I'm still working on an outdoor map but I'm hitting a wall as far as D3 capabilities. I know I can lightmap in external programs so I'd like to lightmap the outdoor sections of my map and combine them with a simple ambient light setup to sort of fake GI. However, I'm at a loss as to how to combine the lightmap textures with the regular textures in D3. I need a second UV set because a lot of my wall sections use repeating textures, and using only one set of UVs would mean my lightmaps would overlap as well.

http://forum.tgmonline.it/showthread.ph ... e-Lighting

I am essentially trying to do that, except without having to duplicate a ton of geometry.

Also, and this is just an aside, can someone give me a rundown on what the state of Doom3 projects and the backend shaders are? I'd like to be able to use some of Sikkpin's shaders in a standalone mod.

You can model whole map and bake AO to diffuse. But then your static lighting won't match your dynamic lighting. And regardless of the approach, you won't get any performance boost or any benefits of faking it in Doom 3.

idTech 4 is what it is, namely unified lighting engine. So either use that to max, or switch to another engine.

Project fake soft shadows directly from light sources. Not only you will be able to improve the scene but they will also interact with models without big performance lose.
Just remember to don't use hi-res textures for shadows. 512x to 256x is fine.

A long time ago I did that with blending textures on terrains and the results were satisfactory.
Call it an extra layer if you want. At least if you want to achieve the unified lighting rules...

You can always do what motorsep said, but the best thing would be to change engine.

Could you please elaborate on all of that?

Either way, without deferred shading (or rendering) and shadowmapping (to get shadows off CPU) idTech 4 bound to be limited to fugly hacks and small indoor spaces.

Second on what Sturm said, I have no idea what he's talking about but it sounds interesting. This is just more of a proof of concept thing to mess around with the engine. The goal is to make the shadows less jagged and make it look better when I combine it with Sikkpin's stuff to do some HDR and tone mapping. I'm still hoping that jmarshall23 will grace us with a deferred renderer. I am almost to the point where I'm going to have to lock myself into UDK because of programming but I thought I'd give idtech4 one last shot.

Why UDK? I mean, sure, tools are better, docs are better, community is larger, but after all, what's the ultimate goal?

Basically I nailed it down to 4 things I needed from an engine:

1) Cross platform.
2) Distributable tools for customs maps.
3) Non-crappy netcode for multiplayer and a dedicated server.
4) Reasonably modern graphics and tool chain.

UDK covers all of these now. CryEngine fails on points 1 and 3, and Unity fails on 2. Also, a bonus to the UDK is that there's pretty much everything I need to figure out already detailed in the forums or on UDN. I'm not a programmer, but Unrealscript comes with pretty much everything I need already in the packed in game code, I just need to make adjustments.

I'd like to use Doom 3, but the problems are insurmountable at the moment. The netcode was fixed in BFG, but now there's no dedicated server. If jmarshall23 ever releases anything that actually works (or even compiles) I still might, but for now I'd like to focus on making a game. This lightmapping thing is just a side project.

Also, I'd like to make a game on Linux using only open source tools, but that's kind of a pie in the sky thing at the moment. I doubt I'm going to find a replacement for things like Z-Brush or the Photoshop plugins I use on Linux.

To put it simply - you are wrong. UDK runs only on Windows and iOS. That's all. To go for Linux, OSX you need UE3 which cost $800k per title (plus royalties).

People don't give a damn about tools. They buy games to play. When you release your game and it will get traction and large user base, then you can start worrying about tools.

Multiplayer? I don't really need to go that far - check out Ravaged and Nexuiz and their empty servers. I don't know where it comes from, but the notion that your multuiplayer game will be popular is false. Even Xonotic (GPL Nexuiz) successor doesn't have a lot of players (that's a free game with userbase from early 2000s).

I am not saying UDK is bad, not at all. But there are 2 things that really bug me - Windows only and royalties to pay from any product associated with the title you license UDK for.

As far as visuals, I highly doubt you can compete with Crysis 3 / CoD / BF3. Just because they can use best art houses from all over the world. So 200+ good artist can easily beat small team of prodigy artists. Not to mention timeframe.

UDK actually cooks to OSX right now. I've tried it, and it works, no 800k license needed. Also, as I sit here Dungeon Defenders, a UDK game, is running on my Linux desktop.

http://i.imgur.com/3uhqN.jpg - there's proof.

As far as licensing, this is a non-commercial project that I'm using to learn how to make a game, I'm not planning to sell it, just release it to the community. I am planning to use the experience I gain from doing this game to get into working for larger teams.

Now, as far as multiplayer is concerned the reason that people don't play Nexuiz or Xonotic is because they are just copies of Quake 3. It's much easier to just play Quake Live. I hate to say that open source FPS games are pretty far behind the times when it comes to actual gameplay, and everyone seems to be just rehashing old arena games that came out in 1996. I never liked arena games in the first place, and I'm not making one now. Multiplayer is still vastly popular in a lot of indie games. Dungeon Defenders is a good example of a really good game made by a small team that is multiplayer and sells well.

Now, about royalties, I'd say the 25 percent is worth it for me, because UDK saves at least 25 percent of my time. Cryengine saves more because it's realtime, but because of the problems listed above I chose UDK. I like idtech engines, but the tools are 10 years old at best. That alone is a reason to use UDK. Just the vertex / mesh painting alone is worth the 25 percent compared to Doom3.

Anyway, I'll finish by adding that you need to realize that *not everyone is making a game for your reasons.* I bought and liked Steel Storm, and judging from the teasers I like what you've done in Steel Storm 2. I was planning to ask you how you did the weather and terrain for that flyby scene. Hell, I had even thought about offering my help in mapping and modeling just for the experience. However, every post I read from you has a sort of "my way or the highway" undercurrent, like you can't fathom why anyone would make other choices than the ones you've made. Maybe that is a good attitude to get things done as an indie developer, but it alienates a lot of people.

Dungeon Defenders is UE3.

http://www.unrealengine.com/en/platforms/ << in Black and White (or rather white on gray in the top right corner): Windows and iOS for UDK.

As far as royalties, do the math. Unless you live in some third world country, and you game makes _a lot_ of money, 25% is something you will need in a long run. Don't forget Steam's share (other distributors take more in some cases). Don't forget you will have a split with your team, and taxes.

However for a free game, as learning project to break into the industry UDK is a way to go (or CE3 or both )

I appreciate your support, and I am glad you like what we do. Business is business. I am not the only developer who understands the difference between modding and commercial game development where "my way or highway" work in favor of project being done and being consistent. If I were to run a company and not develop games, I might have left all the creative work to people who work on the project. But since I am the one who leads the project, I have final word of what goes and what doesn't. It's just a natural order of things. Some people get it and we work together and some don't.

If team has a solution for a problem, or it's out of my area of expertise, I go for it.

I completely expect you to have that attitude when it comes in regard to the game project you're working on, it's the only way to get things done with any efficiency while working in a small team. If I came and worked on Steel Storm 2, for instance, I'd expect to be told what needed to be done, it's your team. However, what I'm talking about is when you're posting about others' projects.

Aye. Perhaps I just come out wrong way because of the info posted, so I apologize. For example when you posted that your goal is different (learning to get into the industry, to work for existing company I assume), then yeah, my outlook was wrong as I was talking about small team working on commercial project.

Still, when I see wrong info, I'd like to clarify things, like with misconceptions surrounding UDK when it comes to platforms, licensing, etc. for indie projects.

Apology accepted, thanks. I'm not personally hurt or anything by being disagreed with but sometimes your mannerisms come off as a bit too harsh for a lot of people on the forums. We all really want you to succeed and are here to help (admittedly I can't really help you with much). If Steel Storm 2 succeeds then it brings more interest in idtech4 in general and perhaps some of the major problems can be fixed, and hell, maybe one of us can get a job on a new team trying to make something out of idtech4.

I'm still working on an outdoor map but I'm hitting a wall as far as D3 capabilities. I know I can lightmap in external programs so I'd like to lightmap the outdoor sections of my map and combine them with a simple ambient light setup to sort of fake GI. However, I'm at a loss as to how to combine the lightmap textures with the regular textures in D3. I need a second UV set because a lot of my wall sections use repeating textures, and using only one set of UVs would mean my lightmaps would overlap as well.

http://forum.tgmonline.it/showthread.ph ... e-Lighting

I am essentially trying to do that, except without having to duplicate a ton of geometry.

Also, and this is just an aside, can someone give me a rundown on what the state of Doom3 projects and the backend shaders are? I'd like to be able to use some of Sikkpin's shaders in a standalone mod.

You can model whole map and bake AO to diffuse. But then your static lighting won't match your dynamic lighting. And regardless of the approach, you won't get any performance boost or any benefits of faking it in Doom 3.

idTech 4 is what it is, namely unified lighting engine. So either use that to max, or switch to another engine.

Project fake soft shadows directly from light sources. Not only you will be able to improve the scene but they will also interact with models without big performance lose.
Just remember to don't use hi-res textures for shadows. 512x to 256x is fine.

A long time ago I did that with blending textures on terrains and the results were satisfactory.
Call it an extra layer if you want. At least if you want to achieve the unified lighting rules...

You can always do what motorsep said, but the best thing would be to change engine.

Could you please elaborate on all of that?

Either way, without deferred shading (or rendering) and shadowmapping (to get shadows off CPU) idTech 4 bound to be limited to fugly hacks and small indoor spaces.

Second on what Sturm said, I have no idea what he's talking about but it sounds interesting. This is just more of a proof of concept thing to mess around with the engine. The goal is to make the shadows less jagged and make it look better when I combine it with Sikkpin's stuff to do some HDR and tone mapping. I'm still hoping that jmarshall23 will grace us with a deferred renderer. I am almost to the point where I'm going to have to lock myself into UDK because of programming but I thought I'd give idtech4 one last shot.

Why UDK? I mean, sure, tools are better, docs are better, community is larger, but after all, what's the ultimate goal?

Basically I nailed it down to 4 things I needed from an engine:

1) Cross platform.
2) Distributable tools for customs maps.
3) Non-crappy netcode for multiplayer and a dedicated server.
4) Reasonably modern graphics and tool chain.

UDK covers all of these now. CryEngine fails on points 1 and 3, and Unity fails on 2. Also, a bonus to the UDK is that there's pretty much everything I need to figure out already detailed in the forums or on UDN. I'm not a programmer, but Unrealscript comes with pretty much everything I need already in the packed in game code, I just need to make adjustments.

I'd like to use Doom 3, but the problems are insurmountable at the moment. The netcode was fixed in BFG, but now there's no dedicated server. If jmarshall23 ever releases anything that actually works (or even compiles) I still might, but for now I'd like to focus on making a game. This lightmapping thing is just a side project.

Also, I'd like to make a game on Linux using only open source tools, but that's kind of a pie in the sky thing at the moment. I doubt I'm going to find a replacement for things like Z-Brush or the Photoshop plugins I use on Linux.

To put it simply - you are wrong. UDK runs only on Windows and iOS. That's all. To go for Linux, OSX you need UE3 which cost $800k per title (plus royalties).

People don't give a damn about tools. They buy games to play. When you release your game and it will get traction and large user base, then you can start worrying about tools.

Multiplayer? I don't really need to go that far - check out Ravaged and Nexuiz and their empty servers. I don't know where it comes from, but the notion that your multuiplayer game will be popular is false. Even Xonotic (GPL Nexuiz) successor doesn't have a lot of players (that's a free game with userbase from early 2000s).

I am not saying UDK is bad, not at all. But there are 2 things that really bug me - Windows only and royalties to pay from any product associated with the title you license UDK for.

As far as visuals, I highly doubt you can compete with Crysis 3 / CoD / BF3. Just because they can use best art houses from all over the world. So 200+ good artist can easily beat small team of prodigy artists. Not to mention timeframe.

UDK actually cooks to OSX right now. I've tried it, and it works, no 800k license needed. Also, as I sit here Dungeon Defenders, a UDK game, is running on my Linux desktop.

http://i.imgur.com/3uhqN.jpg - there's proof.

As far as licensing, this is a non-commercial project that I'm using to learn how to make a game, I'm not planning to sell it, just release it to the community. I am planning to use the experience I gain from doing this game to get into working for larger teams.

Now, as far as multiplayer is concerned the reason that people don't play Nexuiz or Xonotic is because they are just copies of Quake 3. It's much easier to just play Quake Live. I hate to say that open source FPS games are pretty far behind the times when it comes to actual gameplay, and everyone seems to be just rehashing old arena games that came out in 1996. I never liked arena games in the first place, and I'm not making one now. Multiplayer is still vastly popular in a lot of indie games. Dungeon Defenders is a good example of a really good game made by a small team that is multiplayer and sells well.

Now, about royalties, I'd say the 25 percent is worth it for me, because UDK saves at least 25 percent of my time. Cryengine saves more because it's realtime, but because of the problems listed above I chose UDK. I like idtech engines, but the tools are 10 years old at best. That alone is a reason to use UDK. Just the vertex / mesh painting alone is worth the 25 percent compared to Doom3.

Anyway, I'll finish by adding that you need to realize that *not everyone is making a game for your reasons.* I bought and liked Steel Storm, and judging from the teasers I like what you've done in Steel Storm 2. I was planning to ask you how you did the weather and terrain for that flyby scene. Hell, I had even thought about offering my help in mapping and modeling just for the experience. However, every post I read from you has a sort of "my way or the highway" undercurrent, like you can't fathom why anyone would make other choices than the ones you've made. Maybe that is a good attitude to get things done as an indie developer, but it alienates a lot of people.

Dungeon Defenders is UE3.

http://www.unrealengine.com/en/platforms/ << in Black and White (or rather white on gray in the top right corner): Windows and iOS for UDK.

As far as royalties, do the math. Unless you live in some third world country, and you game makes _a lot_ of money, 25% is something you will need in a long run. Don't forget Steam's share (other distributors take more in some cases). Don't forget you will have a split with your team, and taxes.

However for a free game, as learning project to break into the industry UDK is a way to go (or CE3 or both )

I appreciate your support, and I am glad you like what we do. Business is business. I am not the only developer who understands the difference between modding and commercial game development where "my way or highway" work in favor of project being done and being consistent. If I were to run a company and not develop games, I might have left all the creative work to people who work on the project. But since I am the one who leads the project, I have final word of what goes and what doesn't. It's just a natural order of things. Some people get it and we work together and some don't.

If team has a solution for a problem, or it's out of my area of expertise, I go for it.

I completely expect you to have that attitude when it comes in regard to the game project you're working on, it's the only way to get things done with any efficiency while working in a small team. If I came and worked on Steel Storm 2, for instance, I'd expect to be told what needed to be done, it's your team. However, what I'm talking about is when you're posting about others' projects.

Aye. Perhaps I just come out wrong way because of the info posted, so I apologize. For example when you posted that your goal is different (learning to get into the industry, to work for existing company I assume), then yeah, my outlook was wrong as I was talking about small team working on commercial project.

Still, when I see wrong info, I'd like to clarify things, like with misconceptions surrounding UDK when it comes to platforms, licensing, etc. for indie projects.

Apology accepted, thanks. I'm not personally hurt or anything by being disagreed with but sometimes your mannerisms come off as a bit too harsh for a lot of people on the forums. We all really want you to succeed and are here to help (admittedly I can't really help you with much). If Steel Storm 2 succeeds then it brings more interest in idtech4 in general and perhaps some of the major problems can be fixed, and hell, maybe one of us can get a job on a new team trying to make something out of idtech4.