Lens of Truth Discussion *spin-off*

The more I look at the GI solution in the demo the less I'm being impressed, it's as if the secondary light bounce is made of point light instead of a more natural spread. I guess there's only so much you can do with cheap single bounce GI hack. Granted the effect cab be obvious at certain parts of the scene but mostly it's not very visible or impressive given the current treatment.
 
Is it a technical reason why player count is so low? Doesn't seem like anyone is bothered by this...I mean this seems very low for a FPS.
Apparently it's a gameplay decision. I suppose it's due to Free Radical not having much experience with vehicles/big maps.

Oh I have no problem saying the same in regards to C2. :smile:

My wording was off, I personally have issue claiming any game has the best in a category such as "best graphics", "best gameplay", or especially "best game EVAR!!" :LOL: One of the best, I'm fine with though.

If an effect can be proven to be the best by a technical measure, then I have no issue with this, even if the tech can still produce ugly results =p
We're in sync then :p

Yeah, the global lighting, including changing time-of-day, was introduced in MLB The Show 2009. They retained it in MLB The Show 2010. We'll see what happens in 2011. The games are supposed to be 1080p and 60fps. Back to your regular Crysis 2 programming… ^_^

EDIT: Individual games will do things their own ways since the developers tend to focus on different things.
Now I want to see a video of it xD Hopefully youtube has some that show off the effect.


2. That's because they seemed to look, purely, at the visual results. They didn't dive deeply into either tech from a lighting perspective.
Not really. They did delve into the tech, they're just wrong and don't understand it.

3. I would call that a "few" things with destructability. It's not total destructability, so I don't see a problem with that.
They didn't mention the concrete structures. Though this is more nitpicking :p

That's all? They mentioned a lot of things. I guess you agree with everything else. There was a poll with over 2,000 votes. 67% say Killzone 2 looks better. Isn't that the goal (look and functionality)?
They missed the HDR too (big deal). I'm not really sure of the demographics of that site. Besides, it's not the first time they make a comparison full of BS.

http://www.youtube.com/watch?v=eTvYPiEi4eU

*hyper facepalm*

I'm sure you saw that I mentioned shadows. I mentioned very poor because the light doesn't look anywhere near the color space of light from gun fire.
Yeah, they don't cast shadows. Though, what's the colorspace of gun fire? It's certainly not the deep orange of KZ2.

Does anybody or anything in C2 have faces? :) That eases the rendering burden, too.
Sure they do have faces. They're under their transparent masks xD.

Yes, it does.
Now it's just a matter of confirming whether it affects the GI dynamically or not :p
 
The more I look at the GI solution in the demo the less I'm being impressed, it's as if the secondary light bounce is made of point light instead of a more natural spread. I guess there's only so much you can do with cheap single bounce GI hack. Granted the effect cab be obvious at certain parts of the scene but mostly it's not very visible or impressive given the current treatment.
That's because it is done using point lights. How else would you be able to do it?

What's impressive is that they're doing it at all.
 
Proper global illumination is still hideously expensive in computing power, so don't expect a simple 4-5 year old console to get it done in a fraction of a second and still have it look good. If it was that easy we wouldn't need a render farm ;)
 
I guess using point lights for GI simulation in a deferred renderer would come in handy. Maybe they should use a hybrid solution such as baking the highly detailed GI like in mirror's edge while adding dynamic lights on top of it?
 
Well, Pixar's been faking GI for a decade with dozens or even hundreds of manually placed point lights, so it should be possible.

Nevertheless, depending on the game, fully baked environment lighting would still be a viable option in many cases.
 
I guess using point lights for GI simulation in a deferred renderer would come in handy. Maybe they should use a hybrid solution such as baking the highly detailed GI like in mirror's edge while adding dynamic lights on top of it?
The main benefits of 100% realtime lighting are intended for developers. Making as many changes and you like and not having to wait hours to see the results is an excellent improvement over other methods. Allows for more experimentation and polishing.

Gamers benefit too of course, having everything lit in the same way makes the graphics more consistent overall. There's also day-night cycles, destructible geometry, etc...

They still "fake" some of it though:

http://www.youtube.com/watch?v=w-AnWJpzyTA
 
They are making comments about tech they don't understand. If they don't want to look foolish then they should not make guesses about the tech behind these effects on screen.
True, but that doesn't mean everything they say is incorrect. It should still be on a case by case basis. Then, again, a lot of people don't operate like that.


KZ2 doesn't have total destructability either, so it's all meaningless in the end.
Well, this was a meaningless comment. What does this have to do with what I said?

Besides, are they at least comparing MP modes in both? Sorry I can't check while I'm at work.
Yes.

He was giving you a few examples to how they are wrong with their analysis on the tech behind the games. I see no point in spinning his post around.
I just asked for his stance on the rest of the article, so calm down. Are you his keeper?

Also, the "look" of a game is too subjective. I can think KZ2 is the greatest thing since sliced bread, but my friend can think it's a blurry mess that lacks color or appeal. Neither of us are wrong.
True. The only difference between what is considered sane or insane is the majority. The majority has spoken. ;)

To say LoT is right because the majority of the votes agree with their outcome is silly IMO
You would be correct, if I had said that. Since I didn't say that, you don't have to worry about being correct.
 
True, but that doesn't mean everything they say is incorrect. It should still be on a case by case basis. Then, again, a lot of people don't operate like that.

I know people make mistakes, but these guys have consistently proven to be unreliable IMO.

It's hard to take them serious, if they are wrong about that one thing, how are people to know what else they are wrong on or what factors they may have missed due to their lack of knowledge or awareness?

I just asked for his stance on the rest of the article, so calm down. Are you his keeper?

I'm calm :D, misread the comment, apologies.
 
Not a direct sequel as far as i know a novel is coming out telling how they went from the island to new york.

Dev said sp is like 10~12 hours long but consider they mostly do longer about it i reckon it's like 6~8 hours for normal players.:???:

I wish they would have ported Crysis 1 but it is probably too much for consoles to handle.:cry:
 
Not really. They did delve into the tech, they're just wrong and don't understand it.
Where? I only saw a "once over" type surface look at the tech.

They didn't mention the concrete structures. Though this is more nitpicking :p
?

They missed the HDR too (big deal). I'm not really sure of the demographics of that site. Besides, it's not the first time they make a comparison full of BS.

http://www.youtube.com/watch?v=eTvYPiEi4eU

*hyper facepalm*
Yes, they missed the HDR. They also missed the animations, character facial details, characters on screen, A.I., physics, amount of players handled in MP, LPCM, etc. (even bigger deal). ;)

At least that video contained a side by side view of both games. The differences were very, very noticeable, even if the commentary wasn't 100% accurate.

Yeah, they don't cast shadows. Though, what's the colorspace of gun fire? It's certainly not the deep orange of KZ2.
The gunfire isn't deep orange. It's more of yellowish in KZ2. Rockets are kind of orange, though. C2 gunfire seems someone turned on a xenon bulb for a fraction of a second with no shadow.

Sure they do have faces. They're under their transparent masks xD.
Hmmm...

Now it's just a matter of confirming whether it affects the GI dynamically or not :p
That means you can continue to hold out hope, huh? Keep hope alive! ;)
 
I know people make mistakes, but these guys have consistently proven to be unreliable IMO.

It's hard to take them serious, if they are wrong about that one thing, how are people to know what else they are wrong on or what factors they may have missed due to their lack of knowledge or awareness?
The only way to know what people are wrong and right about is to do your research. The only thing a good accuracy rating does is increase the likelihood of the true being told to you the next time. It doesn't gaurantee it. I do understand your frustration, though.
 
It's common knowledge that masses tend to be wrong about most things. Seems like this is no different :LOL:

If information is left out and present article has incorrect info then people are judging based on faulty info. It's nothing different from a newspapper article having wrong/left out info and people getting wrong impressions etc.

That said I looked at their screenshot comparision and both look bad but KZ2 looks uttelry horrible with so little color, 'bloomtopia' and PS2 era reminding bluriness (F3 urban/suburban interiors comes to mind with some blur filter and bloom filter). Pic 11 is also a joke, showcasing lensflare in KZ2 with motionblur blurring the whole scene vs a C2 muted scene due to being inside glass house. They should have showcased C2 lensflares/sunglare.
 
The only way to know what people are wrong and right about is to do your research. The only thing a good accuracy rating does is increase the likelihood of the true being told to you the next time. It doesn't gaurantee it. I do understand your frustration, though.

IMO if they are the one's who are doing an article based on the tech of these games, it's them who should be doing the research. :p

Yeah I know there is no guarantee that what I read will be right 100% of the time, but these guys are consistently wrong, and too many wrongs will destroy your credibility which is exactly what has happened with these guys.
 
Where? I only saw a "once over" type surface look at the tech.
You obviously didn't pay much attention to it. The article is full of technical jargon.

Yes, they missed the HDR. They also missed the animations, character facial details, characters on screen, A.I., physics, amount of players handled in MP, LPCM, etc. (even bigger deal). ;)
Again with the facial details. In a hectic MP? Somebody is trying too hard :LOL:

At least that video contained a side by side view of both games. The differences were very, very noticeable, even if the commentary wasn't 100% accurate.
Seems like you're the exact demographic LOT catters too. That video only proves how incredibly clueless they are.

The gunfire isn't deep orange. It's more of yellowish in KZ2. Rockets are kind of orange, though. C2 gunfire seems someone turned on a xenon bulb for a fraction of a second with no shadow..
Except for the muzzleflash shadows (which are pretty low res in KZ2) it's more realistic. The gunfire flashes don't fade out slowly like they do in KZ2.

That means you can continue to hold out hope, huh? Keep hope alive! ;)
Hope? You mean that Crytek's GI is far superior to other methods? That's confirmed. MLB's is nice but it doesn't really compare from the evidence that has been presented. Sorry to break it to you, but non-PS3 exclusives can have cutting edge tech too ;)

If information is left out and present article has incorrect info then people are judging based on faulty info. It's nothing different from a newspapper article having wrong/left out info and people getting wrong impressions etc.

That said I looked at their screenshot comparision and both look bad but KZ2 looks uttelry horrible with so little color, 'bloomtopia' and PS2 era reminding bluriness (F3 urban/suburban interiors comes to mind with some blur filter and bloom filter). Pic 11 is also a joke, showcasing lensflare in KZ2 with motionblur blurring the whole scene vs a C2 muted scene due to being inside glass house. They should have showcased C2 lensflares/sunglare.
And then comes Nebula with the smackdown :LOL:
 
And then comes Nebula with the smackdown :LOL:

Somebody gotta say it!

Hope? You mean that Crytek's GI is far superior to other methods? That's confirmed. MLB's is nice but it doesn't really compare from the evidence that has been presented. Sorry to break it to you, but non-PS3 exclusives can have cutting edge tech too ;)

Not without the C3ll. lolololo
 
Proper global illumination is still hideously expensive in computing power, so don't expect a simple 4-5 year old console to get it done in a fraction of a second and still have it look good. If it was that easy we wouldn't need a render farm ;)
There's a presentation in their homepage explaining how Crytek achieve Global Illumination on consoles. It seems like Crytek are proud of their solution.

http://www.crytek.com/cryengine/presentations

You can download it under the paragraph where it says "Real-time Diffuse Global Illumination in CryENGINE 3"

I have been reading it and I wonder how many bounces they use on the PS3 and 360, if 1, 2 or 3..., etc.

Also I've been using a search engine and I found this excellent article explaining in great detail how Global Illumination works.

http://labimpossible.com/2010/07/15/global-illumination-with-maya/

I didn't know there can be multiple bounces, not just one, but I didn't find in their presentation how many bounces Crytek are using on consoles versions. I guess they are using an even more complex solution for PCs.

This site is so insanely clueless.
I agree Lens of Truth seems to be so wrong at so many levels. It's a site that seems to be run by people who can actually make a decent comparison but they don't know how to put their thoughts into words because their technical knowledge is not the same as developers and so on.

I would like to know Digital Foundry's opinion on the matter, too.

First of all they are comparing a full Single Player game with the demo of a multiplayer map in beta development. :???:

Some of their other comparisons are slightly -or a lot- better, but I'm not inclined to go back there either due to some really horrible statements that they recently made about some technical stuff. Not only do they seem to think you can compare two games with different approaches and techniques and get it done with, they basically characterize some effects as very poor. Needless to say, this doesn't make me very happy when I read it as I would like to understand well what are the true reasons behind some of their technical statements, and I think a comparison like that should be written after Crysis 2 comes out, not now. By all regards, it makes perfect sense.

I am starting to like the game a little more. I played it the first time and it was like "....... o_O". It had been a lot of time without playing a FPS. But now I find myself casually playing it again every once in a while. xD Not quite sure why. I have to admit, it's been a very long time since a game has actually caused me to *YELL* out loud. :p

Well, I must go in a few minutes. Those are only my two cents about the thing, and I have no idea what the real technical situation is... so... :smile: Take it as you might...!
 
Also I've been using a search engine and I found this excellent article explaining in great detail how Global Illumination works.

What you should realise is Global Illumination (just as Reflections or Subsurface Scattering or Internal Combustion Engine) is not a technology, but a global term that has a wide variety of possible implementations that may have absolutely nothing in common. Just as there are two and four stroke engines, straight fours and V8s, Wankels and Diesels, so are there a lot of options for lighting your scene, from radiosity through photon mapping to whatever Crytek or Epic comes up with.

So the article you've found does not explain GI, it's only a rough and superficial tutorial on how to turn on Mental Ray's photon mapping. I wouldn't make any further assumptions based on just this little snippet.
Oh, and Crytek's solution has absolutely nothing to do with photon mapping, so this article is completely irrelevant to the topic.
 
Back
Top