Free or nonexpensive 3D game engines?

[Kaftan Barlast]

We've just started planning a new action-adventure game and we're in the proccess of trying to decide which engine to use. I was just wondering if theres anyone here who could recommend somthing that perhaps, we've overlooked. We cant afford to pay more than a couple of hundred dollars per license, though. Also, the game is very much commercial so the engine license will have to allow full commercial use.

 

 

Thus far we have looked at

 

Torque3D - Too many faults to count really. unuseable.

Ogre - Just a graphics engine, requires a lot of work to be useful

Nebula2 - DirectX only, lacks proper level editor, only works with Maya7

Source - Major licensing and legal issues

Unity - Is MacOS only

C4 - Seems very good, does all we need it to but we havent heard of anyone using it, and thus not the bad things

 

 

If you've worked with any of the above id love to hear your opinion, but if you have other recommendations that would would be great.

[Kaftan Barlast]

We've also been offered to use a commercial engine that's quite amazing, but they will not give us any direct support. Is it as stupid as it seems to try and use an engine with the in-house documentation only?

 

 

 

This is coincidentally, one of those times I wished the gaming industry didnt have so many silly demands for secrecy. Else, I would be expressing emotions of joy and amazement at said offer right now.

Edited January 19, 2008 by Kaftan Barlast

[@\NightandtheShape/@]

We've just started planning a new action-adventure game and we're in the proccess of trying to decide which engine to use. I was just wondering if theres anyone here who could recommend somthing that perhaps, we've overlooked. We cant afford to pay more than a couple of hundred dollars per license, though. Also, the game is very much commercial so the engine license will have to allow full commercial use.

 

 

Thus far we have looked at

 

Torque3D - Too many faults to count really. unuseable.

Ogre - Just a graphics engine, requires a lot of work to be useful

Nebula2 - DirectX only, lacks proper level editor, only works with Maya7

Source - Major licensing and legal issues

Unity - Is MacOS only

C4 - Seems very good, does all we need it to but we havent heard of anyone using it, and thus not the bad things

 

 

If you've worked with any of the above id love to hear your opinion, but if you have other recommendations that would would be great.

 

This is a tricky subject... Naturally, I assume that you have a small coding team. Although my I add a small coding team is certainly capible of producing a damn fine game engines, but it requires alot of time.

 

It would help if you'd stated exactly what you require from the engine. (platform surpport, specific features etc...)

 

I've had some experience with Torque and I had an interesting conversation with one of the garage games guys at gamefest 2007, where I basically made my disdain for the engine known, although that isn't to say that the engine itself is completely cack. It's a buggy engine from my experience, and a bitch to use, I hated the scripting language with a passion. So I certainly understand any aprehension in regards to its use.

 

I'd say it's a no go for me personally, but as far as completeness goes it's actually fairly powerful, and certainly more useable than some of the other engines you've stated. As the licence allows for direct source code access then I wouldn't really count it out, a modified Torque engine may be infact exactly what you're looking for but it depends entirely upon your coding team, and your needs. So don't count it out as far as it goes it's a fairly complete product with alot of support.

 

Ogre is widely known as a render engine, and other than the fact that it provides the basic gubbins it could be quickly considered to be fairly useless, that said! It should be an attractive concept as it leaves the rest of the systems open to be designed and integrated as needed, you'd certainly have to work on it alot before you saw any results but it's likely that you'd end up yielding some very good results with it.

 

Never heard of Nebula2, or Unity... And I'd certainly say that Source is likely to be a no go for you guys, as you're not intending to make a mod, so the issues you've stated soon crop up.

 

I've taken a look at C4's capibilities, it's okay, there isn't anything outstanding about it, but it seems to have a decent enough tool chain, that said I have no direct experience with it. I wouldn't use it, but that's me... I'd just write my own. That said it does indeed seem to be what you're looking for, in regards to completeness, I don't know how stable the engine is, but it certainly seems friendly enough, and the animation/model format is acceptable, I've come across collada before it's not a complex format, but certainly a capible format. The sound engine appears to be robust, and you have some typical and commonplace shader features in there.

 

Comprehensive bump mapping support. (Which you'd want)

Enhanced parallax mapping. (Nice)

Horizon mapping. (This is basically a type self shadowing)

Ambient lighting volumes. (DUH!)

Gloss-mapped specular reflections. (heh some phong based stuff which works off a texture most likely)

Cube environment mapping. (Claps hands, been around since the dark ages)

Cook-Torrance microfacet surface reflection. (That sounds more impressive than it is, I wrote such a shader a few weeks ago, except I integrated it directly with bump mapping, which gave an awesome result, but cook torrance lighting is fairly good)

Bumpy reflection and refraction. (Nothing astounding about that)

Realistic water shading. (A day's work)

 

^ That's an average bunch. It just about fulfills the list of requirements from my own perspective. That said the material system in my own engine is more advanced.

 

The lighting and shadowing engine looks decent enough...

 

I wish i could say what your results would be like... But unless you have coders who wanna get your hands dirty it may be a little less than what you need.

 

Have you looked at Irrlicht? Its also supposed to be good, but again it'll require getting your hands dirty.

[@\NightandtheShape/@]

We've also been offered to use a commercial engine that's quite amazing, but they will not give us any direct support. Is it as stupid as it seems to try and use an engine with the in-house documentation only?

 

It depends upon skill, some people could, some couldn't. It may require alot of time and patience. What engine is it may I ask?

[Kaftan Barlast]

Im a bit out of my element in this, because Im an artist of the non-technical kind so I cant go into the specifics too much. But the game we're going to make is an action-avdenture platformer with a stylizd cartoony art design a little like psychonauts.

 

I just got a message that we're also thinking about Truevision3D, and at first glance it sems like a top notch engine for the price.

 

 

We've also been offered to use a commercial engine that's quite amazing, but they will not give us any direct support. Is it as stupid as it seems to try and use an engine with the in-house documentation only?

 

It depends upon skill, some people could, some couldn't. It may require alot of time and patience. What engine is it may I ask?

 

 

It's under NDA so I cant tell you exactly, but I am allowed to say that it is one of the big ones. I imagine it would be heaven for us artists, but sheer torture for the programming team.

[@\NightandtheShape/@]

Im a bit out of my element in this, because Im an artist of the non-technical kind so I cant go into the specifics too much. But the game we're going to make is an action-avdenture platformer with a stylizd cartoony art design a little like psychonauts.

 

I just got a message that we're also thinking about Truevision3D, and at first glance it sems like a top notch engine for the price.

 

 

We've also been offered to use a commercial engine that's quite amazing, but they will not give us any direct support. Is it as stupid as it seems to try and use an engine with the in-house documentation only?

 

It depends upon skill, some people could, some couldn't. It may require alot of time and patience. What engine is it may I ask?

 

 

It's under NDA so I cant tell you exactly, but I am allowed to say that it is one of the big ones. I imagine it would be heaven for us artists, but sheer torture for the programming team.

 

Okay well... When you say one of the big ones, I only ever think of three engines (That's not to say its one of those). I also understand that NDA make it difficult to speak about anything directly.

 

You'll be wanting direct control over pixel and vertex shader to help aid in a cartoony result... The ability to tweak shaders is second to none really, psychonaughts infact has a very basic lighting system, lots of diffuse and ambient stuff, not really any specular colouring. Where TF2 uses specular in conjunction with other lighting methods to create the look they got.

 

Truevision3D certainly is ticking all my boxes. The features are exactly what I would aim to integrate myself into a robust multipurpose system.

 

I'd just like to say a few words as to what it is that actually appeals to me about this engine, and I'll go through it's feature headings and try not to be too technical, but your programming team should enjoy the environment as much as the artists.

 

Render System - The main thing I like is that it isn't bloated, which means it supports the basic run of the mill features, which if one was so inclined could be gutted out of the engine. It's also an adequtte start point which allows for controlled expansion.

 

HLSL Shader Support - You really really need this, especially if the game is supposed to be very artistic.

 

Landscape and Terrain System - Now I obviously don't know if you'll be requiring landscapes, but a good terrain system will always go far. The system is robust, more than adequtte and ready to use out of the box, thus the level designer should rejoice.

 

Static Mesh Support - It may sound silly but good static mesh support is always useful, heck you could make a high poly level, and slap a low polygon collision mesh over it with this system in next to no time.

 

Animated Meshes (Actors) - This one is so important, especially if you're making any sort of game where it's character dependant, animation blending it perhaps the one thing that stands out to me personally as it makes the whole process of animating a character 10 times more streamlined than a system without. It supports everything you would ever need. Morph targets should be great for building a lip sync system so you could have some nice cinematics.

 

Material and Lighting System - Does everything you'll need, but you may have to expand it slightly.

 

Particle Systems Everyone thinks about emitters when they talk about particle systems, but attractors are also extremely useful. Althought, the first thing I'd do personally is see what the numbers are when it breaks on an average system. Also, there doesn't seem to be any collision support, it's not something you'd wanna over do but... It's one of those features I tend to smile about and say yeah... I'd like the option for certain emitters. It does on the otherhand support small meshes, so you can actually output a mesh as a particle.

 

GPGPU (General Processing on GPU) - The programmers will enjoy that.

 

Integrated Newtonian Physics Engine - Fairly decent physics engine, I mean its no Havok, but it is certainly good enough.

 

Built-in Special Effects and Controls - It already has some nice post processing features like bloom, glow, motion blur and depth of field. No HDR tho'

 

Supported Languages and Compilers - Language support is excellent! especially when a designer has one of those designer moments and wants to get an idea of a feature, code it up in something nasty and prototype, it can solve arguements and allow for design desicions to take place during development, because nobody gets everything down first time, and idea's don't always work.

 

It's a good engine.

 

If the commercial engine has more features, then you should use that. There will be a huge learning curve but actual commercial game engine experience would be very very helpful for you guys in the long run. I don't know how talented your code team is, but if they're scared of a commercial engine, because they won't have support then I'd be generally scared in regards to their abilities, to me personally there is something wrong with programmers who are afraid to sink their teeth into something and understand why something is written the way it has been, why it may even be hard to read in general, and OMG why there indeed may be some inline ASM laying around. One of the things that it is important to do as a programmer is recognise elegant solutions to a problem, and not get bogged down in the thinking that your way is always the best way.

 

Truevision3D is certainly a good second option if you have pansy programmers who are quaking in their boots over using an industry engine. There are a few things missing from the engine mind, and it's mostly in regards to spatial organisation, so that is a job you may want to work at straight away on the coding side.

[Anthix]

We've just started planning a new action-adventure game and we're in the proccess of trying to decide which engine to use. I was just wondering if theres anyone here who could recommend somthing that perhaps, we've overlooked. We cant afford to pay more than a couple of hundred dollars per license, though. Also, the game is very much commercial so the engine license will have to allow full commercial use.

 

 

http://crystalspace3d.org + CEL makes for a pretty rich setup...and it's free.

[Kaftan Barlast]

I think we're going with Truevision3D, it can do everything we've tried with it so far and it looks really good. It lacks a level editor though, so we're working on using 3Ds MAX as an editor and exporting through some sort of ASCII exporter. Inconvenient, but we dont have time to write an editor and apparently the one are available wont work with what we're trying to do.

[@\NightandtheShape/@]

I think we're going with Truevision3D, it can do everything we've tried with it so far and it looks really good. It lacks a level editor though, so we're working on using 3Ds MAX as an editor and exporting through some sort of ASCII exporter. Inconvenient, but we dont have time to write an editor and apparently the one are available wont work with what we're trying to do.

 

3DS MAX while my pet hate(What smartarse thought Z was up) is perfectly adequette, I personally use XSI when it comes to import/export functionality.

 

Also, here's something that with some clever coding and manipulation may be useful, you can use null's/dummy's to provide points of interest. Via some clever coding and strict naming practices you could potentially assign shedloads of cool things to those positions... i.e. Lighting, particle emitters...

 

For example, by prefixing all the potential point_light sources with for example PL, then some colour info there are a number of ways this can be done, lets just keep it simple and say you have some predefined colours Red, green etc... Then some falloff information say... and perhaps a name tag.

 

What you end up with is PL_RED_10.0_LIGHT1.

 

When you load any "Level" model you can place a point light that emits a predefined red light with a falloff of 10, and also has an internal tag of LIGHT1. This could be done for an entire level, then all you need to do is some basic culling. What you end up with is an increase in your load time, but the benefits can be fairly cool.

 

This certainly makes MAX more effective in regards to level editing. It applies in general to all 3D software packages, infact it's something I've made use off extensively, not in the manner of lighting, but in the manner of particle emission points... The the ability to build your own tool chain and engine greatly increases the amount of stuff you can do.

 

Anyways, that's one fairly good way you can make max more useful as a level editing tool.

[Kaftan Barlast]

That is more or less exactly what we are doing at this point. Im not especially thrilled about this system since we're working blind most of the time, but it works and thats great. Currently, Im working on visual design and leaving the tech stuff to the programmers.

 

 

We wont be using the "big secret engine" for this project, but I have been talking to our producer about getting access to some of the source code to the render engine and shaders so that the programmers might be able to take some of that and incorporate it into our game. Being the artsy type, I dont even know if its possible to do that, but I imagine that it would atleast be a great learning experience for our programming team to be able to take a look at some AAA-class stuff

[@\NightandtheShape/@]

That is more or less exactly what we are doing at this point. Im not especially thrilled about this system since we're working blind most of the time, but it works and thats great. Currently, Im working on visual design and leaving the tech stuff to the programmers.

 

 

We wont be using the "big secret engine" for this project, but I have been talking to our producer about getting access to some of the source code to the render engine and shaders so that the programmers might be able to take some of that and incorporate it into our game. Being the artsy type, I dont even know if its possible to do that, but I imagine that it would atleast be a great learning experience for our programming team to be able to take a look at some AAA-class stuff

 

That would likely be an awesome experience. But also extremely challenging, it's one of those square pegs round holes moments. It can be done, but it ain't easy and would most likely be a serious source of bugs.

[WILL THE ALMIGHTY]

Isn't the Spring Engine freeware?

 

Though I suppose it's crap by today's standards...

[Kaftan Barlast]

We've started to work with Truevision3D now, we moved into our new sweet office today even. It's smelly and on campus, but its a miracle* we managed to persuade the university to give it to us.

 

 

 

 

* our producer may or may not possess compromosing video footage of drunken faculty members

[@\NightandtheShape/@]

We've started to work with Truevision3D now, we moved into our new sweet office today even. It's smelly and on campus, but its a miracle* we managed to persuade the university to give it to us.

 

 

 

 

* our producer may or may not possess compromosing video footage of drunken faculty members

 

Well keep me posted Kaften me olde, I'm interested in seeing if it goes well for you!