So, I:B is a sci-fi game. The reason gamers are drawn to sci-fi games is not merely for the gameplay. That is a very major part to any game, yes but (and a big one) for Sci-fi, it’s also just as important to immerse the player in the universe they are playing in.
Games that play well and look pretty are great for the quick jump-in-jump-out play style, but, true Sci-fi games have needed more and more levels of immersion over the years to accommodate the desires of sci-fi gamers to “lose themselves” in the games they’re playing. Use the game Wing Commander as an example. The immersion levels were unprecedented at the time but only employed very small amounts of immersion than what we’re now seeing in space sims.
I-Novae has already created what would have seemed impossible only 5 years ago. It’s unprecedented to have an entire planet procedurally generated, let alone entire galaxies of such planets. So, finally to the main point of my suggestion.
Immersion
The precedent has been set by quite a few new projects out there including I-Novae’s. So, my suggestion (as others have also suggested) is to add higher levels of immersion. Clouds. Weather. Trees. Cities. Promise these things to come along with I:B and the game mechanics that will also be deployed with the game and I guarantee that the gamers will give answer with money - in abundance!
On the other hand, without those and other levels of immersion (or at least the more common like trees and clouds), the first backers will start playing, release videos, and pass the word but others who hadn’t pledged originally might find themselves saying “that’s awesome and everything but why are there no trees???”
The potential here is absolutely mind boggling so I hope my tiny suggestion at least gets a spark going in I-Novae’s design considerations. I leave you with this thought: Would Minecraft have done as well as it did without trees and clouds? Forgive the use of a game that doesn’t fall anywhere near the depth and scope of your engine but it is meant for nothing more than a “what if” scenario.
I don’t think Tree’s and Clouds are going to be much of an issue, and isn’t really a top priority as such at the minute. As you mentioned, this is a space simulator, so I think the best way to start of an Alpha to a Space Sim is get the Planets, Stars and most importantly Ships into the factor.
I think by time the game hits Beta/Release then yes, Planet detail will be essential, with Trees, Planets, Animals and the rest of nature, this would be pretty simple, however are wanting things like wind and such to affect the trees, and each tree is effected differently?
I think procedural generation of a planet is a lot easier, than generating nature based on the planets environment, I mean are we then going to base the planets on how close they are to the star? And then even the tree’s would also have to be affected by which type of Star, Red Dwarfs would require planets to have purple chlorophyll, i.e the planets environment would be more purple.
I definitely think detail will be implemented, but it won’t be this early, and won’t probably be till after a decent amount of funds are here as it will require a lot of work.
I can understand not having those for alpha but I had read a thread where one of the devs answered a question about clouds/trees and pretty much said they aren’t considering that for game release at this time but might later. That comment also tells me that they might consider it more if they see an overwhelming amount of pledgers through Kickstater.
I just wanted to make sure they were aware of the potential down-stream effects of not doing so.
When you’re running on a tight budget, you have to select your features very carefully. If we were to add all those immersion features, then we would have to cut something else to keep everything in check. Would you rather have ground level details on planets or more game content such as stations / spaceships ? Now if the KS is more successful and gets to the higher tiers, that will hopefully not be a choice we’ll have to make…
Thanks @INovaeFlavien! I do understand budgets. I sure hope Kickstarter gives you what you need to go that distance though. It’s not going to be game breaking but if it were to be compared to a painting, we’d want to get those “happy little trees” in there.
What if the game engine could be updated to automatically generate cloud layers in the atmosphere but only volumetric once the player is within a certain radius of them for the sake of view distance. On the ground and in space, the player would just see the “2d animated” layer(s) to give the look of realism which you wouldn’t need volumetric at those distances.
I’m sure my tiny little brain is thinking all wrong but it seems it should be simple enough. I’ll concede to your statements though as I ultimately do understand budgets as I said and I know choices must be made at first. Here’s hoping for a successful Kickstarter! I’ll spread the word to my friends at the Star Citizen forums as well.
You’re talking about completely updating the engine to fit this in. It wouldn’t require much change to the mass amount of code, but it requires a lot of work as not only do you have to have the weather/clouds/nature to attach to a planet, you have to generate it off of the stats on the planet, therefore you actually have to implement stats for the planet.
The following stats would need to be added for each individual planet;
Planet size; you don’t want tree’s or clouds appearing outside in
space
Planet Temperature; how close is it to the sun
Gravity, Gravity affects all types of nature generation
I’m stating this because if iNovae would even think about doing this they wouldn’t do a half hearted job and throw the things in there. iNovae care about the experience by the player, and a player would have a better experience playing in a realistic universe with no trees, than a realistic universe with crappy trees.
Budget only slightly comes into this, it’s a massive update to the mechanics of the entire game, and although I stated about graphics and performance earlier, it will be affected. It also increases the games size.
There are a lot of things to take into account with this suggestion, and we all agree it would be stunning, but it’s also more difficult than it sounds. Well I say it’s difficult, the iNovae team seem to do a good job at making the impossible, possible and I’m no expert but it doesn’t seem easy to me.
Indeed. It would be far better to wait with such things as vegetation on planets untill all the more important aspects of the game are in place and working well. Once that is done, then they can start playing with all these sorts of details, which further down the line could be part of the MMO game (still hope Quest for Earth is the “end goal” here)
I agree with both @MrGibbs and @SPECTRE for the most part. My last post actually just focused on clouds and how easy that would be to. I watched one the of the tech demo videos for I:B - one where they fly through the rings of a planet and then into the planet’s atmosphere. How they have worked out the way you see the rings from a distance but then start seeing the actual asteroid debris that makes up the rings only when you get a close enough distance.
I understand the trees and other planetary aspects not being something that is possible with the time/resources they have right now. I would even understand the clouds if the devs see that as not possible using something similar to the planet rings method.
You forgot that realizing the clouds up close isn’t the hardest part.
Problem is that you have to simulate some kind of aproximation to cloud arrangements on a planetary level.
Not necessarily on a planetary level. Remember their engine is already built for super large scale and procedural generation. From a programming perspective, you shouldn’t have to literally program a lot more or even graphically design anything that covers entire planets or accommodates the different sizes. The great thing about procedural is that as long as the engine reports an objects size specs, other things around it can and will compensate automatically if programmed accordingly.
An atmosphere is already generated for each planet. Mathematically, you can come up with a standard algorithm that would determine the troposphere in accordance with the outer edge radius of the atmosphere and the average surface radius of the planet. Using the earth’s troposphere radius between its outermost atmosphere level and its surface as a template for that algorithm should work just fine even. Then in your programming, simply tell the game engine to create a hollow sphere with 0 wall thickness and place a 2d animated alpha channel clouds texture. This is what the player would see from a distance (on the surface of the planet and upon distant approach to the planet). As they get closer, just like with the rings of the planet in the video I mentioned, volumetric clouds would replace the 2d textured hollow sphere - at the same radius.
I’m actually not a programmer but I do understand the general logic that programming is based on. Whether this applies to their game engine and how it could fit in their code, that I don’t know unfortunately but they do and they can decide of course whether this suggestion is feasible or not based on how I described it or any other way.
It’s fun just talking about possibilities at the very least.
Even if you could make a single 2D texture map (created how? by hand?) the size of an entire planet with a high enough resolution to be adequately detailed so that the some generator could pull 3D volumetric clouds from it (seriously, making a system to build anything in 3D clouds off a 2D texture map isn’t easy). You’re still going to have planets transition from a 2D cloud texture to being fully volumetric at a single point? have you any idea how jarring/ugly that would look?
You think not having clouds would be bad for immersion? With this you’d end up having clouds that spring into existence like pages in a pop-up book.
Personally, I don’t feel that clouds or trees have any impact on immersion what so ever. The simple reason being, that the ability to be immersed in any world relies solely on how you’re able to interact with it, and it’s apparent effect on you.
Sure, they would give a better sense of scale, but wouldn’t effect the feel of the game itself. For true immersion in a space game, you need to worry about how your view/camera reacts to the movement of your ship as you pull off manoeuvres; how physics are portrayed and how your controls respond; how your ship shakes and rattles as you dive into a planets atmosphere or how it shudders as it’s armour shrugs off incoming fire. That, along with appropriate and accurate sounding audio, are the sort of things that will let you immerse yourself in the experience.
@Naiba is correct, creating even a basic planet wide weather system starts getting very tricky very fast and this is primarily due to 3 things:
The sheer size of the volume of the atmosphere surrounding a planet
The expense of doing the computational fluid dynamics on this massive volume
Sub-surface scattering, which is how you render sexy photorealistic clouds, is a very expensive operation in and of itself. Even if we can figure out the solution to #1 and #2 we’re still faced with the problem of finding an efficient way to render an animated volume using one of the most expensive rendering techniques in computer graphics
The memory and computational requirements start getting out of hand pretty quickly. Also, while it’s easy to procedurally generate a static set of clouds those techniques start to break down when you want to do an accurate simulation of animated clouds because of how fluids interact with each other.
In the near-term the most likely cloud solution is static, procedurally generated clouds. Just getting those to be volumetric at the scale of an entire planet is pretty challenging. It’s going to take a lot of R&D effort to create really great looking clouds that don’t make your computer cry tears of blood and that’s why we just don’t have the time for it at the moment.
Isn’t it possible to simulate and pre-compute the 3D weather maps on the servers and simply send the baked maps to clients so players don’t render anything? Render those maps in 1-hour chunks and have players download those chunks as they need.
Perhaps make these kinds of computations shared/distributed across multiple clients like the BOINC or torrent systems on a volunteer basis, especially for those with high-end GPUs.
Also i’ve noticed the scale of planets and rings/asteroids is not quite right, i believe it’s a factor of 10 at least. Or maybe the demo’ed planets are actually really small so dunno…
Meh don’t mind me, i’m just rambling because i want to see clouds and trees
It’s completely possible to add weather and nature to planets. It’s not the question of is it possible that’s the issue, it’s the fact that it’ll require a lot of time and editing of code to implement for something that isn’t even ready to be rolled out yet.
It’s an additional feature that would like to be added after the core game is completed basically but first the core items need to be completed
Certainly we can create distributed systems to calculate weather and clouds but, as @MrGibbs mentioned, it’s a matter of time and money. As for the scale they are correct. Keep in mind that distance has a huge effect on your perception of scale. If a super large object is very var away it will look exactly the same as a very small object that is much closer. We’ve reduced the distances between the planets so that they’re larger in the sky however their sizes are absolutely to scale. You’ll see once we release our demo =).
I’ve been waiting for almost 20 years for someone to take up the mantle and tradition of those games. Elite 4 was being worked on from the year 2000, but development broke down several times. Then, I heard of Infinity / I-Novae in 2007… and now Elite: Dangerous (Elite 4!)… so my heart filled with hope again. And now we have so many procedural space sims coming, that I am overjoyed!!
I-Novae is obviously more impressive than Elite 2 and 3 because of the far higher level of detail, which is due to the very hard work of Flavien Brebion, but also as a natural evolution in the technology since 1993. A lot of things have changed from 1993 to 2013.
Personally, the most advanced animated clouds I really care about seeing is a very localized fluid simulation of the clouds reacting to your ship passing through them.
animating a noise function enough that the clouds aren’t always exactly the same is pretty trivial. Do it slow enough, and you won’t have to worry about how it looks animated, because you’ll only ever see it still, or moving slow enough to not notice any movement.
and I’m no expert but it doesn’t seem easy to me.
Have you not heard of these old games?