What other games on the market have been using this technology?
I give up.
- Thread starter Richard Bong
- Start date
Almost everybody. There are a couple of competitive technologies, but not nearly as successful. I used to work in simulation development (as opposed to game development... they are close but have different objectives)
The technology itself has been available to implement since some time before 2003., and is a mature technology. I originally saw it as part of Blender, in 2007. But some of the other tools also got their start around the same time. If you're up to the task of learning how the tech works, you can just download Blender. It's free, but has a fairly steep learning curve.
Unity3d incorporates this into their game developer tools, which IIRC are free until you reach some income level from games you develop. You can import your own mesh i.e. from Blender, one of it's competitors, or from their own store, or any other store that gives you a license to do so. You can add color maps using tools like GIMP (free open source), or Photoshop (Adobe has a bunch of tools) or any of their competitors.
There are other implementations of the tech as well. I mean what game developers need mostly depends on what can they punch through their game development pipeline, but mostly they use some version of a mesh development tool.
Thanks, I was just not aware and appreciate the explanation. I find it to be really interesting stuff, just dont know much about it.
So what is CIG's work on planetary meshes doing above and beyond this technology that we should appreciate when we see it in game?
Is this explained in one of the Citizencon videos or a Star Citizen Live?
So what is CIG's work on planetary meshes doing above and beyond this technology that we should appreciate when we see it in game?
Is this explained in one of the Citizencon videos or a Star Citizen Live?
Last edited:
I'll say, the level of detail (LOD) is awesome both in the planets and moons, but also the ships and other game assets.
It's not so great with in-game character development, hair styles (are a joke) and other things like rendering flesh, even how they walk have big problems that border on a bigger problem.
This has to do with how human game characters appear to real live humans. If you try to make characters too close to human, you better be perfect. Approximations of humans are things that either need to be greater than 95%, or less than 60% accurate - or they can give the audience a reaction called "uncanny valley". It's akin to the reaction people who have been to an open casket funeral of a friend or relative can have when the makeup artists didn't do a very good job.
In most game engines the deformation capability is built in. Once the pieces are streamed in, and deformations applied, then the deformed objects are streamed to the clients.
Last edited:
So far, they aren't up to the level I described. Granted Carrara only, realy, has VUE as a competitor for terrain and landscape generation, they are specifically designed to do this work. When designed Carrara and Vue's terrain systems were only useful in gaming at very low resolution, but technology has caught up. (They have both been around for a while.) Astrosynthesis and Fractal Terrains where originally designed to help Game Masters build maps for tabletop RPG's. However the ability to output to Bryce made the workflow work. The current version of Fractal Terrains can output height maps directly, though I haven't personally tried it, so I don't know how well it works.
It is my observations based on the pyro videos and in game planets and moons. There are tiling errors in the textures. The shaders don't properly handle slope and altitude. The volumetric clouds in Pyro, are place holders. And, of course, the unfinished state, not just of the Pyro system but also the Stanton system. By now the entire verse could, and probably should be complete, minus some polish.
Don't get me started on the avatar system. :)
I've no idea if anyone is. The licensing doesn't require recognition. I would suspect ED is using part of the workflow. They turned out a lot of systems in a very short period of time. I've seen demos with both Unity and Unreal of the Carrara export, but don't know if those demos ever turned into anything. I do know, from personal use, the workflow works, since I used it to create visuals for my Traveller game sessions back in the day.
So they reinvented the environment building tech to their own tastes and needs. I'm fine with that.
What I'm not fine with is the fact that they have been advertising the tools they made trough multiple episodes of their online shows and during CitCon tech panels for years now as a huge time saver!
Yet we see no speed up. Even if they can now make planets faster and better and all that, we are not shown any of that. All I saw was the same planet from a year ago with some extra rocks and flowers added which is just ridiculous!
CIG, show me proof that the tools you spent so much time reinventing actually pay off! Drop a minute long cut of all the different planets you already Completed during the last year with these tools! And the number of those planets better be over 20! Than I'm gonna go Wow, spending all this time to make better tools was worth it! No context of what those planets are is needed, no spoilers. Let us speculate, have fun with it!
@Richard Bong
"By now the entire verse could, and probably should be complete, minus some polish."
Yeah well, that polish is important. Stuff like the major landing zones (cities) we got makes the game. I got a feeling that they want to have at least a few major locations like those per system, and those do actually take time to build. Without those, we end up with planets like ED which is barren and devoid of any character and thus utterly disappointing and boring, even though it has like a million landable moons. Still, there should be a lot more progress, which I'm guessing went into SQ42 planets. Which they should just show off as I mentioned above! If this is not the case and there is nothing to show, we are in really big trouble indeed cos a 100 years won't be enough time to get a dozen star systems done, let alone 50 or a 100.
I had some fun with VUE back in the day for VFX, it is a pretty powerful tool for generating backdrops.
What I'm not fine with is the fact that they have been advertising the tools they made trough multiple episodes of their online shows and during CitCon tech panels for years now as a huge time saver!
Yet we see no speed up. Even if they can now make planets faster and better and all that, we are not shown any of that. All I saw was the same planet from a year ago with some extra rocks and flowers added which is just ridiculous!
CIG, show me proof that the tools you spent so much time reinventing actually pay off! Drop a minute long cut of all the different planets you already Completed during the last year with these tools! And the number of those planets better be over 20! Than I'm gonna go Wow, spending all this time to make better tools was worth it! No context of what those planets are is needed, no spoilers. Let us speculate, have fun with it!
@Richard Bong
"By now the entire verse could, and probably should be complete, minus some polish."
Yeah well, that polish is important. Stuff like the major landing zones (cities) we got makes the game. I got a feeling that they want to have at least a few major locations like those per system, and those do actually take time to build. Without those, we end up with planets like ED which is barren and devoid of any character and thus utterly disappointing and boring, even though it has like a million landable moons. Still, there should be a lot more progress, which I'm guessing went into SQ42 planets. Which they should just show off as I mentioned above! If this is not the case and there is nothing to show, we are in really big trouble indeed cos a 100 years won't be enough time to get a dozen star systems done, let alone 50 or a 100.
I had some fun with VUE back in the day for VFX, it is a pretty powerful tool for generating backdrops.
That is part of being done. For me polish is tweeking shaders, adjust things like rivers that don't flow correctly, etc.
Carrara does better trees and plants, and has a more complete tool set. Vue integrates better with other 3D graphics software, but requires you are mixing your render engine with Vue's render engine, or render a backdrop, and foreground with Vue possibly causing shadow and lighting inconsistencies. Vue was used very well in Crystal Skull.
I wouldn't mind if they built their own terrain generation tech, if they actually managed to finish most of the systems by now.
The issue really comes down to the fact they are using an old engine for something it was never designed for. We have to remember that the original engine is a single-player fps engine while it did great graphics for its time that's parent company is struggling. If you look at other games in this engine like Hunt Showdown you can see where the strength of the engine lies. They are smaller maps that are prebaked prior to release and at level, load is loaded in their entirety into memory.
So when SC's scope changed around 2.0 they needed tools to create large-scale maps and then a way to load and unload all of the assets that occupy the world without requiring client PCs to contain more memory than high-end servers. At this point, they really should have ditched the engine that would have worked for an updated wing commander but no longer was capable of handling what SC was becoming.
So yes they could have been done with a game like No man's sky with limited assets and a lot of procedural code to fake the scale but that has never been Chris. My issue has never been with the tools as they need to create the tools to be able to make the assets for the game in a time period where we would not be four generations long in its creation. And while these tools would need to be crafted for any engine they have also had to do a lot of engine modifications because they stuck with an old engine that is really not suited to what they are attempting to accomplish (The good part is because of the issues with Crytek they were able to acquire some of the original developers of the engine which is something most game studios do not have and instead of having to pay for / ask for assistance on engine modifications)
Things like the RASTAR tool for base building, using the planetary tool to make mountains or painting planets and then we cannot forget that in 2019 they introduced the updated Planetary TechV4 which caused them to have to rework all of the Stanton System.
And this comes to the heart of the issue. As shown just updating a tool like the Planetary Tech has lots of knock-on effects and while it reduced the planetary time sync from months to weeks it still caused them to have to rework the whole Stanton System and the same goes for the current implemented ship catalog every time they add a new feature. To think we have been in the Stanton system since June 2017 and while I fully acknowledge the impact covid had on the last two years there we are still looking at a development of 5 years held up by fundamental engine requirements like server meshing while things like gen12 render so that the graphics engine can still keep up with triple A tile looks there is a lot of development time that is not accounted for.
The only reason I even launch the.... ummmm..... program? now is so I can show my kids the interiors of my ships. I don't really want to call it a game at this point.
Any modeler can build bases, in fact IIRC they use 3DS Max for that. Carrara has the terrain and plant tools (Yes Carrara also has a passable modeler for things like bases, but Max does that better.), and Astrosynthesis plus Fractal Terrains builds the systems. Making the engine support the assets is one thing, building things to create assets, that already exist, is another.
And yes, Cryengine is definitely the wrong base engine.
it all sounds like an uphill battle, would any engine they didn't make themselves be capable in all the areas they need it to be, like 64bit presision, no loading screens, aparrently infinate maps, multi-layer cloathing and cloth physics etc etc etc?
Would having to rebuild the engine as they go explain having to remake the wheel for things like the example in this thread which one would assume would be able to be taken off-the-peg in other cercumstances...?
Would having to rebuild the engine as they go explain having to remake the wheel for things like the example in this thread which one would assume would be able to be taken off-the-peg in other cercumstances...?
The modeling, which
Building terrain, plants, props, etc. isn't supposed to be done in engine. It is done in software, like Carrara, 3DS Max, Maya, Lightwave, Blender, etc. You may do your UV mapping in the modeling software or a different piece of software. Modeling generally includes the moving parts, and LOD's. Things like terrain usually require a specilaized tool, like Carrara, or a plugin for 3DS Max, Maya, etc. Different types of objects are generally, in a bigger studio, are handled by different people. Building a starship requires different knowledge than building organic shapes, like people.
Texturing isn't done in engine, it is done in a 2D or 3D graphics program like Photoshop or Substance Designer and is, generally, a different discipline.
Rigging, which defines how an organic object, like a person, moves, is usually a different discipline, and specialized work, and is usually, but not always, handled in specialized software.
Animation is another discipline. Again, not usually handled in engine. Some modelers do it well, but have limited export ability. A specialized tool, like Motion Builder,, is generally used.
Shaders, which takes your various texture maps, and applies them to the object, is done in engine because it is render engine dependent.
Texture and poly budget is engine dependent.
The list goes on.
The game engine determines the poly budget, the texture budget, the render engine (and subsequently the shaders), the netcode, the physics, the data base(s) and that kind of thing. With Cryengine, there is less, current, community support than with Unreal or Unity.
For example you can go to the Unity store and pick up assets, like trees and people or plugins that do things like Netcode or improved physics.
Yes, you still have to do a bit of coding for a game engine to make it do what you want it to do. Terrain, isn't part of that.
I'll break down some of the more common tasks required.
Building terrain, plants, props, etc. isn't supposed to be done in engine. It is done in software, like Carrara, 3DS Max, Maya, Lightwave, Blender, etc. You may do your UV mapping in the modeling software or a different piece of software. Modeling generally includes the moving parts, and LOD's. Things like terrain usually require a specilaized tool, like Carrara, or a plugin for 3DS Max, Maya, etc. Different types of objects are generally, in a bigger studio, are handled by different people. Building a starship requires different knowledge than building organic shapes, like people.
Texturing isn't done in engine, it is done in a 2D or 3D graphics program like Photoshop or Substance Designer and is, generally, a different discipline.
Rigging, which defines how an organic object, like a person, moves, is usually a different discipline, and specialized work, and is usually, but not always, handled in specialized software.
Animation is another discipline. Again, not usually handled in engine. Some modelers do it well, but have limited export ability. A specialized tool, like Motion Builder,, is generally used.
Shaders, which takes your various texture maps, and applies them to the object, is done in engine because it is render engine dependent.
Texture and poly budget is engine dependent.
The list goes on.
The game engine determines the poly budget, the texture budget, the render engine (and subsequently the shaders), the netcode, the physics, the data base(s) and that kind of thing. With Cryengine, there is less, current, community support than with Unreal or Unity.
For example you can go to the Unity store and pick up assets, like trees and people or plugins that do things like Netcode or improved physics.
Yes, you still have to do a bit of coding for a game engine to make it do what you want it to do. Terrain, isn't part of that.
Cheers, appreciate your kind efforts to explain :-)
So at what point will CIG declare that its 'engine is complete' for Star Citizen as conceived today?
Radegast74
Space Marshal
Here you go...I finally found the best summary of this situation:
View: https://www.youtube.com/watch?v=AV69DyOuzbE
honestly, no engine can do what CIG wants to do for the size of the game CIG wants to make. Having hundreds or thousands of people in a game that uses a twitch shooter tech and all the calculations involved therein is very possibly too big an ask for any engine, much less a less than modern one. (and Lumberyard or even Star Engine is not a modern engine, for all it's changes)
Simply put, if you want 1000s of players in a single instance, you can't have twitch gameplay. Especially when spread all over the world. I don't think we are going to have the "single world" CIG has talked about. I think the best we will be able to do is regional world servers. So EU, AUS, Korea/Japan/Vietnam/China (mainland China may need to be it's own server), North America, South America and other markets where appropriate.
Simply put, physics will prevent their vision. If this was EVE and you had a time based system for attacks and such, lag would not matter as much. For twitch based fights, it does matter.
Personally, I think we will get a more and more complex game, but 1) it will never be finished, and 2) it will never be released in a bug free state...
Simply put, if you want 1000s of players in a single instance, you can't have twitch gameplay. Especially when spread all over the world. I don't think we are going to have the "single world" CIG has talked about. I think the best we will be able to do is regional world servers. So EU, AUS, Korea/Japan/Vietnam/China (mainland China may need to be it's own server), North America, South America and other markets where appropriate.
Simply put, physics will prevent their vision. If this was EVE and you had a time based system for attacks and such, lag would not matter as much. For twitch based fights, it does matter.
Personally, I think we will get a more and more complex game, but 1) it will never be finished, and 2) it will never be released in a bug free state...