Killzone 2 pre-release discussion thread

[vanquish]

Did the gun bang the moment the button was pressed, or as there a brief delay? My TV/monitor hasn't got a game mode, so may be introducing the lag I experienced, though I haven't noticed it in any other games.

Thanks for supporting my post Shifty.

 

[patsu]

Doesn't KZ2's use of deferred rendering make it more susceptible to input lag than other traditional methods of rendering?

Why ?

 

[Sigfried1977]

Only lag I noticed (both in the Demo and in the promo code) was because of the huge controller dead-zone. The thing doesn't seem to respond one bit unless you push it past the halfway point.

 

[Neb]

Doesn't KZ2's use of deferred rendering make it more susceptible to input lag than other traditional methods of rendering?

I dont think it would have anything to do with it being deffered but instead of some delayed rendering for perfomancce reasons. I thought I read GTAIV also has input lag (consoles) and it is also uses deffered rendering. But Stalker/Clear Sky (deffered rendering) has no input lag at all.

 

[patsu]

Yes, forward rendered games can have lag too based on the gamasutra articles I linked to.

 

[vanquish]

Yes, forward rendered games can have lag too based on the gamasutra articles I linked to.

I said deferred rendered games are probably more susceptible to input lag, not they have lag while forward rendered games don't.

Are we sure GTA4 uses deferred rendering, because on 360 it uses tiling, and apparantly tiling & edram + deferred rendering is a big no no.

Having said that, I did notice lag in GTA4, especially when you're in cover and pop out to shoot someone, and the gun doesn't fire when you pull the trigger.

 

[AlNom]

Are we sure GTA4 uses deferred rendering, because on 360 it uses tiling, and apparantly tiling & edram + deferred rendering is a big no no.

The light pre-pass renderer is a deferred technique.

 

[patsu]

I said deferred rendered games are probably more susceptible to input lag, not they have lag while forward rendered games don't.

Are we sure GTA4 uses deferred rendering, because on 360 it uses tiling, and apparantly tiling & edram + deferred rendering is a big no no.

Having said that, I did notice lag in GTA4, especially when you're in cover and pop out to shoot someone, and the gun doesn't fire when you pull the trigger.

I know, but deferred rendering is not more likely to cause input lag than forward rendering. I think you are thinking of delayed rendering.

EDIT: Reqouting the gamasutra articles in case people are interested: http://forum.beyond3d.com/showpost.php?p=1265444&postcount=1789

I don't know what the latest story is. GG mentioned that it could be a game design decision or a demo-only issue. Feel free to report issues but deferred rendering may not be the cause.

 

[Neb]

Are we sure GTA4 uses deferred rendering, because on 360 it uses tiling, and apparantly tiling & edram + deferred rendering is a big no no.

The PC version code says so.

 

[Aeoniss]

... Ok so I just looked at that interview by guerrilla games where it showed the breakdown of their engine...

Remember how he displayed all those stats? SPE time, RSX time, Memory etc?

Well I watched the video like 12 times and wrote down all the memory statistics for KZ2 just for the laughs- and the number I came up with is really odd.


I got (total) 458MB being used by KZ2 in that one level (Train level) based on all the information on that screen..

But the screen didn't display the Gbuffer- how big was that exactly? 30MB?

^ could somebody help answer this one?

Couldn't we find out how much memory Guerrilla has on the PS3 simply by adding in all the figures?

Right now I got 54MB unused as stated... But how big is the Gbuffer? (as it isn't listed in the memory statistics).

H\o while I try and find an image.....

http://i184.photobucket.com/albums/x250/Vulgotha/PS3stuff.jpg


All the memory added up = 54MB unused.

 

[RenegadeRocks]

Even R2 uses Deferred techniques for Lighting . It has nothing to do for lag in controls. Frankly speaking, I don't experience any lag in KZ demo, but maybe I am not as nitpicky as you guys. The only thing that is slightly annoying sometimes is the acceleration as I anot able to line-up headshots accurately.

 

[Kittonwy]

Hey, I was wondering about the MP gameplay. What happens if everyone on the enemy team is a Saboteur ? Wouldn't it be an enormous advantage since everyone would look like a friendly soldier at first glance (especially in a small map) ?

They don't get to use the standard assault rifle or sniper rifle, saboteurs have sort of a disadvantage in terms of available weapons.

 

[vanquish]

well I did some more tests today, that avoided the effects of acceleration on the controls and perceived input lag.

And I am pretty sure that there still is some noticeable, eg. If I gently nudge the stick in a direction, and let go of the stick, the reticule will finish moving only after my finger is free from the stick.

What's 100% certain though is that the slowness of the controls, the acceleration, and any lag (which may or may not be present in the final game), make the game significantly harder than it would be otherwise (similar to Resident Evil, though not as bad).

 

[minimoke]

I don't understand how the perceived input lag can have anything to do with the rendering. If it did then the complexity of the scene should have some affect on the lag. For example the lag should be greater when there are a lot polygons, effects etc. on the screen and somewhat less when there isn't much happening e.g. the corridor before the warehouse in the demo.

I've done some non-scientific tests and the reaction time/turn speed/jump response etc. is constant in both complex and simple scenes.

Just my $0.20 worth.

 

[patsu]

What's 100% certain though is that the slowness of the controls, the acceleration, and any lag (which may or may not be present in the final game), make the game significantly harder than it would be otherwise (similar to Resident Evil, though not as bad).

Gah... I changed the sensitivity all the way up (for both X and Y axes) just to see what the big deal is, there is some differences but it's barely noticeable for me.

The demo was easy, even though the AI made each replay rather unique. It's the first level afterall, and we have plenty of cover. As an average FPS player, I could run around, flank and gut Helghasts quickly. Because the Helghasts usually hide behind covers and expose their heads, landing a headshot/helmut shot was not too difficult. Watching the MP videos also didn't surface any noticeable aiming issues by other players. I'd say an accomplished FPS player should be able to adapt to the scheme with some practices.

The RE5 demo is more restrictive and more complex in comparison (e.g., Limited weapons, limited ammos, enemies don't drop ammos, need to manage inventory, can't shoot and move at the same time, can't clear obstacles, ...). It's a totally different kind of game altogether.

 

[Phil]

I don't understand how the perceived input lag can have anything to do with the rendering. If it did then the complexity of the scene should have some affect on the lag. For example the lag should be greater when there are a lot polygons, effects etc. on the screen and somewhat less when there isn't much happening e.g. the corridor before the warehouse in the demo.

My impression is that the post-processing is causing some lag. If you add so many post-processing effects after most of the rendering is complete, you're going to see some lag between the refresh of the page and when you pressed the button.

 

[vanquish]

Gah... I changed the sensitivity all the way up (for both X and Y axes) just to see what the big deal is, there is some differences but it's barely noticeable for me.

The demo was easy, even though the AI made each replay rather unique. It's the first level afterall, and we have plenty of cover. As an average FPS player, I could run around, flank and gut Helghasts quickly. Because the Helghasts usually hide behind covers and expose their heads, landing a headshot/helmut shot was not too difficult. Watching the MP videos also didn't surface any noticeable aiming issues by other players. I'd say an accomplished FPS player should be able to adapt to the scheme with some practices.

The RE5 demo is more restrictive and more complex in comparison (e.g., Limited weapons, limited ammos, enemies don't drop ammos, need to manage inventory, can't shoot and move at the same time, can't clear obstacles, ...). It's a totally different kind of game altogether.

If you find KZ2's controls easy, then you aren't a very experienced FPS player I'm sorry. Because coming from COD4, Halo3, Uncharted, Gears, Resistance, MGS4 etc, I can tell you it's definitely noticeable. Though definitely not as bad as RE5

I also try and shoot the helghast heads when they're behind cover, while I can easily do this in Gears etc. in KZ2 at 90% sensitivity it is very easy to over/undershoot due to the lag and acceleration, which means headshoots are much harder to make.

 

[assurdum]

I dont think it would have anything to do with it being deffered but instead of some delayed rendering for perfomancce reasons. I thought I read GTAIV also has input lag (consoles) and it is also uses deffered rendering . But Stalker/Clear Sky (deffered rendering) has no input lag at all.

I understood correctly? GTA IV uses deffered rendering?

 

[patsu]

If you find KZ2's controls easy, then you aren't a very experienced FPS player I'm sorry. Because coming from COD4, Halo3, Uncharted, Gears, Resistance, MGS4 etc, I can tell you it's definitely noticeable. Though definitely not as bad as RE5

I said the lag's barely noticeable and doesn't affect me that much.

Uncharted is more tricky because as you or RudeCurve mentioned before, the pirates dodged more suddenly compared to the KZ2 demo. Halo, CoD, Unreal and RFOM are more twitchy and easy to pick up. Resistance 2 took me a while to get used to despite its fast and lag-free gameplay. Bioshock has more clumsy controls for me, but people seem to manage. MGO controls are slower and more complex than KZ2, mainly due to CQC. RE5 is not even a FPS, so I'll leave it aside. Overall, KZ2 feels natural once I get used to the default controls.

Everyone is different.

I also try and shoot the helghast heads when they're behind cover, while I can easily do this in Gears etc. in KZ2 at 90% sensitivity it is very easy to over/undershoot due to the lag and acceleration, which means headshoots are much harder to make.

I don't know, I shoot off their helmets pretty frequently. Most of the time I prefer close combat in KZ2 because I like to set the pace (instead of waiting for the enemies to recover). While KZ2 may not be as simple as the twitchy FPSes (which I enjoy), I don't think the KZ2 demo is significantly harder. It's just a simple, introductory demo.

 

[RenegadeRocks]

Why is it so difficult for people to adjust to a new mechanic in a game ?
Every game is different, and I found aiming and shooting very difficult in MGS3 initially, but adjusted and enjoyed the game immensely thereafter.
Its more about mindset than lag, I feel.
There might be a actual lag which I do not notice, or maybe My TV doesn't exagerrate it so I don't notice, but for a Single Player experience, I think it is nothing to rant about. Adjust and enjoy, if you can't, there are many other games to play. No Sweat !

Personally, i think I find it very refreshing to have an "Anticipation-> action-> follow through" in the camera movement when I move in KZ2. Same with gun animation. It feels very fresh and more like controlling a person than a floaty camera.
Just a thought, but maybe the intricate camera movements associated with looking around might be responsible for what you call "Lag".