financial investment and that I and a bunch of other guys were eagerly anticipating. We should know this stuff and its important that its out there. I'd have fought the fact that both IW and Neversoft have backed this article was enough to shut up the retarded SDF brigade, but nope, their stupidity and idiocy knows no bounds.

If technical data shows up your console of choice as inferior then man up and learn to liive with it, reality sucks. Some of us actually like to educate ourselves about technology rather than live on the words of our PR messiahs.

The DF bunch are great guys, if you've got an issue with their objective research then go over to Beyond3D and bring up your issue. Don't just bitch and moan, whilst trying to isolate yourself in blissful ignorance.

 

[Koren]

yeah even the fastest LCD monitors have around 15ms input lag. Unless you are playing with a CRT running at same refresh rate as your fps.

Last generetion Sony HDTV (W4000 and W4500) were under that (0-10ms in game mode according to HDTVTest, and I'm indeed unable to see a difference with CRT, while I was able to correctly estimate at 35ms the lag in normal mode simply using the Minna no Golf swing bar).

Unfortunately, the new generation (W5500) is, even in game mode, far over that (over 30ms, IIRC). I'm pretty worried, things are not always going in the good direction, especially since I'm feeling I'm in a really small minority really bothered by this issue.

 

[The Faceless Master]

I don't know about this. I tried playing PacMan:CE on my wireless and actually had issue making proper turns with it after playing with my wired all the time. It was actually very frustrating.

I would bet you're adding another 40 or so ms to your lag with a wireless controller. I would love an official study on it though. That would put it all to rest.

I found this study quite interesting. Good stuff to know.

the tech MS uses and the Bluetooth Nintendo and Sony use all say they have 2-3ms latency.

in a 60fps game, one frame is 16.7ms, and in a 30fps game, it's 33.3ms.

we're talking 1/5 to 1/15 of a frame latency.

40ms? really? sounds like you're paranoid.