Sep 22

Around the ‘Verse: Episode 3.08

This post is a transcription of Around the ‘Verse: Episode 3.08, material that is the intellectual property of Cloud Imperium Games (CIG) and it’s subsidiaries. INN is a Star Citizen fansite and is not officially affiliated with CIG but we reprint their materials with permission as a service to the community. Enjoy!

As is with any information on our transcripts and summaries, everything posted is subject to change by CIG and in some cases may not always be 100% accurate at the time. While we strive for accuracy, mistakes do happen so please let us know if you find something amiss that we didn’t catch. Enjoy the show!

Around the ‘Verse: Episode 3.08 Transcript

Intro

Chris Roberts(CR): Hello and welcome to another episode of Around the ‘Verse. CIGs official weekly look behind the scenes of Star Citizen, I’m Chris Roberts.

Sandi Gardiner(SG): And I’m Sandi Gardiner. We’re what, three weeks away from CitizenCon?

CR: Yeah! Well, two and a half really. Everyone’s working really hard.

SG: They are and not just on CitizenCon though, we’ve got teams hard at work on 2.6 which as many of you know includes the long awaited Star Marine game mode.

CR: Yup that’ll be good to have in your guy’s hands. We’re also aiming to have new flight balance changes going out to the evocati test group to get some early community feedback which is just one part of the Arena Commander Improvements for 2.6.

SG: And we’re deep into prep on performance capture shoot for 3.0 which is currently slotted early after Citizencon so yes, lots to do.

CR: Yeah in fact I’m getting on a plane a few days after CitizenCon to do that, so there you go.

SG: Very busy. On Today’s episode we have Evo Hertzig explaining all the tech that went into the vision stabilisation in FPS which we shared last week with you all.

CR: Yeah, everyone really liked that, but first let’s head to our Frankfurt office where Brian Chambers will update us what they’ve been working on.

Studio Update

Brian Chambers (BC): Welcome back to Germany. I am Brian Chambers, Development Director of the Foundry 42 Frankfurt office – this update from Frankfurt we will start off with the weapons team, where David will walk through some of the recent weapons they’ve been working on.

David Sibbe (DS): Hello, my name is David and I’m a Junior Weapons Artist here at Foundry 42. Today I would like to show you some of the weapons we’ve been working on for the past couple of weeks. These weapons are still work-in-progress and could slightly change until the release of Star Marine. We are working on the complete rework of these weapons for Squadron 42 – one of our goals is to generate a uniform style for the manufacturers with similar materials and overall look, it’s important for us that you can see what manufacturer created the weapons from distance, also for gameplay reasons – we were making a polish pass for existing weapons with PBR materials using new software and we changed the look of existing weapons and applied the style guides of weapons you would see in future. So I hope you like the quick look at the weapons we did for you and see you in the future.

BC: Thanks David – it’s always cool to see weapons with their polish pass – look forward to playing those in the game. As you guys know, Star Citizen universe is vast and we need to create systems that are going to allow for the scalability so we can populate the universe to the level of detail that we need and have the gameplay in there. With that said, let’s go to the level design team where Ben and Tobias will show off some of the modular systems they’ve been working on.

Benjamin Dare (BD): Hello, I’m Ben, I’m a Level Designer here in Frankfurt and I’ve been working on a modular system from which we can construct satellites. Satellites conserve a huge number of different purposes and they are going to be really cool when they are in-game. You can think of all the different gameplay options that can come out of these diverse array of satellites which we can build. Now, my job was to develop a system whereby we could build numerous different satellites from a number of modular pieces. The purpose or the function of a satellite will be determined by the modular pieces which it is made up of. So when I started on this modular system, I created a number of test modules which serve different purposes – like one would be shields, another power generation, and in a lot of ways their functionality is similar to ships but the difference being on satellites all of these systems will be automated.

It was important when designing these that the silhouettes of the modules were distinctive from one another. What this means is that as a player when you are arriving at a satellite, you can take a quick look at it and instantly grasp what it does – what modules it is made up of and what it’s purpose is in the universe. When these individual modules get damaged their specific functionality will fall out of the satellite – so if it’s shields then shields will be down – then a player will have to EVA over to those modules to fix them and bring that module back online – other interesting things could be stealing data from a satellite and this data could be breadcrumbs that could lead to a mission. The size and the shape of the satellites themselves will be influenced by where they are in the universe. For instance, a satellite very close to a sun will need a lot of shielding – it’ll need some sort of liquid cooling. All these systems to be able to dissipate the heat into space. You can see it is easy to swap in and out the difference modules of the satellite.

The hub in the middle is the backbone from which all of the other modules branch out. The type of power generation will be dependant on the purpose of the satellites – how much power it needs to cool and also for instance a military satellite will not want to have solar panels because they are liable to get damaged. Each of the modules you can see here has six connection points – this allows the satellite to be build out in every different axis and when the art team have got hold of them they’ll look a lot more attractive and in keeping with the rest of the art style of Star Citizen. Thanks for listening guys, bye.

Tobias Johansson (TJ): Hi, my name is Tobias Johansson, I’m part of the Level Design here at Foundry 42 in Frankfurt. I have been working on a modular system for small planetary locations for 3.0 – humans have always wanted inhabit new locations or find them and it’s the same in space. So for this first iteration we are just going to have small camps and research facilities that are placed on planets for humans to live and stay. And now I’m going to show you this- how this system is gonna work in rough ideas here in CryEngine behind me.

So, with this system that I’ve designed so far is- we’re gonna build a location using components. We start picking an exterior – as you can see here I just have a small one. As I told you this is what we’re aiming for in 3.0 – after you pick an exterior, you go inside and you gonna be able to switch between different interiors depending on what you want the location to be like. So as you can see I’m switching between different rooms that we’re just going to be able to scroll through fairly quickly.

After deciding what we want the location to be – or the building – we can also add some exterior assets to it to make it more distinct and recognizable from far away. This is all used to how the gameplay and everything that this is built around so they- we can populate planets faster, inhabit planets faster.

As that is just what we’re aiming for now, but as you can see, this aswell can be scaled up to larger buildings with more slots – this is also a building with just one slot and here we have one that has three room slots where we should be able to switch all the rooms between as we want to.

For 3.0 however this is more what it is going to look like here, where it is these small camps with just one or two of these smaller buildings and a corridor connecting them so you don’t have to go through an airlock every time you walk between buildings and here we have another one that’s just three slots.

As you saw in the presentation, you get an understanding for how we can faster build our locations that we’re going to place on planets – we still have to place this manually however this system is going to make it faster so that we can fill up the world, make it feel more alive and then more gameplay for you guys.That’s all from me this time, and thank you.

BC: Thanks guys, I look forward to making all the variants you guys have been putting together and how we’ll eventually encounter them as you jump from planet to planet. For the rest of the team, they are incredibly busy and we look forward to showing off more work from other disciplines in the very near future but for now that wraps us up for Frankfurt – appreciate you guys watching and thanks so much for the support.

Back to Studio

CR: Thanks Brian, so those reworked weapons were looking really good.

SG: They were, they were looking very cool. I feel like we’ve talked about modular sets a lot on AtV.

CR: Well, you know, it’s one of those systems that you have to have when you are making a game the scale of Star Citizen.

SG: Cool. So in the spectrum between completely custom locales where everything is hand crafted and the opposite end of the spectrum which is procedurally generated locations where would these modular systems fall?

CR: So hopefully best of both worlds – the modular sets allow us to build out the different locations very rapidly and then we dress them up with props and banners and different art to give them an individual feel and character.

SG: Nice. So next update, let’s check in with Tyler Witkin for this week’s Community Update.

Community Update w/Tyler Witkin

Tyler Witkin (TW): Hey everyone. Tyler Witkin, Community Manager, in the Austin, Texas studio here to bring you this week’s Community Update. Last week the Vanguard Warden won the title of Galactic Tour’s Combat Ship of the Year landing in the spot on our pledge door for one week and that sale will end tomorrow. Fast forward to this week and the battle continues. This time between the Aegis Retaliator and the Anvil Gladiator competing for the coveted title of Galactic Tour’s Bomber of the Year. It’s looking to be a very close finish, so make sure to log in to our website, cast your vote and we’ll post the results tomorrow.

Now I said it last week, and I’ll say it again. The Bar Citizen fever is spreading around the globe. Last Saturday I had the privilege to attend the Bar Citizen event in Orlando, Florida, and let me tell you it is an unforgettable experience. We have some upcoming events. One in Denver, one in New York and even one in France. You can find out all the details about those events and more at TinyUrl.com/BarCitizen. Now, it’s convention season. CitizenCon is right around the corner, but first TwitchCon. Alexis, Ben, Jared and myself will be present wandering the show floor, so hopefully we’ll run into some you guys while we’re there. Last week also brought us a new issue of Jump Point for subscribers packed with awesome content and even a fun in-fiction piece that’s definitely worth checking out.

Now it’s time for this week’s MVP award. A huge congratulations to Utho Riley for his talented efforts in creating some fan Star Citizen music. Browsing through his YouTube channel piece after piece continued to blow my mind, so congratulations, Utho. You’re this week’s MVP. And lastly the week would not be complete without Reverse the Verse. We’ll be broadcasting live at Twitch.TV/CIGCommunity tomorrow at 7:00 AM Pacific Daylight Time, so make sure to tune in to catch all the talk about everything that you saw on today’s episode. Thanks again everyone for your support, and we’ll see you in the ‘verse.

Back to Studio

CR: Thanks Tyler. So love the passion and creativity from this community. I mean they do amazing stuff all the time.

SG: They do amazing stuff. In last week’s newsletter the sneak peek was a video demonstrating the changes to our Vision Stabilisation System.

CR: Yes. So as a first person universe, the transitions between flight and ground base gameplay, obviously the players perspective is an essential element to get that right. So you have to have it feel both smooth and realistic.

SG: Without motion sickness.

CR: Yes, that too. So anyway, it’s been an ongoing process of trial and error to get it right.

SG: Very cool. So here’s Ivo Herzeg to explain the tech and the improvements that we’ve made.

Behind the Scenes: Vision Stabilisation

Ivo Herzeg (IZ): Hi, I’m Ivo Herzeg, Lead Animation Engineer at CIG and today I want to talk about experiments we did to get the camera stable in the first person view.

The first person mode we have in Star Citizen is a bit unusual because it is unified with the third person body and it’s also driven by the animations of the third person body. And it took quite a while to get this kind of setup to work. If you just take a camera and put it at a point between the eyes of the third person model and move around, then it is pretty crazy what you get on the screen.

So that’s how it looks with a camera attached to the head. As soon as you start to move you get you get this crazy camera shake and it almost looks like the camera is randomly bouncing around. So this is nothing we can use in the game and it’s certainly not how humans perceive motions in real life. But the funny thing is this is actually the real motion of the head: this is all part of the mocap data. If you take a GoPro camera, attach it to your forehead and move around then it looks exactly like this. So this means head motions can be pretty extreme and it also means that our eyes and our brain are doing a pretty good job to compensate it. That’s why we are not even aware that this issue exists in real life.

And the first thing we did to improve this was eye stabilisation. And that’s something humans do all the time. It’s basically a counter-rotation of the eyeballs to compensate for the movement of the head. And this is pretty easy to implement: all you need is a camera with a focus point and that’s just a point in the distance you can control with the mouse when you’re aiming.

So that’s how it looks with eye stabilisation. When I move around everything feels much smoother than before and the image is now perfectly aligned with the horizon. And by using this focus point we are actually trying to emulate the biological principle that keeps the image stable on the retina of our eyes. But instead of counter-rotating both eyes we only counter-rotate the head camera in the direction of the focus point. And with this simple trick we can eliminate about 80% of all the camera noise.

But if you look now on the screen, you can see the image is still not perfectly smooth and that’s because eye stabilisation only works well if you move forward in an open environment and focus at a point in the distance. It doesn’t really work when you move close to walls and it’s something you can’t avoid in narrow corridors. And we have lots of them on a spacestation. It also doesn’t work when you straffe in front of walls. And strafing is something we do all the time in a first person game. And it looks pretty terrible with all the stop and start motions that we have. So when you start to run and stop directly in front of an object then you get this “bounce back” effect.

So all this is nothing new – we already had it in PTU 2.0 – and nobody was really happy with the result, mainly because of all the camera noise and head bobbing. And then some people started to complain about motion sickness and for many it was just an immersion killer and not realistic at all. And this all makes sense because eye stabilisation, if you look at it as an isolated feature, it can only compensate the rotation of the head and that’s it. It doesn’t help much with the translations of the head but that’s why we still have all the trouble with the camera.

Now the interesting thing is these head translations: they happen all the time in real life and it’s never a problem. When I walk into the kitchen to get a coffee then my whole body and my head is moving of course, but the whole motion feels more like floating through the corridor. And maybe that’s what we expect to get on the screen. But if I repeat the same experiment with a camera attached to the head, and even if I use exactly the same mocap data, the result on the screen looks more like an earthquake.

So this issue was keeping us busy for a while and we spend some time trying to understand how humans are doing vision stabilisation. But it turns out it’s a pretty complex mental process and there wasn’t a practical way to get that into a first person camera. So that was a dead end. And all the techniques to stabilise handheld cameras: they don’t work in our case because our camera is attached to the head and driven by a human body. So we had to find something else.

And we found something it was just a bit unexpected.

We learned that birds, or at least most types of birds, they have a pretty interesting problem. They can’t roll the eyes around the way humans can and that makes it a bit hard for them to keep the vision stable and move the body at the same time. If you can’t keep your vision stable by moving your eyes, well then the next logical step is to try to do the opposite and just keep the head stable. And that’s what they do: birds have a long neck so they just counter-translate all the body motions. It’s kind of a camera stabiliser invented by nature. But the really cool thing is this one operates only on joints and that means we have a pretty good candidate for implementation into an animations system.

Now to get all this to work with the human rig in the game we design a full body IK system to control the hands, the legs, and the head independently. And we only use the special IK only for the hands in first person. And what we see here is a combination of eye stabilisation with head stabilisation. And stabilisation itself happens mainly on the head camera and only for a few extreme motions we distributed the rest over the entire body. And the adjustments on the body are only a couple of centimeters. But as you can see this is already enough to keep the image perfectly stable on the screen.

The unified rig is not the only unusual thing in the shooter. Another pretty big difference is how we spawn the bullets. In a typical shooter they come from the center of the camera and they go to the center of the screen. Which means you are actually shooting with your eyes and the gun in your hands has no practical purpose. But in Star Citizen we spawn the bullets directly from the gun barrel and that changes a few rules.

For example iron sights and scopes don’t point directly at the target so you always need to aim a little bit higher, maybe two or three centimeters. And because now all body motions have a direct impact on the gun barrel, running and shooting doesn’t works so well: there’s way too much bullet spread going on. And the same goes for recoil: when you’re firing too fast while the gun is recoiling then it’s very hard to hit something.
And as a player when you are moving around in a shooter you always have this gun in front of you and most of the time you are not even aware that there actually is a body. And you only notice your own body when you look down at the ground and see your own shadow and how the feet are moving. And that’s the moment you realise that the recoil, all the hand animation, all the body animations – every single frame – is actually identical in the shadow because it is one single rig and all the animations are shared.

Back to Studio

SG: Thanks Ivo. So he mentioned in the beginning that most games have different animation sets between first and third person perspectives. Why was it important for you to unify them?

CR: Okay so Star Citizen is a multiplayer game and it’s very important to ensure that what you see in first person and what your friend standing next to you sees is the same thing.

So we do a lot of things for instance if where our gun is pointed exactly where we’re going to shoot as opposed to most first person games where the center of the screen is where you’re going to shoot, so actually in a lot of first person games you’re going to point not exactly in the same direction as where your bullets will fly. So that’s a problem, but also when you’re sitting next to it and I do something, you’ll see exactly me doing that thing and if you wanted to shoot me, not that I hopefully you, I wouldn’t want you to shoot me, that you would be able to where I think my body is, is where it actually is and where you can hit it and there’s a lot of games that doesn’t happen because they fake the first person view, it’s different than where the actual physical body is.

So sometimes you’re going to be playing a first person game and you’re going to be hiding behind cover and then get hit where you wouldn’t thought you would have been hit because your first person view is slightly different than what your body really is. So for us because we’re a multiplayer game and there’s going to be so many interactions, your friends are going to be next to you, say flying a ship like a connie or something and you’ll be seeing people do stuff, that we needed to unify it so whatever the character was actually doing they were doing, but it was also what you were seeing in first person and then the problem is if you place say a Gopro on someone’s head and you run around and you look at the footage, the footage goes like this and it’s really sort of barf inducing, that’s sort of the blair witch issue that you solve, there’s sort of found camera footage stuff.

So the vision stabilisation is just basically mimicking what our brain does when we’re bringing in the imagery because if you run around, your vision is quite stable. So we spend a lot of time to do that so we can have one unified set of animation and character assets whether it’s first or third person, it works and the view that you have is great and nice and smooth and feels more akin to what you do in real life.

Not many, in fact no games have done this, I think Arma is one of the few games that have managed to do it. We’re very proud of what we’ve done, it’s taken a lot of work. Most people don’t try as hard, if we were doing just Squadron 42 we’d probably wouldn’t have done it because that’s a single player game, but because Star Citizen is such a multiplayer game, it’s super important so we’re very proud and Ivo and the rest of team doing that did an awesome job.

SG: Very cool, I look forward to trying it out.

Outro

SG: That is it for this week’s Around the ‘Verse. As always we’d like to thank our subscribers whose monthly contributions allow us to make this extra community content.

CR: Yeah, thank you guys, and also to all our backers and supporters out there who got us here in the first place. Thank you very, very much guys.

SG: Yes, thank you. Be sure to check into Reverse the ‘Verse tomorrow at 7am Pacific, 3pm GMT, or 4PM Frankfurt time where Brian Chambers will be chatting with Ben Dare and Ivo Herzig from this week’s episode as well as revealing some exclusive new footage.

CR: Cool. So on next week’s Around the ‘Verse we’ll explain some of the changes to the flight model that you’ll see in 2.6. So make sure you check it out.

SG: Definitely, and thanks for watching everyone.

CR: We’ll see you..

Both: Around the ‘Verse

About the Author:

Leave a Reply

*