Procedural voxel terrain (1 man team)

This forum is for anything that doesn't specifically have to do with Better Than Wolves
Post Reply
User avatar
Luciferous Leo
Posts: 20
Joined: Mon Nov 05, 2012 6:36 pm
Location: New Zealand

Procedural voxel terrain (1 man team)

Post by Luciferous Leo »


Only video where he explains his process.

This man has been working on his procedural voxel engine for a while (you've probably seen it in earlier versions).
But recently he has added some sandbox editing (minecraft style in cubes of various sizes as well as curves


His work is amazing and interesting to read, so I just thought I'd share it with you all.

http://procworld.blogspot.com/
User avatar
Wafflewaffle
Posts: 369
Joined: Sat Nov 03, 2012 7:17 pm
Location: Carnaval land

Re: Procedural voxel terrain (1 man team)

Post by Wafflewaffle »

I am by no means tech savy. I've read some of the commotion when this voxel engine was first announced but people said it was not entirely game changing as it proposed itself to be. Never quite understood why.


Refresh my memory will ya? This voxel engine is "better" then the regular polygon based because the amount of detail placed in say 1 cubic meter of digital space doesnt interfere with its processing speed right? I probably got it all wrong, if someone would like to explain please do as if you are talking to a infant. thx!
Oh great, now nothing can stop the inbred train

Paradox Interactive:
CHOO CHOO!
User avatar
FlowerChild
Site Admin
Posts: 18753
Joined: Mon Jul 04, 2011 7:24 pm

Re: Procedural voxel terrain (1 man team)

Post by FlowerChild »

A big part of it is the procedural part, in that an artist doesn't have to hand draw each and every in game object for them to have a varied and unique appearance.

With something like this, each rock can look different, and no one has to ever create a specific texture or geometry for it, or place it by hand.

It basically allows for an extremely varied and organic looking world, in a way that wouldn't be feasible normally due to the development effort it would require.

Put another way, it's like applying the concepts of MC's world generation to each individual in game object.

In terms of detail, it's again like MC. If instead of being made up of polygons, the world is made from small (not 1m cube) voxels, it allows for a level of detail you don't generally get with polygon engines, and it also allows for them to be more easily modified on the fly (say through player interaction), just like with Minecraft.

It's not faster. Not by any means. The reason why engines like this don't exist is because the processing horsepower they require far exceeds what is need by a polygon engine. The big deal about something like this is what it makes possible *if* someone gets it to run fast enough to be practical, NOT that it is somehow inherently faster than a polygon approach.

This is all theoretical of course, and I'm just talking about what he's trying to achieve here. As per usual with such things "I'll believe it when I see it", and I never get excited about stuff like this just based on videos.

In short though, it's the difference between making a world out of atoms, or making it out of geometric solids that can not be further subdivided. Or, to put it yet another way...it's like super high-res Minecraft.
User avatar
Foxy Boxes
Posts: 421
Joined: Sun Apr 01, 2012 1:52 pm

Re: Procedural voxel terrain (1 man team)

Post by Foxy Boxes »

I'd like to note that is has also been done many a time on a smaller scale, say units in a RTS, but to apply it to something as expansive as terrain? Wow.
On the internet you can be whatever you want. It's surprising so many people choose to be stupid.
User avatar
Shengji
Posts: 638
Joined: Sun Jul 10, 2011 8:35 pm

Re: Procedural voxel terrain (1 man team)

Post by Shengji »

Foxy Boxes wrote:I'd like to note that is has also been done many a time on a smaller scale, say units in a RTS, but to apply it to something as expansive as terrain? Wow.
One of my earlier games was the original Delta Force, it used a voxel engine and we experimented with a proc gen approach. Needless to say, we couldn't get it close to being optimised enough, though I'd love to try it again on modern machines (This must have been 20 years ago), although it was nowhere near as sophisticated a voxel engine as minecraft and the likes.
7 months, 37 different border checks and counting.
User avatar
FlowerChild
Site Admin
Posts: 18753
Joined: Mon Jul 04, 2011 7:24 pm

Re: Procedural voxel terrain (1 man team)

Post by FlowerChild »

Shengji wrote: One of my earlier games was the original Delta Force, it used a voxel engine and we experimented with a proc gen approach. Needless to say, we couldn't get it close to being optimised enough, though I'd love to try it again on modern machines (This must have been 20 years ago), although it was nowhere near as sophisticated a voxel engine as minecraft and the likes.
Wait a second, you worked on the first Delta Force?

Dude...I have very fond memories of that game. I played the shit out of it in online multiplayer. If I remember right, SAW + claymores was my thing, and it offered a gameplay experience unlike anything else out there at the time.

Kicking it old-school ftw.
User avatar
Shengji
Posts: 638
Joined: Sun Jul 10, 2011 8:35 pm

Re: Procedural voxel terrain (1 man team)

Post by Shengji »

FlowerChild wrote: Wait a second, you worked on the first Delta Force?

Dude...I have very fond memories of that game. I played the shit out of it in online multiplayer. If I remember right, SAW + claymores was my thing, and it offered a gameplay experience unlike anything else out there at the time.

Kicking it old-school ftw.
Indeed, Novalogic were the last games company who didn't make me want to slit my wrist rather than make it in to work and that game was very special - the SAW was amazing, completely over the top but pure joy to use!
7 months, 37 different border checks and counting.
User avatar
Magnavode
Posts: 42
Joined: Fri Sep 14, 2012 6:51 pm

Re: Procedural voxel terrain (1 man team)

Post by Magnavode »

I love those dev blogs. Here's another one: Atomontage Engine
Ozziie
Posts: 128
Joined: Tue Sep 11, 2012 2:41 am

Re: Procedural voxel terrain (1 man team)

Post by Ozziie »

Shengji wrote:the SAW was amazing, completely over the top but pure joy to use!
Wow man, Delta Force was the first FPS I ever played :) Loved that game!

Funnily enough, the wikipedia page mentions "Delta Force was praised for emphasizing "gameplay over graphics" in its time." Sound familiar? ;)
User avatar
FlowerChild
Site Admin
Posts: 18753
Joined: Mon Jul 04, 2011 7:24 pm

Re: Procedural voxel terrain (1 man team)

Post by FlowerChild »

Shengji wrote:Indeed, Novalogic were the last games company who didn't make me want to slit my wrist rather than make it in to work and that game was very special - the SAW was amazing, completely over the top but pure joy to use!
Oh wow man...respect. I was in the industry at the time too, and I really dug what you guys did there.

Had no idea you were a veteran yourself :)

Loved the gameplay, but I remember I was also keeping a close eye on Novalogic at the time as I was also experimenting with voxel/height-map engines back then. These were the days of Magic Carpet (Bullfrog), Terra Nova (Origin I believe) and of course Commanche (which you guys also did) right before 3DFX blew that whole development trend right out of the water with the release of the Voodoo. I think I first picked up Delta Force to check out your technology at work, then wound up playing the shit out of it at home when I realized how much fun it was.

I've always loved voxels. It figures I'd be working on a game involving them 20 years later when they finally resurfaced again :)
User avatar
Luciferous Leo
Posts: 20
Joined: Mon Nov 05, 2012 6:36 pm
Location: New Zealand

Re: Procedural voxel terrain (1 man team)

Post by Luciferous Leo »

Magnavode wrote:I love those dev blogs. Here's another one: Atomontage Engine
Wow, that looks amazing. The LOD they've implemented is brilliant.
Post Reply