Sony's Kin - the next big ego shooter?

[Phil]

Netherlands developer Lost Boys is busy beavering away on a top secret, triple-A first-person shooter for PlayStation 2, which insiders are already billing as a potential Halo-beater. Known simply as 'Kin', details of the project are extremely thin on the ground, but the title is believed to be due for release in Spring 2004 in Europe, with a debut showing possible at this year's E3.
Kin is the first major project for Lost Boys, established in 2000, but the developer has two further next-gen titles in the works - Knights and Call of the Dragonfly. Despite the fledgling nature of the developer, SCEE was apprarently hugely impressed by the early stages of Kin, and quickly snapped up the title as a PS2 exclusive.

With Bungie's groundbreaking Halo being a serious system seller for Microsoft, Sony would dearly love to have a second-party title to match. One well-placed source told us: "This is a very big deal for SCEE, and it expects great things from Kin."

We contacted Lost Boys for further information, but a spokesperson would only admit: "We can confirm we're working on a title for Sony Europe, but I can't give you any further information," referring to a deal that was made public way back in 2000.

And Sony Europe was - remarkably - equally reluctant to spill the beans, telling us: "We can confirm the existence of the game but we have nothing further to add at this time."

It's expected that Kin will make its debut at E3 although this is currently unconfirmed. But with such excitement already in evidence behind the scenes, we'll be all over this one like a rash in the coming months. Stay tuned for more soon.

http://www.computerandvideogames.co.../news_story.php(que)id=89091

Additionally, I found this interesting tid bit on another forum:

About the game engine. The engine can support "unlimited terrain sizes" meaning that you could have an endless game enviroment. The engine uses Voxel Terrain, meaning it can be manipulated. Basically it will catch on fire and leave burn marks, blow craters, leave tire tracks, etc. It also uses Vectors on top of the Voxel Terrain. That means you can have buildings, etc, on top of the Terrain. The engine also uses Voxel Models. Voxel Models are not polys, but poly models can be converted to voxel models, meaning you can have as high a poly count as you want. The engine also supports Weather effects, true water effects (depth clarity) and wind water effects, meaning the water will be affected by the winds movement, etc

This sure does sound interesting IMO. It does sound somewhat impressive what the engine could be capable of if this is correct. What are your thoughts? What can we expect - is it likely to be become one of the next technological masterpieces (that perhaps was not thought possible on PS2-tech) or the next big letdown that at least had potential?

 

[London Geezer]

voxels on PS2... now thats interesting... really looking forward to see how this thing turn out to be...

 

[PC-Engine]

I hope for the sake of PS2 owners that this game will be more substance than hype. It does sound interesting though.

 

[McFly]

No screens till now?

Fredi

PS: Info about voxel gfx:
http://www.flipcode.com/voxtut/voxtut-struct.shtml

 

[Tagrineth]

I was wondering when someone would put all that EE muscle into something interesting.

 

[London Geezer]

I was wondering when someone would put all that EE muscle into something interesting.

exactly the same thoughts here... apart from nifty particle effects no one ever put some care into it... really wanna see this *water affected by wind* thing :?

 

[Cyborg]

I wouldnt put much faith in the quote you found on "another forum"

http://www.gamesdomain.com/gdreview/depart/aug97/aowint.html

3DI: Sure. The engine uses Voxel Terrain, meaning it can be manipulated. Basically it will catch on fire and leave burn marks, blow craters, leave tire tracks, etc. It also uses Vectors on top of the Voxel Terrain. This means we can have buildings, etc, on top of the Terrain. The engine also uses Sprites, and Voxel Models. Voxel Models are not polys, but poly models can be converted to voxel models, meaning you can have as high a poly count as you want. It only gives slight detail loss.

:?

 

[Qroach]

I think this is total BS. The biggest problem with voxels is that they require a lot of memory. PS2 doesn't have a lot of memory, so I can't really see someone wanting to create an entire terrain using this technique.

Even some of the last PC voxel games required 128-256 megs of ram since draw distance was important...

 

[vers]

I think this is total BS. The biggest problem with voxels is that they require a lot of memory. PS2 doesn't have a lot of memory, so I can't really see someone wanting to create an entire terrain using this technique.

Even some of the last PC voxel games required 128-256 megs of ram since draw distance was important...

no problem , div the terrain some blocks and read visible blocks from dvd

 

[Grall]

How about generating the terrain procedurally then? Too involved with voxels even for the mighty EE to cope?

*G*

 

[Evil_Cloud]

So this game will be in development for 4 years?
Sounds like a real triple A title to me...

 

[Qroach]

You could generate it procedurally, but I don't think you'll make a good memory saving. You still need to store the terrain you generated, unless you want random terrain generation.

 

[Thowllly]

You could generate it procedurally, but I don't think you'll make a good memory saving. You still need to store the terrain you generated, unless you want random terrain generation.

If you generate the terrain procedurally it will be "random", but you still don't have to store it. If you use the same seed number every time, you'll generate the same "random" terrain every time.

 

[Grall]

I should have been more exact. I didn't mean generate procedurally from a random generator function, but from higher-order surfaces. Some kind of splines or such could describe large expanses of terrain at a low memory cost I guess, and the tesselator could LOD the surface in real time too I guess, making the voxel mesh denser closer to the camera...

Provided the EE is up to the task of course.


*G*

 

[Qroach]

If you generate the terrain procedurally it will be "random", but you still don't have to store it. If you use the same seed number every time, you'll generate the same "random" terrain every time.

Of course you have to store it. Unless you want to genetrate a terrain that will change everytime you turn around or something (like a hill) would move off into the distance and vanish when you returned. that wouldnt' be very useful at al in a game.

Dispalying voxels on screen requires memory. i dont' think there any way around it from what I've seen.

 

[Saem]

I wonder if they're using some really fancy compression techniques to squeeze all the voxels in there.

 

[Jabjabs]

I wonder if they will get it above 10fps, voxels generally require loads of CPU power and can't hardware accelerated. If the frame rate is decent lets just hope they don't cut back on the voxel density.

 

[marconelly!]

Ahem, people... do you even read what is posted? Voxels thing was someone's joke.

 

[Thowllly]

If you generate the terrain procedurally it will be "random", but you still don't have to store it. If you use the same seed number every time, you'll generate the same "random" terrain every time.

Of course you have to store it. Unless you want to genetrate a terrain that will change everytime you turn around or something (like a hill) would move off into the distance and vanish when you returned. that wouldnt' be very useful at al in a game.

Dispalying voxels on screen requires memory. i dont' think there any way around it from what I've seen.

Yeah, the voxels you display at the moment would have to be in memory, but they don't all have to be in memory. If you regenerate the terrain you will get the same terrain! Let me explain.

To generate a terrain you need random numbers. To get random number on a computer you use some algorithm for converting one number into another number. Repeat the process to generate a series of numbers. But these numbers are not random! If you start out with the same number (the seed number) you will generate the same sequence of numbers each time. So if you use the same seed number you will generate the same terrain.

This was used in The Settlers. You could play on random terrains by entering a 16 digit number. If you got a terrain you liked, you could write down the number to use later, and the next time you used it you got exactly the same terrain. I even believe that the developers encouraged the players to send in codes for cool terrains.

 

[Phil]

If you generate the terrain procedurally it will be "random", but you still don't have to store it. If you use the same seed number every time, you'll generate the same "random" terrain every time.

Of course you have to store it. Unless you want to genetrate a terrain that will change everytime you turn around or something (like a hill) would move off into the distance and vanish when you returned. that wouldnt' be very useful at al in a game.

Dispalying voxels on screen requires memory. i dont' think there any way around it from what I've seen.

Yeah, the voxels you display at the moment would have to be in memory, but they don't all have to be in memory. If you regenerate the terrain you will get the same terrain! Let me explain.

To generate a terrain you need random numbers. To get random number on a computer you use some algorithm for converting one number into another number. Repeat the process to generate a series of numbers. But these numbers are not random! If you start out with the same number (the seed number) you will generate the same sequence of numbers each time. So if you use the same seed number you will generate the same terrain.

This was used in The Settlers. You could play on random terrains by entering a 16 digit number. If you got a terrain you liked, you could write down the number to use later, and the next time you used it you got exactly the same terrain. I even believe that the developers encouraged the players to send in codes for cool terrains.

If you'd use this method to regenerate the "same terrain" - wouldn't this be far too taxing on the system? I guess it really depends how big of a terrain we're looking at, but if it's of imense size, you wouldn't be able to maintain a steady framerate + all the other things that need to be calculated. I believe this is what Qroach was trying to point out - at least on todays systems...