LAIR Thread - * Rules: post #469

Someone please elaborate on this "real-time" progressive mesh technology. What else can we expect from a second iteration ?

EDIT:

Capeta, for your list of impressive things, you forgot about the water. I wonder about the wind too (since flags and smokes are present). The slow down could be due to other things happening in parallel. Hopefully they have enough time to polish these. They promised a solid 30fps.

As for exploding sail ship... there may be explosives on board.

 

I don't understand the gripes. I was skeptical when I only had the screenshots to go on, but the game in motion is very striking. Its seamless expansiveness is unprecedented at that high a level of detail and activity. Frame-for-frame it doesn't look as good as corridor shooters or other games of lower scene complexity and world activity, as expected, but it's far from shabby for the kind of game it is.

 

Next step would be if you could land and take off at will.

Then it would be like GTA.

But you can only land in certain places (which are probably more detailed), according to the developer walkthrough.

This kind of flying game would be interesting if you could play online in co-op mode. Have a real wing man and so on.

Warhawk would be interesting if it was like that too.

Or maybe they do a jet fighter game with ground targets and in an online game, some players would control ground-based units like SAMS and maybe scramble fighters while other players would be attacking land targets.

 

I think the ships contain fire powder that's why they were exploded like that.

 

These are directions unrelated to progressive mesh though. Landing anywhere seems to be an artist and time issue. Online is game design, network code and time (again) but there may be too many update messages in such a huge game world.

Within the progressive mesh technology itself... is it related to destructible buildings (since there are some in the game) and level of details ? How does it work with streaming (what gets streamed) ?

 

I wasn't saying that free-roaming gameplay had to do with progressive mesh or any of the other technology they were talking about.

However, maybe instead of doing fluid simulations, which does make the water look great, they could use the CPU for other things.

For all we know, the AI of these enemy units could turn out to be stupid, just be sitting ducks.

So instead of making pretty water, maybe the SPUs could have been running better AI code.

It seems like they've developed a lot of technology for this so I hope it does well and there is a sequel. Then they could possibly leverage the assets and the tech. to enhance the gameplay in areas like having ground combat (off the dragon) and being able to land and take off at will

Not to mention online play, although in an earlier interview, I think that Julian guy said that to make a real immersive experience, they couldn't do online in this iteration.

 

Yes but there are better ways of doing it in terms of realism, like burning then small consecutive explosions of the hull etc followed by the ship slowly tilting/sinking.

I think it's a simple geometry tesellation routine. As far as streaming, I think they said the textures get streamed which makes sense.

 

I saw the developer walkthroughs, and they are certainly more impressive than some of the more recent shots I've seen in this thread.

It's still quite clear to me, though, that Factor 5 no longer has the knack. Their technology is still good, but it's no longer unquestionably bleeding-edge.

And I know PR and sales demand it, but some of this Cell stuff is just getting out of hand. You'd think we never saw such mind-blowing features as TEXTURE STREAMING or WATER DYNAMICS or PROGRESSIVE GEOMETRY until Cell came out. I used to feel sorry for Xbox developers because they were *always* asked what they were doing to tap the additional power, as if they have to sell the console or at least the console's potential in one game. Now I feel sorry for PS3 devs for the same reason. Unfortunately, the answer "we could do the same but somewhat less on other platforms" just isn't a good PR answer.

 

Err... very little that goes into games is actually new these days. What's new is the additional horsepower to better realize existing concepts to the degree that developers have previously only dreamt about.

 

The developers walkthroughs actually, finally, give this game a little credibility. That was a huge city scene with what looked like full lighting and shadowing. The intro cinematics were excellent quality too.

The actual comments were somewhat bewildering. SPUs are easier to program than EE. How much programming of EE did Factor 5 do? And progressive mesh? Impossible until this gen??

It's LOD. I don't understand the big deal, but maybe I'm missing something? Maybe it' the scale he's talking about? As a technology it's been around for years. DX7 came with a demo of a progressive mesh tiger, where you could scale down the number of triangles from thousands to a few hundred. Terrain has come with progressive meshes for years. Someone more knowledgeable will have to comment on whether Lair's progressive mesh stuff is really something special or just mindless prattle.

As for the water dynamic, was that really a smart thing to invest in for a flying game? If you're not interacting with the water, using a hack that looks similar is IMO a smart move, to leave you resources for other things.

 

Hmm... I'm sure there is a simple explanation for this (Out of development time or run-time resources for example).

Thanks guys for the anwers. Shifty, the F5 guy clarified in the walkthrough that it's the scale he's refering to... not the technology. I asked the "progressive mesh" question also to find out what scale (How many triangles) we are talking about here.

True... I was thinking of the same thing. Someone might have decided that "normal" water looked ugly/unnatural when super large critters emerged out of it to engage in battles. The presenter also mentioned some level using fluid dynamics in interesting ways. There might be quite a few fight scenes at the sea (more than just sinking boats) to justify the effort.

I'd rather they remove the flag simulation (for more important stuff like AI and better fire). I wonder how much run-time resources it took to simulate the waters.

Might be new hires (but experienced PS2 programmers) they picked up along the way.

 

Those developer walkthrough vids have easily some of the best visuals seen in a video game. Not sure about the gameplay.. Games like these will surely get people exited about the PS3.

 

Maybe because its a dynamic progressive mesh with the wather ripples and all instead of a fixed one? Could also be the dynamic battles taking place, I'm just winging it.

But it's good that we got better quality videos shooting down the low res textures comments.

 

Doesn't such a claim call for an example of another more advanced console engine? Sorry for being anal but empty words hurt.

 

Nice idea...except the weak particle effects of water droplets really mess up the visual impact of realistic waves! Perhaps they just wanted to bag the prize for 'best ever water effect in game' which has always been a strange sort of landmark?

 

EDIT: (can't seem to find the edit option)

@Shifty Geezer

What I tried to say is that there should be a difference between a mesh calculated in real time compared to a pre-calculated one found in games like GTA.

And all the water effects are a way for Factor 5 to show the "power" of their mesh technolegy.

 

In turn Lair with flying dragons, realtime day and night shows where the deficiency of GTA4 lies if it's really without an airplane.

 

You need sufficient time and posts in the forum before editing is enabled.

Snowblind Studios had that on PS2 and XBox in their engine. Originally in puddles in the BG: DA game, and then in large water bodies in CON. And as I said, what's the point if precalculated meshes or other cheats look pretty similar? Seems like a lot of processing for little gain.

Still, there may be other reasons. Googling the subject, I saw suggestions of Factor 5 doing download titles. A water engine might be part of those, and so they integrate it into Lair seeing as they have it. It becomes a component, more than a feature.

 

Totally agree.

I wasn't sold from seeing some of the other stuff (which in comparison looks really early/unfinished).
Those two latest developer videos contain the best next-gen graphics I have seen.

My mind has been blown.

 

The interview by Dean Takahashi has some interesting comparisons in types of gameplay:

http://blogs.mercurynews.com/aei/2007/04/qa_with_factor_5_chief_julian_eggebrecht.html#more-3517

He talks about differences compared to Gears of War, about being able to do macro-micro gameplay. He compared it to like a combination of Ace Combat and Gears of War gameplay.

Sense I got though, when he talked about doing enough of Star Wars and wanting to do original IP, is that Sony gave them a big budget and/or cut them a nice publishing deal.

The business arrangement is probably at least as importanta to F5 as their enthusiasm for PS3 design. So for all the talk about risk, it's probably Sony's money at stake.