Star Citizen Live: Invictus Launch Week Vehicle Q&A

It is still being worked on. That reworked shield tech is mostly to do with SDFs, the idea that previously we would have the ship encased in a mesh that someone would make, but now that is calculated automatically with this SDF. An SDF, if you haven't heard of it before, is this data structure that lets you query how far away a point is from the ship, so you can effectively work out very easily if a bullet or projectile has hit the ship. It'll result in much more reliable shield hits and stuff like that.

Of course, SDF is being used for quite a few things, not just the shields. On ISC, SDF is being used for the new reentry effects. I'm trying to think of other things SDF is being worked on right now, and I'm blanking.

VFX is a huge one. There's so much VFX that you really want to conform to the ship, and it's time-consuming to author all these effects for every different ship, so this automates a lot of it.

Especially when you have to deal with animating ship parts and bits that fall off the ship, you get that. Not for free, but it's significantly easier than how the current systems work.

Indirectly, yes. So something we're planning out at the moment is a lot of the precursor work. So anyone that was around before Item 2.0 will have been familiar with avionics motherboard items. So we're doing the planning for the revamp of those to go in the ships. And obviously a big part of that is all the blades you put in. So we obviously need to get the base computer item in first and then we can work on the blades. But that's all being planned out in the very near future.

Right, and just for folks who maybe have joined the project since the last time we mentioned AI blades, they are components that go into the vehicles that will allow you to automate the actions of the manned turrets.

Yeah, so the vehicles will all have computers, small ones, medium ones. They each have a varying amount of blade slots in there, and you can put all these different sorts of blades in. The most popular one will be: I want to slave a manned or remote turret to the pilot seat. That would take up one of the blade slots, or more perhaps.

So yeah, the plan was the Carrack team would move on to the Merchantman, but as we were coming towards the end of the Carrack, we had the team do a bit of pre-scouting of the Merchantman, and based on the work that we did on the Defender and all the problems we hit with that, which we discussed around this release, we wanted to spend a lot more time concepting it to make it a really solid ship. Because I know the Merchantman [backers] are known as super vocal about things. Looking at the concepts, you can see there's quite a lot of inconsistencies with some of them. I want to rule all of them out, have it all planned on paper, so to speak, before we start modeling it out. Paul Jones and those guys are going to be spending time tightening up the concepts, and then the production team will hit it at some point in the future. The Carrack team has essentially been split into two smaller ones that are working on two other ships.

We learned a lot of lessons with the Banu Defender about how to do the organic style design for the Banu as far as the ship. And I imagine there are a lot of lessons we had to take from that and apply to the Merchantman, which is a ship at a much larger scale.

Yeah, it's like we have with other ships. You can do a ship at a small scale in a manufacturer and we have an idea of how it's going to work at a larger scale, but it's not a one-to-one translation.

It's something that we're definitely trying to move more towards. In the past, it was very much, we've done the ship, so let's just release it. Starting this year, the end of last year, where we can, we try and release some of the gameplay with it. So the Cutlass Red was a good example in that we had a very basic form of medical gameplay in. The Mercury is obviously twofold. It's a cargo runner and a data runner. So it's not as critical to ship with data running because it does have another purpose alongside it. We'll see as we get closer to the time. Because data running is essentially cargo running with a data commodity.

I think it's important, and we've touched on this topic before a couple of times in the past. Obviously, when things can align the way we hope they do, yes, we would always like to release a ship with its associated gameplay. But as is often the case, the associated gameplay is done by teams that are outside the ship team, for instance, and different schedules and different workloads and different priorities. Sometimes those two don't always align and the ship is ready to go before the associated gameplay. I think one of the bigger examples is the Reclaimer.

Yes. So it's when you have that asset just sat there, ready to go, beyond any gameplay. It's a very hard decision to go, well, shall we just keep this to ourselves for another 3, 6, 9, 12 months? Or do we give it?

And I imagine that plays some factor into your scheduling, too. When you're trying to schedule what ship is going to be built next, you have to look at, well, is there a chance that the associated gameplay can come online with it? And if there isn't, maybe we don't prioritize making that ship sometimes.

Yeah, and the vehicle content schedule is currently mapped out 12 months from now, and I literally sent an email just before this call about what things we are going to put in the next slots after that.

Yeah, there's a lot of technical obstacles to that. Some of them we're actually making progress on, like you need docking and stuff for the Hull C to be useful. And John can speak to all of the other stuff. There's just so many technical hurdles to get that ship out.

Yeah, so docking is obviously a big one, which we're handling on the vehicle side. Then the other things really are having multiple physics grids in a ship and having vis areas and area shapes scale dynamically at runtime. Area shapes are basically just a markup that we can use to define a room's volume, or how large a room is, so you can have the audio tied to it, the air tied to it, and vis areas control the visibility in those areas. The engine assumes they are always static and they never change. All you can do is you can turn on and off a portal to look in or move in and out of them. Obviously, the Hull C has this big section that expands in the middle, so these rooms have to dynamically scale at runtime, which is something the engine was never designed to do, and it needs adjusting to do that. It's the same for physics grids. At the moment, we have one physics grid per ship. Having multiple ones for the front and the back is a relatively easy change, but having that third one that then dynamically scales at runtime is the other big issue. So docking and those core engine problems are the big first hurdles, and then we move into performance and optimization. At the moment, SCU is a one-to-one match, so the Hull C is like 4,000-odd SCU of cargo. At the moment, that would be 4,000 boxes stuck to the outside of the ship, which is not going to be great for performance. So we need to have the cargo system, which is another team outside the vehicle team completely, work on using the different sized boxes. So if I buy 32 SCU of stuff, don't fill my ship with 32 individual boxes. Fill it with one 32 SCU box. And then that just scales for the Hull C. So instead of having 4,000 boxes on, you might only have 100 boxes. And that pretty much touches on them all. So a lot of those, they've been in limbo for a while. I'd say the vast majority of those are either in progress now or going to be imminently in progress.

I think the discussion with that is that it should be almost prohibitively expensive, such that you could do it by yourself, but it's just so much money that it'll cripple you. If it's spread over a group or an organization, it'll be much more feasible and manageable.

It's technically possible, but not realistically possible. Yeah, you can fly one of those ships by yourself. You could hire an NPC crew instead of other human players. That just adds more cost. If you fill it with humans, they're going to want some share of whatever you're doing. It's realistically not something that you use as your daily driver around the 'verse. Although someone will. There's always someone that will have their Javelin out every day.

Let's not kid ourselves. There will be some magnate who corners the market on Quantanium or something and is a Star Citizen billionaire. And they will, just as a matter of status and a matter of effort and a matter of 'I'm going to do it the hardest possible way', they will be solo operating a Javelin out there. And it will happen.

It's all still planned, those customization elements. People obviously have components in-game that they can customize now. There's going to be more of them. There'll be more variety of them. AI blades we talked about a few minutes ago. Tuning of thrusters is a good one. I know there's been some very early milestone or Kickstarter rewards with tuning kits that we'll deal with. Something that we do want to do is have different thruster sets for ships, which I don't know if Dave and Rich want to touch on.

We've done a huge amount of the prerequisite work to support that, and we're pretty much there in terms of the core vehicle system actually being able to do that. Now it comes to a matter of actually doing it and fitting it in with the economy and having the shops all supported and everything, which is a lot more work.

Yeah, there's definitely plans in the very near term to have at least one ship support having two different thruster sets that you can purchase in-game.

Not to mention as well that obviously having multiple thruster sets almost doubles the tuning work on the flight balancing of that ship, because now you have to make sure that ship works in a reasonable capacity for two different configurations.

So they have [been paying attention]. And the exciting news is that I think everything that you just said, Jared, is pretty much... We've been reviewing it and making sure it's at a reasonable standard that we are happy releasing it. And all of it, we're pretty much happy to go. So there's the aerodynamics. We've overhauled the aerodynamics. And the efficiency curves obviously make a huge difference. If you haven't heard of them before, they apply efficiency curves to thrusters, so things like atmosphere, particularly atmosphere, will affect the output of the thruster. There's various other subtleties. There's loads of stuff happening in that. I don't want to go in too deep because we're going to actually have a more in-depth segment on ISC, I believe, coming up in the future. But suffice to say, it's almost here. It's really close.

It's been really fun to tune it as well. I'm really excited to get it in the players' hands and just see what they can do with it.

Yeah, so we're basically getting there, we tune it, we'll push it to its limits and bring it back and get ships working, and then we'll provide feedback to David, then we'll sit together and work on some solutions for the problems we've found with the feature maybe, and then we'll just carry on to improve it. And then it's kind of our job to really finish off the fine-tuning of all the ships and provide the unique character that we're going for in these ships, because we've got small ships and then we've got massive ships and they've all got a behavior in atmosphere. So it's about providing that kind of unique performance that these will have. And so I am excited to get that into players' hands and see if they enjoy it and see if they enjoy the differences as well.

The big advantage we have with the vehicle experience team is that Rich and his guys have this really holistic vision over the design and feel and gameplay of vehicles. Whereas vehicle feature programmers and designers tend to be narrowly focused on doing their thing. What'll happen is we'll work on aerodynamics or whatever, but then Rich and his guys will have this perspective of how that fits in with this other thing and this other thing. It's their responsibility to make sure that the holistic experience is tip top.

Before the event happened, the two teams were literally sat next to each other in the office, so it's very collaborative. It's not like one team does something, sends it off to another floor in the building, they play with it, send it back. Everyone's all just sat together collaborating on it. Normally.

Normally. We're basically just on Teams at the moment having lots of calls all the time, which is very different. But it's working pretty well for us.

A lot of it is just how the pipelines got better over time. So if you look at a lot of the older ships, a lot of them didn't really consider component placement or metrics. The Redeemer is a great example of a ship that didn't adhere to metrics particularly well. So when we get into production, things have to change to fit. And for the Mercury in particular, the concept did a great job of fitting everything in, but it's been a year plus since we did that concept, and things in the engine update over time. So when that concept was done, we were only just dealing with landing gear compression. So the landing gear that were concepted worked fine for how the game worked then. But now we've learned a lot of things. We need to beef up the landing gear, change their position ever so slightly so the ship settles better. Because you can make a nice asset in Max or whatever, but until it's a physical thing in engine and you see how it lands and settles on its landing gear, you're never 100% sure how it's going to work. So we had to beef up the landing gear, slightly change the placement, and all those things knocked on to: if you have bigger, stronger landing gear that perhaps have a greater length, so you have better compression, so it lands easier, you need a bigger volume to put these things into the ship. So that then has knock-ons to the hull, because you need a bigger thing. You need bigger doors to cover the landing gear. They need somewhere to go. One little thing like 'I want to have half a meter more landing gear clearance' can have huge knock-ons to the underside of a ship.

Because we don't cheat our ships. They are one-to-one. The interior fits in the exterior, unlike the Millennium Falcon. I also have to imagine, and I don't say this with any information, this is just a question, the air ducts, the compartments, getting around, one of the features of the Mercury, that probably necessitated some further development, just getting the player model to crawl through there and stuff.

I would think so, yeah.

It's probably one of my favorite ships, the Vanguard, and I think it fits pretty well with the weapons it's got. It's got a good balance of weapons and missiles.

I disagree with saying it's got little firepower.

Yeah, because it's got quite a good kill ratio in the PU really.

Yeah, looking at the analytics in AC and PU, it is top three, top five at worst, patch to patch on the amount of kills it gets. I can see it's not ideal because it has the mix of fixed and gimballed weapons. So if you don't like that mix, your only choice is to go all fixed, which gives you more DPS, which only improves the situation. And when we did the rework, we had to make that decision at that point of, do we stick to the concept of those four fixed nose guns? And we talked with Chris a lot, and it was like, yeah, that's an iconic part of the ship. We decided to keep them, which ties us into various problems. I know there's been lots of solutions, like get rid of those nose guns, make them gimbal. If they gimbal, there's no way for them to gimbal. They're all going to clash into each other. Get rid of the chin gun, stick it under the wings. If you put it under the wings, there's no space for landing gear, no space for missiles. Where it is, we're pretty happy with. Maybe the bespoke guns could have some tuning to them, but it's not going to get another rework.

So it's probably down to how we see it working in the future when we've got improved turrets, because it's got a turret on the Vanguard. And we've always got some improvements coming to turrets that are going to make them much more effective. So that's kind of where we see it going. So I think it's perfectly reasonable for them to feel it's probably not as powerful, but in the future, when we have a much improved turret experience and they're much more balanced towards actually working and being able to shoot rivals, I think that's kind of where we see it going.

And there's always the wild card of when a new ship comes into the game and upsets the meta. The Prowler is overperforming compared to what it should be doing, which knocks into the Vanguard owners' territory. So is it that the Vanguard sucks or is it the Prowler is too good? And I would say that the Prowler is too good.

I think a lot of people would say that fixed weapons in the current meta, whatever the current meta is, are somewhat limited.

Yeah, that's actually another good point and something that we're fixing for 3.10. We know fixed weapons have been pretty rubbish, certainly compared to gimbal assist and gimbals. It's something we've worked on and it seems to be a viable choice now between the two.

Well, I can't promise we're going to solve the world with this, but we feel for 3.10 it's a really good step. Best I've ever played. So we've had some really positive playtests within the group where we've got in a Hammerhead, for example, and we've felt like kings, basically, so that wasn't possible before.

We had one person flying a Hammerhead and two people in the turrets. So one was in the top turret and one was in the side turret. And we were just constantly spawning waves of AI. And within about five minutes, we'd killed 50 people just with turrets, which is pretty unheard of with how turrets have been. And there's a whole slew of turret improvements. It's not just one thing we've done. We've pretty much reworked them from the ground up. And I believe, if we're not doing an ISC episode on it soon, there is going to be a post detailing a lot of the combat and flight changes for 3.10 that will come before 3.10 that will explain it all.

It's time to get those conspiracy theories about the Constellation Taurus out of the way. There's nothing really in particular that's stopped it beyond other ships have taken priority. At the end of the day, it's another cargo ship. We have cargo ships in the game already. It's more work than it seems to do it because it has exterior hull geometry changes and significant interior changes. Unlike the Phoenix, where the changes were constrained to one central section of the ship, it has that second secret cargo area in the front and the extended middle section. So there's a lot of work there. And when it comes to do we finish the Taurus or do we get the Prowler in game or do we get something else in game, the poor Taurus always seems to lose out. But it is back in the schedule. When we're confident of the release date, it will go on the roadmap and all the Connie variants will be closed out. There's no secret entire rework of the Connies on the cards.

So docking requires ports on the ships to match up. We're planning on two different, effectively, metrics or shape sizes of docking ports that are not compatible with each other. And so if you want to dock one port to a matching port, then they will work. If you want to dock that port to another size, then it won't. And I think most of the ships already have their docking ports there. I think, for example, the Constellation has the smaller docking port size on the sides, the round one, and then there's a larger docking port size. And then the stations will have docking ports of varying sizes, so that ships, particularly the larger ships that maybe are too big for the landing pads, will just dock with the station, particularly rest stops and things like that. But otherwise, yeah, docking is progressing. We're working on it at the moment and trying to resolve a lot of the issues that come up with it. There's some, shall we say, special cases where, for example, the Merlin docking with the Constellation, that's a special case because it doesn't use either of those two metrics. Another special case is the detachable command pod on the Caterpillar. So things like that. I think the Hull C is also an exception. So we need to work through how those are going to work and how we integrate them with the rest of the system.

For literally years, I wanted to do the space A-10. And it was like, no, nobody wants that. I was like, I want that. So let's do it. And finally, we managed to get the Ares out. So I haven't got a new one that I want to do, because that took a long time to get through.

Although it's heartening to know that even you have to fight to get a ship into the game.

Top of my mind is older ships. I like the idea that our universe has this whole history. I would love to see some historical ships in the lore of the universe, kind of showing you how the development of technology has progressed. That is a very high-flying ambition, though, but it just appeals to my sense of lore in the game's universe.

Like a '57 Chevy or something, like a classic version of an existing ship. I'm with you. What about you, Richard?

Yeah, so I'd probably go for a mix of an M50 and the Herald. So something kind of bigger, beefier, with massive rockets on, but that can also really perform as well.

Would that be a racing truck?

Yeah, like a racing truck.

I gave up on my Origin stealth ship, the 404, a spaceship not found. But no, I still want the jalopy. It's actually kind of like what Dave wants. I want an old, beat-up ship that you can't pledge for. You can only get it in the game. It is an absolute junker, and it's the kind of thing that you'd buy to get from A to B, like when you're stranded on one planet and all your ships are somewhere else. But then there could be a whole quest line where you can restore it, all completely in-game. And then it would be a point of pride at the end. You'd have that old junker completely restored, and people would know that you put 100 hours of work into restoring it. Either that, or the one racing type ship that we don't have yet is our demolition derby. I want an Anvil racer that's just got junk plates attached to it. Your M50 can do the maneuvering and your 350R can beat you on the straightaway, and then a demolition derby racer will go through them.

I'm thinking of the trucks from Mad Max, a spaceship version of that.

It is definitely a lot faster. I know back when we did the flight presentation at CitizenCon 2948, I think it was, we showed off all the new flight tools, and we continue to improve them. There's new tools for the aerodynamics and all sorts of stuff. And I know there's various other new tools for the other aspects of ships, but I think John mentioned the other day that our ships have also gotten more complicated. So there's more to do, but the tools are definitely years ahead.

It's definitely quicker. Using the tuning example, when I first started, it was pretty much you open the XML and you look at these sets of numbers and you just type some of the numbers in there and play it and go, no, that doesn't feel quite right.

You would have to stop the game, though.

Yeah, stop the game, type it in the XML document, restart the editor, go. And then over time, it got live update and then coming out of Excel. And then now we have all the cool tools that Dave's written and Rich uses day in, day out that do it live and give you all sorts of information. And then on the art side as well, obviously the old way we used to do damage, you would essentially model the ship four times. And as each one went down to 75, 50, 25, 0%, you would swap that part out. Now we have the UV2 shaders. We just model the ship pristine. We model some gubbins inside and let shaders do all the magic in between.

Yeah. We're looking into ways of doing this. Particularly in atmosphere, it's a lot easier. But part of the issue is that our terrains are quite rough. And so we're carefully caught in this balance of: we want these low-riding, exciting ground speeds, but the planets are filled with all sorts of stuff. And so we have to kind of go, OK, well, we could increase the height that they hover and then you can go faster. But it's this weird balance where we're trying to push them as high as we can without it being annoying that you're just constantly hitting stuff. It used to be a lot higher when we first did it with IFCS 2. And then it was just really annoying, so we had to lower it. But we're trying. We want it to be faster, and we're trying to make it faster.

Have you ever considered just increasing the speed and then going to the environment team and being like, that's the slowest we can make them, guys, you need to clear some of these rocks out?

I've tried it.

Yeah, we have tried that. It's one of those weird things as well. You get it in the real world. If you're driving in a low-down car versus an SUV... We could make you be higher, but you have to go so much faster to get the same sense of speed as just skimming along the ground. We have the same problem in-game. So where we are is sort of comfortable-ish.

We could probably go a little bit faster and just let you have a little bit more responsibility about not hitting stuff. And then we'll see all the threads about why we hit stuff.

The range is not purely based on the fuel tank size, though. Obviously, the usage rate of the quantum fuel is different for each ship. So that's a primary factor, but there's various other reasons why it's not as simple as just the fuel tank size, right?

Yeah, you've got the drive that factors into it as well. And another thing which gets confusing is the ship matrix, and it says this ship has one small quantum fuel tank and this ship has one small quantum fuel tank. And eventually we want to have those tanks be physicalized things on the ship you can target, which means they need to not just be a metric box like all of our other components, because of fitting them in everywhere. Like in a real spaceship, you have it all in the wings, you have it under the floor, wherever you can find space for it. So long term, they'll be bespoke to the ship and they'll be almost modeled. So you can see them wrapped around, and it'll be a combination of the drive you have in and the size of the tank, which will be bespoke to the ships. The current system of one size one versus one size one is just a very, very loose approximation of that for the short term.

I think in a simpler sense, though, they're saying that the Terrapin is a ship for exploration and it should have a longer quantum range than the Pisces.

If there's stuff like that, then stick it on the Issue Council and we'll look at it.

So I do have an answer for this. Having moving objects at rest on moving platforms is a really hard problem. And obviously physics grids solved this problem for inside ships, but elevators and ramps and stuff, they don't use physics grids. And it's partly because the way the physics grid system and vis areas work means that having exterior physics grids has some weird issues. Like the vis areas will cause shadowing artifacts and other issues that we haven't been able to resolve yet. There's also the issue that John mentioned earlier that we can't have multiple physics grids on one ship. The alternative is then to not use physics grids, but then we have these elevators which teleport their position because they're animations, like the ramp is an animation and stuff. When you have teleporting objects trying to interact with a physics collision system, the system gets very confused, because the objects move in ways that are not predictable to the system. If it was just a velocity, it would be predictable, right? And so if you imagine a really complicated suspension system sitting on a teleporting platform, the collision system gets quite confused. So we either need to break out those parts and simulate them manually with velocity and everything, or we need physics grids to work in a more complicated way. It's a really, really hard problem to get that right. We're still trying to work through the issues that we have with getting that fixed.

One of the big pieces that is missing for that is the ability to have items, such as the missile racks of the Retaliator, inside the object containers of the vehicle, which are essentially the rooms inside the vehicle, attached to a vehicle and all communicating together. The inside of the vehicle currently doesn't talk to the outside. I think we're going to talk about this more in a future episode of Calling All Devs. At the moment, the missile racks on the Retaliator are attached to the outside of the ship. So if you took away that interior room and put the dropship module in, you'd just have these two torpedo racks still stuck on there. So we need to make those items be part of the object containers and swap in and out. And the work for that is actually scheduled in the upcoming quarter. Whether it's going to release in that quarter is not for me to say now.