Lag is a term that console players do not understand....200ms is very large.... it makes competitive play barely playable.

Now if you have a 200ms lag just from the controls, you're gonna be at 300+ when playing online? that just suxs...

@Deneidez

The -2 frames that Gamasutra applied was because of their TV setup. A different setup (LCD vs plasma, tv postprocessing or buffer modes etc) can bring that down to 1/50th of second ie half frame.

Anyway yes, in that video Resistance 2 seems to have 2-3 frames of latency. Good job Insomniac :)

@Hephaestos: I was thinking that too... I don't see a reason for 200ms lag from controls alone... That's the ping time I got back in the 56k modem era.

Well, 4-5 frames at 30fps mean 133-166ms of rendering lag really.

I'm a PC gamer and I code for a living (though admittedly not games) so I think I can add something about that. The "lag" you see when trying to find a server to play online is the netcode lag (let's say it is 30-100 ms). The rendering lag and the netcode lag will not simply add up, it depends on how the game is coded because when I pull the trigger the action doesnt need to wait for the rendering to be over before it's sent to the server. That's why netcode is always asynchronous and uses predictive code for trajectories that are then adjusted.

@NJ5

Again, network lag and control responsiveness are two very different beasts. And you might not "see the reason", but if most console fps games have a control response time of more than 100ms, the devs must have chosen their performance tradeoffs with good reasons, given the limitations they had to work with.

@WereKitten: I know they're different. The difference being that network lag is mostly a function of physical limitations and connection quality (i.e. it's not under the control of game developers), and input lag can be optimized.

I am a programmer too (not games either although I've done some amateur stuff), and I find it strange that these games get such big input lag. Apparently it's not due to controller lag since some games get less lag, so it must be lack of optimization. I mean what are they doing with the controller data, buffering it for a while or do they have a long pipeline before it gets to the engine? Very strange.

In other words I am curious to see if those lag times are measured properly and whether they happen on many games.

Well, I can't understand this one either. Are they really checking inputs under 30 fps or is there something even more strange behind it? For example my engine does check inputs currently 100fps and it doesn't depend anyhow on rendering fps. Even if you throw a lot of stuff on the screen inputs would be checked at least once per frame and shouldn't that be enough for 0-1 frames lag? :P

Yeah I can't imagine any good reason for it to depend on how much stuff there is going on, and I'm sure all the programmers here understand what I mean.

Unless they made some lame optimization which only checks inputs every 5 frames. If that's the case then we'd see response time anywhere from 0 to 5 frames, averaging around the middle of course.

@Deneidez

polling controls at least once per frame won't ensure 0-1 lag if, for example, the physics engine takes more than 33ms for its loop responding to your input and updating the world model. Not saying it does, just that there are several things happening between your input and the scene model changes and then from that to the rendering.

The only way the physics engine can take that long is if they're pipeling all the code stages throughout several frames, and I never heard of any game doing that. Wouldn't that entail having two copies of all the game data in memory for the different pipeline stages (one input and one output)? Sort of how backbuffers/frontbuffers work.

Isn't it great that as soon as a valid reasonable argument gets presented (by WereKitten and Deneidez) every detractor either lose interest or ignore the post completely?

It goes to show how much respect reason gets around here. Kudos guys for dispelling the 'lag' issue of KZ2, if the same gamer can tolerate similiar input lag for games like Halo3 and Farcry2 then they can do the same for KZ2.

If you have issue with the design of the control thats fine, you're entitle to your opinion, but acting as if this input lag which you've put up with the entire gen is suddenly game breaking is hypocritical at best.