Digital Foundry Article Technical Discussion Archive [2011]

Status
Not open for further replies.
It is not a case of game breaking but rather when you put it side by side next to kz3, you would see all the the shortcomings of crysis 2 a lot more clearly. Sub HD, low res DOF, strong aliasing, pop ins, lack of AF and flat textures are just some standout issues in the demo. Then again, we all have different standards.

My point was very specific about his claims of the graphical issues effecting gameplay, nothing more.

I agree the demo has it's issues, I've mentioned a couple of them in the Crysis 2 thread, but to turn the current topic of DF's analysis into some kind of comparison thread is unwarranted.

He raises some valid points so just because more people are disagree with him doesn't make him wrong. A lot of things here are speculations anyway.

The way I see, what points he does have are by coincidence and nothing more. I mean when you point out so much of a game as flawed (whether or not it really is), you're bound to hit some valid points in the end. His intentions are very clear and can be seen by some of his ignorant comments.

If these points could be discussed in an intelligent and open manner, then great, but I haven't seen any such intent.

KZ3 has a lot more dynamic lights on screen from what I have observed, Crysis 2 does have better lighting as in HDR and GI and I believe that's the double edged sword which should not be used here.

None of the comparisons should be used here. The direction each developer has taken in how they prioritize their engine and especially the artistic direction used in each game are far too different IMO to be compared accurately especially based on one single level in a demo that seems to be older code. Not to mention how pointless I think it is to compare a studio's first attempt on console with a multi-platform game versus a 1st party studio who has had experience on the platform they are developing on.

The whole thing is beyond silly and has no place in this thread, that's always been my point. I'd much rather discuss the article at hand, the points it brings up, and maybe some things I would like to see in future tech analysis articles. I'd be saying the same thing if he dragged Crysis 2 into a discussion about DF's analysis of KZ3.
 
Last edited by a moderator:
This isn't really a thread about Crysis 2 vs. Killzone 3, so I'm not really going to continue that discussion. :LOL:

It'll be interesting to see how the final game turns out, considering all the improvements in the past few months. If Crytek can reach the quality level of the 'Be Strong' trailer on consoles, eg:
Shot10.jpg


...I'll be really impressed :D

(that shadow bokeh effect really is a waste, though :p)
 
That's not what was said on the video, was it? ;) "This creates the illusion of colliding particles..." "...virtually colliding particles at once." Plus, the particles didn't even have multiple bounces like in KZ2.

It's his wording, doesn't change the fact it needs to do collision detection else it falls through objects. In other words simple collision detection based "physics". ;)

"New type of particle that creates an illusion of collision on the GPU instead of calculating expensive simulation on the GPU".

Also you can see the particles bounce multiple times on the ground after bouncing on the car. You should watch the video instead of quoting someone else. :p
 
But is this in the actual game? I think that part is the reason I checked for sparks and saw none using gun.

It should be relative to material. Different materials have different friction values. Also factor in the speed of particles and fade out time.
 
It is not a case of game breaking but rather when you put it side by side next to kz3, you would see all the the shortcomings of crysis 2 a lot more clearly. Sub HD, low res DOF, strong aliasing, pop ins, lack of AF and flat textures are just some standout issues in the demo. Then again, we all have different standards.

You know much of that goes both ways. I am sure examples been given in this thread, both ways.


KZ3 has a lot more dynamic lights on screen from what I have observed, Crysis 2 does have better lighting as in HDR and GI and I believe that's the double edged sword which should not be used here.

Not all lights are dynamic, they get switched out by LOD. Dont think every single light visible at any distance is an active-at-the-moment dynamic light. I am sure you can find out how they LOD lights in GG public documents.

Then for C2 multiplayer it obviously has low amount of lights becouse... well it's daytime and like in reality you dont have christmas decoration in summertime. Does KZ2/KZ3 even got sun based dynamic light?
 
You know much of that goes both ways. I am sure examples been given in this thread, both ways.
It's actually being kept at minimal in KZ3's case, hence the stark contrast.


Not all lights are dynamic, they get switched out by LOD. Dont think every single light visible at any distance is an active-at-the-moment dynamic light. I am sure you can find out how they LOD lights in GG public documents.
Yes true but the vast majority of them are dynamic, they didn't use deferred render for nothing after all. I'm sure GG have freed up more resources by dropping HDR in order to get more dynamic lights also.
Then for C2 multiplayer it obviously has low amount of lights becouse... well it's daytime and like in reality you dont have christmas decoration in summertime. Does KZ2/KZ3 even got sun based dynamic light?
Not even the Central station night level has that many lights in comparison. Why would kz use sun based dynamic light when it's not intended for day&night cycling? See this is why GG are smart with their resource utilization, they don't waste nothing for the sake of checklist.
 
Just a little OT , but LBP2 would greatly deserve more tech discutions and analysis.
Hiden behind it's naive cuteness ,there is a lot of amazing tech(s) going on here.
 
Yes true but the vast majority of them are dynamic, they didn't use deferred render for nothing after all.
I'm guessing most of them are particle point lights, not actual dynamic lights. It makes sense though in distant objects; what's the point of doing advanced lighting calculations if you're not going to see them?

Really though, just stop this discussion or spawn it somewhere else. It's completely off-rails ATM.
 
This isn't really a thread about Crysis 2 vs. Killzone 3, so I'm not really going to continue that discussion. :LOL:

It'll be interesting to see how the final game turns out, considering all the improvements in the past few months. If Crytek can reach the quality level of the 'Be Strong' trailer on consoles, eg:
Shot10.jpg


...I'll be really impressed :D

(that shadow bokeh effect really is a waste, though :p)
Interesting...There certainly seems to be some jagged edges,smells like "good ol" temporal AA.And even though its video capture IQ does not seem to be out of this world:smile:
 
Interesting...There certainly seems to be some jagged edges,smells like "good ol" temporal AA.And even though its video capture IQ does not seem to be out of this world:smile:
The bad IQ is probably due to the horrible EdgeAA/blur that we saw in Crysis. Temporal AA works well for distance objects (as you get sub-pixel antialiasing), but EdgeAA just seems to be pretty nasty in general :LOL:. Wish they would use NFAA, though (which is possible, since they hired the guy who developed it :smile:).

The video quality wasn't great, though. IQ was decent in other sections ;)
 
The bad IQ is probably due to the horrible EdgeAA/blur that we saw in Crysis. Temporal AA works well for distance objects (as you get sub-pixel antialiasing), but EdgeAA just seems to be pretty nasty in general :LOL:. Wish they would use NFAA, though (which is possible, since they hired the guy who developed it :smile:).

The video quality wasn't great, though. IQ was decent in other sections ;)

You mean Crysis 2 becouse the Edge AA in Crysis wasn't horrible.

Crysis default EdgeAA 2 setting.
http://forum.beyond3d.com/showpost.php?p=1520884&postcount=1833
 
Normal Filter AA?

Speaking of which I am curious to see how sub-pixel affecting 'MLAA' (SRAA) by Nvidia will be used.Seems cheap with 1.8ms for 1280x720 res though dunno what GPU and how it could be adapted to consoles if possible.

http://forum.beyond3d.com/showthread.php?t=59443

http://research.nvidia.com/publication/subpixel-reconstruction-antialiasing

Subpixel Reconstruction Antialiasing (SRAA) combines single-pixel (1x) shading with subpixel visibility to create antialiased images without increasing the shading cost. SRAA targets deferred-shading renderers, which cannot use multisample antialiasing. SRAA operates as a post-process on a rendered image with superresolution depth and normal buffers, so it can be incorporated into an existing renderer without modifying the shaders. In this way SRAA resembles Morphological Antialiasing (MLAA), but the new algorithm can better respect geometric boundaries and has fixed runtime independent of scene and image complexity. SRAA benefits shading-bound applications. For example, our implementation evaluates SRAA in 1.8 ms (1280x720) to yield antialiasing quality comparable to 4-16x shading. Thus SRAA would produce a net speedup over supersampling for applications that spend 1 ms or more on shading; for comparison, most modern games spend 5-10 ms shading. We also describe simplifications that increase performance by reducing quality.
 
It'll be interesting to see how the final game turns out, considering all the improvements in the past few months. If Crytek can reach the quality level of the 'Be Strong' trailer on consoles, eg:

Well we know they updated new effects for both cloak and strength/armor mode.

You mean Crysis 2 becouse the Edge AA in Crysis wasn't horrible.

Crysis default EdgeAA 2 setting.
http://forum.beyond3d.com/showpost.php?p=1520884&postcount=1833

Probably a dumb question, but is this Edge AA cheaper in performance than MSAA?

I plan to install and play Crysis 1 some time soon so I can finally beat it before Crysis 2 is released. I was originally holding out until I can build a computer to run it at settings that would do the game justice, but I much rather have the backstory when playing the sequel.
 
Probably a dumb question, but is this Edge AA cheaper in performance than MSAA?

I plan to install and play Crysis 1 some time soon so I can finally beat it before Crysis 2 is released. I was originally holding out until I can build a computer to run it at settings that would do the game justice, but I much rather have the backstory when playing the sequel.

On my 4890 it takes a fraction of a frame in hit. Game uses by default EdgeAA mode 1 but you can set it to the superior mode 2.
 
KZ3 has a lot more dynamic lights on screen from what I have observed
How many? Besides, most of the environment lighting in KZ2/3 is lightmapped whereas every single light and shadow in C2 are realtime.

Not even the Central station night level has that many lights in comparison. Why would kz use sun based dynamic light when it's not intended for day&night cycling? See this is why GG are smart with their resource utilization, they don't waste nothing for the sake of checklist.
Again, HOW MANY?

I see plenty of dynamic lights here:


All the lights are realtime, every plasma shot from the aliens rifles are light sources as well.

Here, all those small lights in the ceiling are realtime as well:


Just a little OT , but LBP2 would greatly deserve more tech discutions and analysis.
Hiden behind it's naive cuteness ,there is a lot of amazing tech(s) going on here.
Agreed, I love the shadow casting particles.
 
Crysis on consoles gets pretty big plus signs for being multi-platform and having a significant and rather cool creation and development environment behind it. It makes detailed comparison with a platform specific title almost pointless.

That said, I can't give any impressions personally because I checked yesterday, but the Crysis 2 demo doesn't seem to be available to me on Live.
 
What I found to be the biggest flaw with the demo, much bigger than any graphical error, is the horrible lag which effected the hit detection.

On my 4890 it takes a fraction of a frame in hit. Game uses by default EdgeAA mode 1 but you can set it to the superior mode 2.

I see, awesome, will have to try it out. Thanks.

How many? Besides, most of the environment lighting in KZ2/3 is lightmapped whereas every single light and shadow in C2 are realtime.


Again, HOW MANY?

I see plenty of dynamic lights here:


All the lights are realtime, every plasma shot from the aliens rifles are light sources as well.

Here, all those small lights in the ceiling are realtime as well:


Let it go man, just let it go. :p
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top