Support NeoGAF

[squicken]

http://www.eurogamer.net/articles/digitalfoundry-lag-factor-article

CoD4 67ms
GTA4 133ms-200ms
Killzone 150ms (Patched version)
Halo3 100ms Shooting/150ms Jumping
Guitar Hero 50

As Neversoft itself is responsible for most of the latest Guitar Hero games, where latency is hugely important, it is perhaps not surprising that Mick West took such an interest in this subject, and his conclusions are intriguing.

* The lowest latencies a video game can have is 50ms (three frames) - the PS3 XMB runs at this rate, but few games reach it.
* Most 60FPS games have a 66.67ms latency - Ridge Racer 7, for example.
* 30FPS games have a minimum potential lag of 100ms, but many exceed this.
* Game developers should test their own games using the camera technique in order to weed out bugs - West says that Heavenly Sword's response slows down to 300ms just by turning the character, and reckons it's a technical issue that should have been resolved before going gold with the game.
* Citing GTAIV as an example, West suggests that a 166ms response is where gamers notice controller lag, which could also explain the Killzone 2 furore too.
* Game reviewers should accurately measure latency for their reviews where controller lag is an issue, in the hope that sloppy game response times come under far more scrutiny.

 

[Fyodor Dostoevsky]

So much goodness:

Heavenly Sword's response slows down to 300ms just by turning the character, and reckons it's a technical issue that should have been resolved before going gold with the game.

holy fucking shit

West suggests that a 166ms response is where gamers notice controller lag, which could also explain the Killzone 2 furore too.

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

 

[Stoney Mason]

Interesting article. Full list from the article.

Game Latency Measurement
Burnout Paradise 67ms
BioShock (frame-locked) 133ms
BioShock (unlocked) as low as 67ms
Call of Duty 4: Modern Warfare 67ms-84ms
Call of Duty: World at War 67ms-100ms
Call of Juarez: Bound in Blood 100ms
Forza Motorsport 2 67ms
Geometry Wars 2 67ms
Guitar Hero: Aerosmith 67ms
Grand Theft Auto IV 133ms-200ms
Halo 3 100ms-150ms
Left 4 Dead 100ms-133ms
LEGO Batman 133ms
Mirror's Edge 133ms
Street Fighter IV 67ms
Soul Calibur IV 67ms-84ms
Unreal Tournament 3 100ms-133ms
X-Men Origins: Wolverine 133ms

 

[DoomGyver]

Unreal Tournament 3 100ms-133ms

That's surprising.

BioShock (frame-locked) 133ms
BioShock (unlocked) as low as 67ms

So that's why the two modes felt different.

Call of Duty 4: Modern Warfare 67ms-84ms

Hopefully MW2 gets closer to 50ms.

 

[Septimus]

So much goodness:



holy fucking shit



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

In GTA4, hell no. In Killzone 2, yes.

 

[McBacon]

Never noticed it. Not even in Killzone.

 

[Unregistered007]

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.

 

[Stoney Mason]

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.

So I assume you never played in the playstation 1 era or indeed for even many games in the playstation 2 era.

Or the rampant slowdown in some genres like shumps that plagued earlier generations.

 

[Fyodor Dostoevsky]

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.

There's a difference between you just now learning about this stuff and it actually being there. I definitely didn't notice it as much before learning the terminology and how to spot it.
Before it was just "this game sucks" :lol


This gen does have it's problems, but hey, we can patch things now and sometimes developers do.

 

[karasu]

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.

lolz @ your absurdity.

 

[Stoney Mason]

Never noticed it. Not even in Killzone.

I notice it quite a bit. I noticed it in the Killzone demo when I played it. And I notice it in other games. Not saying that makes them bad games. But I definitely notice it and I definitely prefer when it's lessened. It effects how you perceive how responsive the game controls are.

That's surprising.

So that's why the two modes felt different.

Yep. Generally speaking that's why a consistent high frame rate is desirable. It increases responsiveness although you also have to have good code in general that is responive to controller input. The extreme example is playing a game with lousy frame rate and how it feels delayed and laggy.

 

[Dabanton]

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.

Well stop playing then.

 

[Erebus]

So according to that list, Burnout Paradise has the least latency from all the games tested?

Criterion is awesome confirmed!

 

[shuri]

I dispute the validity of this test.

 

[Panajev2001a]

This is what the OnLive guys are betting money on... to have fast enough servers to stream the game to you in that time it would normally take for your input to be registered and to modify on screen elements.

If:

"time_it takes_the_signal_from_controller_to_reach_OnLive_device

+

time_it_takes_the_input_to_reach_server

+

server_processing_time

+

time_it_takes_the_frame_from_server_to_reach_OnLive_Device

+

time_it_takes_the_frame_to_be_displayed_on_TV"



is close to 60-100 ms then their service can work... and with FPS's too.

 

[Erebus]

Yep. Generally speaking that's why a consistent high frame rate is desirable. It increases responsiveness although you also have to have good code in general that is responive to controller input. The extreme example is playing a game with lousy frame rate and how it feels delayed and laggy.

Heavenly Sword comes instantly to mind.

 

[Lagspike_exe]

I notice it quite a bit. I noticed it in the Killzone demo when I played it. And I notice it in other games. Not saying that makes them bad games. But I definitely notice it and I definitely prefer when it's lessened. It effects how you perceive how responsive the game controls are.

Killzone 2 (if you are refering to the KZ2 and not KZ1) demo had a really huge deadzone. It's very hard to distinguish input lag from stick deadzone.

 

[kitch9]

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.

The drama queens are dancing in dresses that way -------> That maybe more your bag.......

 

[squicken]

Never noticed it. Not even in Killzone.

I definitely noticed something off with GTA and Killzone. But I only played Killzone pre-patch, and the author seems to think most complaints should be attributed to the dead zones.

GTA, Niko is a tank. Walking out of your damn apartment can be an exercise in frustration.

 

[Stoney Mason]

Killzone 2 (if you are refering to the KZ2 and not KZ1) had a really huge deadzone. It's very hard to distinguish input lag from stick deadzone.

I won't get into a whole Killzone 2 thing here. I enjoyed the demo but when I played the demo it definitely felt delayed. As in the way when I play certain games it wasn't as responsive as I ideally desired it to be. Some games do have that as part of a design for a specific context although if a game is generally doing it all over the place that is why people tend to compare games and say X game doesn't feel as responsive as Y game.

 

[bee]

pretty obvious stuff really, vsync doubles the lag as seen in the bioshock example and 30fps is laggier than 60, wow shocking

 

[Pepto]

Onlive vindicated?

 

[Fyodor Dostoevsky]

I definitely noticed something off with GTA and Killzone. But I only played Killzone pre-patch, and the author seems to think most complaints should be attributed to the dead zones.

GTA, Niko is a tank. Walking out of your damn apartment can be an exercise in frustration.

lol, true. enclosed spaces in GTA are hard to navigate. Not so much in open ones, I found.

 

[vpance]

It felt pretty bad in pre patch KZ2 since there was delay and exaggerated ramped acceleration for aiming.

 

[gofreak]

This is what the OnLive guys are betting money on... to have fast enough servers to stream the game to you in that time it would normally take for your input to be registered and to modify on screen elements.

If:

"time_it takes_the_signal_from_controller_to_reach_OnLive_device

+

time_it_akes_the_input_to_reach_server

+

server_processing_time

+

time_it_takes_the_frame_from_server_to_reach_OnLive_Device

+

time_it_takes_the_frame_to_be_displayed_on_TV"



is close to 60-100 ms then their service can work... and with FPS's too.

That's interesting...never thought about input lag when considering the viability of 'remote play' services. But yeah, given the surprising amount of input lag in some games, they may have pretty reasonable windows of time to work with.

 

[Stoney Mason]

I definitely noticed something off with GTA and Killzone. But I only played Killzone pre-patch, and the author seems to think most complaints should be attributed to the dead zones.

GTA, Niko is a tank. Walking out of your damn apartment can be an exercise in frustration.

There is a difference between his walking like a tank necessarily and input lag. You could have good input lag and he still feels like a tank because of the animation system. Your input is registered but it takes you awhile to complete an action because the animation takes awhile to complete what you want to do.

In the case of GTA 4 there is some input lag also combined with a very variable frame rate. A lot of open world game suffer the variable frame rate thing due to the nature of their games.

The animation system input was slightly improved in TLAD btw although I'm guessing a large portion of the controller lag still remains.

 

[MvmntInGrn]

Heavenly Sword felt horrible in the demo IMO, now I know why. Killzone 2 felt different but I got used to it quickly, haven't' tried new patch controls.

Onlive vindicated?

More than anything I think it just says that it should be feasible, at least lag shouldn't be a huge concern for most gamers (as it should fall under the proposed 166ms point of notice).

 

[Quasar]

So...is wireless controllers having impact on this?

 

[squicken]

There is a different between his walking like a tank necessarily and input lag. You could have good input lag and he still feels like a tank because of the animation system. Your input is registered but it takes you awhile to complete an action because the animation takes awhile to complete what you want to do.

In the case of GTA 4 there is some input lag also combined with a very variable frame rate. A lot of open world game suffer the variable frame rate thing due to the nature of their games.

The animation system input was slightly improved in TLAD btw.

I understand. My complaint is that somewhere between the animation system and the controller latency, it was difficult to navigate the character in tight spaces.

Some of the newer Samsung LCDs pack in 200ms of lag in standard mode. So you are talking a third of a second in total latency between your input and its on-screen reaction.

 

[More Fun To Compute]

BioShock (frame-locked) 133ms
BioShock (unlocked) as low as 67ms

The sluggish control in the Bioshock demo on the 360 ruined the experience for me. I eventually played on PC which was better but not good enough IMO. Didn't think off changing too many settings but I might try that now, if I ever go back to it.

 

[dejan]

Factor in display lag (there are LCDs out there with 100ms+ input lag) and start to weep.

 

[brain_stew]

Important to note is that these figures do not include any extra lag added by your display device. Most modern HDTVs will add at least 2-3 frames of input lag (if you're lucky) and the bad ones can easily add 2-3x that amount. Unless you're using a 1:1 mapped display in a "PC" or "game" mode, then expect the worst. So the vast majority of people will be playing Killzone 2 with input lag beyond the 200ms threshold, as such its no wonder so many complained about the controls in that game. 200ms+ of input lag is quite frankly broken and borders on unplayable.

Before the fanboy wars rage, a lot of work and research has went into this article, I've been following the construction of it quite carefully, offering suggestions where possible.

 

[Stoney Mason]

The sluggish control in the Bioshock demo on the 360 ruined the experience for me. I eventually played on PC which was better but not good enough IMO. Didn't think off changing too many settings but I might try that now, if I ever go back to it.

Bioshock unlocked on the consoles was a much better experience than having it locked to me. It's more inconsistent because of the frame rate swings but when it's doing above the locked rate it feels much better. If you can tolerate the swings I thought it was a better play. I started off with it locked and then unlocked and I never switched it back.

 

[Lagspike_exe]

I won't get into a whole Killzone 2 thing here. I enjoyed the demo but when I played the demo it definitely felt delayed. As in the way when I play certain games it wasn't as responsive as I ideally desired it to be. Some games do have that as part of a design for a specific context although if a game is generally doing it all over the place that is why people tend to compare games and say X game doesn't feel as responsive as Y game.

You are correct, the game did feel unresponsive. It's just that input lag wasn't the one to blame, but stick deadzone. The result is the same (your input takes some time to get displayed on the screen), but the cause is different.

 

[nightez]

I think your brain automatically compensates for any latency. I honestly never notice any latency when playing games even the games with the most extreme numbers on that list. Its the same with playing music instruments live, there is quite a bit of latency by the time you hear the sounds on speakers but your brain compensates automatically.

....

 

[brain_stew]

Bioshock unlocked on the consoles was a much better experience than having it locked to me. It's more inconsistent because of the frame rate swings but when it's doing above the locked rate it feels much better. If you can tolerate the swings I thought it was a better play. I started off with it locked and then unlocked and I never switched it back.

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.

 

[squicken]

Important to note is that these figures do not include any extra lag added by your display device. Most modern HDTVs will add at least 2-3 frames of input lag (if you're lucky) and the bad ones can easily add 2-3x that amount. Unless you're using a 1:1 mapped display in a "PC" or "game" mode, then expect the worst. So the vast majority of people will be playing Killzone 2 with input lag beyond the 200ms threshold, as such its no wonder so many complained about the controls in that game. 200ms+ of input lag is quite frankly broken and borders on unplayable.

Before the fanboy wars rage, a lot of work and research has went into this article, I've been following the construction of it quite carefully, offering suggestions where possible.

http://www.hdtvtest.co.uk/reviews.php

These guys are the only reviewers who test for screen latency when doing TV reviews. When I had a Samsung LCD, the lag was such that I had to lower the difficulty in all games.

Sure, they have game modes. But those modes turn off all of the processing that made the picture so pretty in the first place.

 

[More Fun To Compute]

Bioshock unlocked on the consoles was a much better experience than having it locked to me. It's more inconsistent because of the frame rate swings but when it's doing above the locked rate it feels much better. If you can tolerate the swings I thought it was a better play. I started off with it locked and then unlocked and I never switched it back.

If I remember correctly the lag got really annoying for me in the demo when the action was busy so frame rate limiting probably wouldn't solve that problem. I tend to stick VSynch and frame limiter options on as default and forget about them because I notice screen tearing even more. I would much rather have an option to vsynch to a steady 60fps with downgraded graphics.

 

[Stoney Mason]

I think your brain automatically compensates for any latency. I honestly never notice any latency when playing games even the games with the most extreme numbers on that list. Its the same with playing music instruments live, there is quite a bit of latency by the time you hear the sounds on speakers but your brain compensates automatically.

....

You can ignore it. If a person ignores something for long enough he has a tendency not to see it but it still exists. Here is a quote of mine about Section 8, a fps game below a few weeks ago on the demo on the 360. If your a fan of a certain genre and like it for very specific reasons when you play something that doesn't feel right it's immediately noticeable

The game controls poorly in my opinion and has control lag relative to other console fps games.

Your natural expectation as a gamer is that when you press a button you get an immediate on screen action or at least an animation reaction to start the process of what you want to do. The longer the delay between you pressing the button or moving the stick and that happening, the more you notice something feels wrong. It's more noticeable in genres where very specific timing is required.

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.

The tearing was pretty bad in that title but in a trade off between the two, I prefer the tearing.

 

[Pepto]

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.

Bad framerate and bigger input lag vs tearing... The choice was easy for me.

 

[Minamu]

50ms in the ps3 xmb huh? I KNEW I felt something there! Perhaps it's just sluggish movements I feel, like in Mirror's Edge, but you can definitely sense that something is off there. I haven't thought about in GTA4 though. If this is such a big problem though, why not design gameplay speed around it if it's not fixable? I bet that's why I haven't noticed it in GTA.

Edit: Where in GTA is this noticable? Navigating building is just difficult due to animations mainly from what I've seen as has been mentioned.

 

[dallow_bg]

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.

Same, I can't stand tearing.

 

[Stoney Mason]

50ms in the ps3 xmb huh? I KNEW I felt something there! Perhaps it's just sluggish movements I feel, like in Mirror's Edge, but you can definitely sense that something is off there. I haven't thought about in GTA4 though. If this is such a big problem though, why not design gameplay speed around it if it's not fixable? I bet that's why I haven't noticed it in GTA.

Edit: Where in GTA is this noticable? Navigating building is just difficult due to animations mainly from what I've seen as has been mentioned.

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.

 

[Joel Was Right]

I bought a DS3 for Grand Theft Auto IV and I felt a little "lag", or controller latency, when moving the character and especially driving. As you can see from the tests, GTAIV isn't the quickest game with the thumb but I swear the added weight of the controller emphasised that (at least for me).

Whenever I feel this issue of lack of responsiveness, I would use the Sixaxis.

P.S. Guerilla made a massive design mistake. There were a lot of returns of the game in the local game stores near me.

 

[RockmanWhore]

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

 

[CrunchinJelly]

What was Killzone 2 pre-patch?

The guy even says the patch was most likely just tweaking the stick deadzones, so would it not have made sense to check it pre and post-patch to check?

It's also not in the full list??

 

[Joel Was Right]

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

Resistance was very well done IMO. I felt in total control.

 

[cyberheater]

Resistance was very well done IMO. I felt in total control.

Yep. Controls were spot on for this awesome game.

 

[DDayton]

Strictly speaking, these are software/game lag tests, not controller lag tests.

I only say this as I came into the thread expecting to see a comparison of lag times for the various controllers on the market, wired versus wireless, etc..

 

[DoomGyver]

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.