As per usual, anything said during the show is subject to change by CIG and may not always be accurate at the time of posting. Also any mistakes you see that I may have missed, please let me know so I can correct them. Enjoy the show!
30 bugs currently.
Kiosks are getting some tweaks after receiving feedback from testers.
22 bugs currently, 91 tasks remaining.
Miles Eckhart is getting some more animations to flesh out his animation set.
There was a bug where spawning a second ship would make everything go black for everyone on the server and no one could move.
18 bugs currently, 48 tasks remaining.
Working on 2D radar for ships as well as integration of radar into Starmap
Making the Starmap easier to use and more intuitive
Overhauling the mission app to allow for more details such as where it’s from, who it’s from, time remaining, payout.
Will allow you to see what missions are around you in space and track which ones you want to focus on.
2 Bugs currently, 18 tasks remain.
Working on space to atmospheric flight characteristic transition, second stage afterburner, gravlev polish.
Tuning normal flight velocity control to afterburner control to prevent instant transition.
Performance isn’t a single patch to make everything smooth, but constant tweaks over time.
Studying Evocati performance captures carefully to find entities eating up performance.
Quantum drive now is quite linear, but Chris Roberts wanted a ramping up especially with the larger maps involved
By using one higher than acceleration – the proverbial 11 – one gets a constant jerk and the desired curve and by then basing the quantum special effects off of speed, it gave the desired curved effect
Converting quantum travel to 2.0 required changing old code to get to use the native Item 2.0 advantages
Original effects for quantum travel consisted of more of a tunneling effect, now a residue appears from the front of the ship and moves to the back for a more Star Treky sort of effect
Chris wanted more colour mixed in to effects
Tags and triggers are provided by game code like spooling, entry, the traveling effect and flash effect
Five ‘phases’ of Quantum Drive
Initial ramp from current speed to afterburner
Ramp up phase to figure out where your maximum speed is
Middle phase is cruise, hitting maximum speed of your drive
Last two phases of ramp down and speed down
The math it took to create the five phases and have it calculated accurately to be able to track the ship throughout all the phases and know where it will ended up was a huge challenge, but they pulled it off by using something called a Quantum Snapshot which while has no performance impact, is able to know where you’re going, how fast it will take, how fast you will be going in all the phases, and where you will be throughout the phases.
The formula scales well with multiple players, and cinematics.
Sandi Gardiner (SG): Hello and welcome to another episode of Around the Verse, our weekly look at Star Citizen’s ongoing development, I’m Sandi Gardiner.
Chris Roberts (CR): And I’m Chris Roberts, I’ve just come back from Europe.
SG: Yes, he has. Today’s feature sets its sights on the quantum travel system, with so much distance between celestial objects, this feature is an essential part of where your journey starts.
CR: It sure is but before we get to that, let’s see what progress the team has made on Alpha 3.0 in this week’s installment of Burndown.
Eric Kieron Davis (EKD): Welcome back to Burndown, our weekly show dedicated to reviewing progress on the release of Star Citizen Alpha 3.0. We’ve been working through each of our designated categories as laid out in our previous episode as well as the production schedule which are: shopping, missions, ship and vehicles, traversal, mobiGlas, and overall performance and stability. Let's go to the team and see how each are progressing.
Erin Roberts (ER): So commodity shopping polish?
Todd Papy (TP): So basically there is… Chris prioritized some of White's tasks. There’s some Simon White polish that is needed, however Chris pointed the item 2.0 stuff taken care of first so that there’s just the final tasks that are on Simon White.
CR: So basically it feels like you’ve got stuff to get like all the features in and the content in and then you’ve got polish stuff like the shopping or the ships right?
TP: Fair enough, fair enough, okay.
Steven Egan (SE): So these are the kiosks we’ve been looking at, these are the original ones, they’re very kind of… the feedback we’ve got was that they’re very tacked on, the screens are a bit small so the experience isn’t necessarily best for the player. So basically we’ve been looking to take this generic kiosk and create something a bit more bespoke for the stores. So it’s very close to being complete, the template is all set up so we know that any work that we’re doing now can get plugged in and it will work. This one here is finished for Port Olisar.
I’m currently working myself on the counter for Levski, or the kiosk for Levski which is almost done, members of the team are pushing hard to get the counters for different areas: Bazaar Bargains, no sorry not Bazaar Bargains, Dumpers Depot, and other stores from around the game will have their own unique and bespoke kiosks and shopping counters and stuff. Yeah it’s very close to being complete.
Ricky Jutley (RJ): Just to move on to above the physical shop polish, and actually Eckart and Ruto. Todd’s got to sync with Jake and Reininger for the work being done on that because he’s still got some feedback.
TP: I already did.
TP: That’s what I was doing before. So jake is very confident they’ll be able to finish it in that time.
Rob Reininger (RR): So for 3.0 we’ve got several NPCs that we’re working on, a couple of which are mission givers. We’ve got Eckhart whose obviously on Levski which you guys know about. We’ve got Ruto whose also been coming along as well. These are both two very important characters because they’re going to be responsible for doling out some of the content you guys are going to have access to.
We’ve been capturing a few more animations for Eckhart which is going to help fill out his set of animations and allow us to do the intro conversation and the follow up conversations after that, so really separate those kind of experiences for the players.
We’ve been handing this work off to the Frankfurt office. So we’ve got Vitis, Kunis, we’ve got Chris Speak who’s also helping out, Dan Trepids team. So it's really been kind of a global effort to get these guys together as we focus on closing out the shopping here.
Matthew Webster (MW): So QA literally like 10 minutes ago just gave me this issue where spawning a second ship blacks out the entire environment and removes all movement controls for everybody on the server which is absolutely bizarre.
Dev1: That is such a weird..
MW: Yeah exactly.
Dev1: bouncy bug, it’s like…
MW: So this was present on the A14 build that we kicked up before launch.
Chad McKinney (CM): And I’m looking at the spookiest bug which is if you spawn a ship and then a second ship, all the lights and all the stars go out and you disappear and everybody, this happens to everybody and you can’t move.
Developer Calix Reneau (DVR): The programming head is working on the overheat field on the heat MFD. I have discovered, I noticed we aren’t displaying the global heat pool anywhere which is a thing by the way, so we should probably have that so I’m going to be adding that in today, probably talking with Max about where I get that information from that I wasn’t able to track it all down last time, but that’s the current thing I’m working on.
Matthew Lightfoot (ML): And at the minute for the closedown of 3.0 I’m looking after three main chunks of work that really is six individual features. These three segments of work are: IFCS and flight balance and that’s comprised of atmospheric flight, second stage afterburner, and also gravlev polish and tuning.
So these were two tasks that we checked with Todd this morning which we pumped to the epic, which was the modified atmospheric flight model which was the first one we talked about. The second one was the atmospheric switchover triggered by the level of ship drive was the second one I mentioned. As I understand it..
Paul Reindell (PR): And Todd said this is high priority.
ML: So Todd said that the proposal that we took to Todd, the first one, the modified atmospheric flight model was the one that we wanted. The second one which was the atmospheric switch over hopefully will be easy to get, however if it runs into any problems, we’re also to push that to 3.1 so the minute that becomes a longer task than anticipated we can shove that to 3.1.
David Colson (DC): Recently I’ve been like… since that meeting I’ve been as I’ve said smoothing this transition and modifying the accelerations of the flight control system as you’re flying around so that you don’t get these sudden changes in acceleration which you feel very strongly as your ship suddenly feeling not maneuverable and before that it was a case of finding some sort of a system where we could consistently control this and for example we need to deal with decoupled and that was quite a challenge because currently coupled and decoupled fly quite differently and the system needs to work consistently on both sides. So I’ve put a lot of time in making sure you can use it in decoupled and it sort of behaves similarly so that you’re not like… You don’t want this huge discrepancy between how it works in two different modes. So we sort of worked and I worked and implemented the system where it will only go to the higher speeds if you’re in a straight line and if you try and modify your velocity it will kick you out or try and slow you down again.
Andrew Nicholson (AN): Because of some of the issues we’ve had with this afterburner velocity on the straight line going up, we found that you would easily switch but from SCM and you’d want to kick into afterburner and you find yourself going really fast into these higher velocities. One of the ways we considered fixing that was to go back a little bit on some of the the stuff we’ve done before where you have SCM on a much smaller scale. We want to give the combat maneuvering a larger range so that you wouldn’t need to rely on afterburner quite so much. So I’ve been doing a large on pass on all the ships in the game right now on their SCM velocities and all the accelerations based around that.
Basically scaling them up so you have more velocity to play with, but I haven’t messed with the acceleration so that should carry it over from the current build.
RJ: On the actual Starmap basically Leos done most of his stuff on the code side and Todd said it was okay, the main thing is just cause we rearranged some of Zane's stuff we just want to see where some of Zane’s stuff comes in now.
TP: Starmap in the mobiGlas. Starmap in the ship is, scroll down Ricky.
RJ: It’s a later sprint that one, the ship to radar.
TP: It’s just under ships and to where that one is right now is Oleg is finishing his work for basically what a 2D one would look like versus what a 3D one would look like and then at that point ship art and tech art in sprint with us right now so they can do their work needed for the radar item and get those things positioned so then Bone and his team can do their work.
Leo Vansteenkiste (LV): The past few months I’ve been working on getting the radar into the Starmap system and have decided to go kind of in a component based system so if you add Starmap then you can interact with Starmap and if you add ontop of that radar component you can zoom in to yourself and suddenly this radar pops out kind of and if you remove the Starmap again then you just see the radar on your display. So that’s kind of the way I’m trying to go and we can always add different components to that so it will expand the functionality of this display.
Zane Bien (ZB): I’ve just kind of wanted to go over with you the switching of galaxy and local area because I think most times you see horizontal zoom sliders, usually the macro if use to the left rather than to the right so that’s just kind of why I had that on the galaxy on the left side.
ZB: Yeah. Then there’s probably some stuff I need to clean up in regards to the nav map and radar scale button so that they’re clearer that they’re toggles, but in general I think it would be a good idea at least to get Chris’ eyes on this if he hasn’t seen this yet.
TP: Yeah definitely.
ZB: The textures look different.
CR: Yeah I want to have, yeah. So for the cases where you press buttons and turn you know dial turn knobs and things like that I think it should be diegetic, like you actually have it labelled in the universe and you can see it and then that’s why you have that thing, the finger press press icon that use it when it’s there and optionally you can have translation which is the case here and then when something becomes a bit more complicated, you have more than one thing you can do with it then it would have a multi option.
Hugo Silva (HS): So the mission app manager in the mobiGlas went to review, we got some feedback and we are changing some stuff we got from the reviews. We have polished a couple of features that we’ve got.
So we now have the ability to actually view the remaining time that a mission has. We have all the information, the description, who is it from? How much time? What’s the payment going to be? We can accept these missions, we can track them on our accepted list. We can also untrack them now. There’s the ability to know which missions were actually tracking. We also have the possibility when abandoning a mission, we can actually go to the history and actually know now which missions are abandoned, completed, midway basically. If we see here like we have now information about all of these missions.
It’s a much sleeker user experience that they have now making it easier for the players to actually keep track of all the objectives that they need to do to finish the missions, actually having a proper way to manage their missions. Imagine if they’re halfway through the mission and they see like, Oh I have a really nearby mission I can do, they can actually easily manage these options.
Niklas Munck (NM): As for the performance topic, it’s of course that the closer you get to release, the more important it becomes, but on the other side you have to see that performance is nothing that will be done in one point, it’s not like a feature we can complete and then it’s done, it’s something we will always have to deal with that’s overarching the whole project and as I said it’s of course becoming more and more important the closer you get to a release because we of course want to provide the best experience for the players, smooth.
That’s why we’re now looking into the performance captures from the Evocati players. Basically always so it’s not just testing a specific feature, but we’re always looking into those captures and see where we can possibly shave off another millisecond or another frame, where can we possibly save something or optimize so get the smoothness to you.
EKD: This week alone we’ve checked in 756 updates to our 3.0 branch in perforce in every possible category. You’ll see from category to category how unpredictable progress can be as the Evocati testers and our internal QA work together to uncover new issues while leadership evaluates what’s absolutely necessary to address to release 3.0 to PTU as quickly as possible.
So come back next week to see the progress we’re making on Burndown.
SG: If you would like more info on our schedule as we work towards the upcoming release of Alpha 3.0, be sure to check out the website for our weekly production report.
CR: Definitely do that but now it’s time to spool up this week’s feature on quantum travel. Where we look at the updates coming in Alpha 3.0 that make the system more dynamic.
SG: With the old quantum travel system, you would select the destination, spool up the quantum drive, zip through space to that location. It was the same experience regardless of where you were going.
CR: Yeah, works but considering how distance is going to be much more of a factor moving forward, 3.0 and beyond, we knew it was something we wanted to improve. That’s why we decided to very your travel speed and effects based on the distance traveled. That way a trip between Earth and the moon would look and feel very different then a trip between Earth and Jupiter or in our initial case the Stanton system. Between somewhere like Levski and I don’t know…
CR: ArcCorp perhaps, yes.
SG: There you go. To learn more, let’s check out our feature on the updated quantum travel system.
Michael Dillon (MD): Anywhere you want to go that's not like a hundred feet from you, you're ... probably a quantum going to be involved. So, there was the old quantum system, which I believe is what's up on the live server now, cause we're still working on the new stuff.
Mark Abent (MA): You look in some random direction, and you fire off, and you go. It was great. Had visual effects and everything. Problem is the designers weren't happy, and C.R. wanted it to be a little bit different, because we're going to have bigger maps, and he wanted it to function differently. The previous system basically said, “I want to fire from here to here and go.” Set max velocity, everything, and you just went. There was no concept of ramping or … Well there was a slight concept of ramping, but it was very linear. So if you was a short jump or a long jump it felt the same, and C.R. wanted the experience of having something completely different where if there was a moon and I'm going … say from the Earth to the Moon. It's far, but it's considered a short jump. [Finger Quotes] A long jump would be from Earth to Pluto. So from Earth to the Moon he wanted a kind of feel like a little bit slower, cause you don't reach that max top speed, while as the one at Pluto you reach that top speed, but he didn't want a linear increase to it. He wanted more of a dramatic build up to get to there.
So we were experimenting with a bunch of ways of getting to that point, and one of the ideas that came out in the very early prototype staging was the concept of a … really nailing down the concept of a short jump and a concept of a long jump, and the first implementation was just a binary system. By that I mean, “Oh if we're at this certain distance we're a short jump. If we're at this distance, we're a long jump, and what that gave us is different speeds. So when you went from Earth to the Moon, it took a little bit longer to get there, because you weren't going as fast. You could kind of enjoy the scenery, and you'd just ramp, and then when you went to Pluto we gave you a much bigger speed so you just zoomed by to get to there, but you still have the problem of it being linear, and it didn't have that build up. You just started increasing, you got there, and you just stopped. We're missing that ramping up smoothness, and since it was a binary system if you have the distance just slightly off it would change from short jump to a long jump. So if I'm going from Earth to Mars, if that distance was just enough and I flew a little bit, it'd be considered long. If I flew back, it'd be considered short, and since depending upon what jump you're going to do the effects, the sounds, and all that was completely different, and it didn't scale at all, because it's one or the other.
So went back to the drawing board going, “Alright, he want's that ramping and design really wants that curve, and what we're doing was we use motions … we use a constant acceleration”. So you start out at a velocity and you had an end velocity and you literally did a linear curve as you accelerated over time or your velocity increased over time, but that gives you that straight line which is not what we wanted. So, what design wanted was like a little curve up. You could think of it like a … What was that movie? Back to the Future. You're in the Delorean, you put the pedal to the medal, and you get to 88 – that gives you a nice little curve, but then soon as you hit that 88 you ramp up like crazy. So, you want that initial speed-up so you can kinda go, “Alright, I'm accelerating. I'm accelerating”. You hit that boom phase, and it just goes whoosh.
So, went back to the drawing board, took a look at it and said, “What can we do? Well, alright. Acceleration is giving us linear. What happens if we go one higher?” In that famous movie we'll go to 11. Well, in this case we'll go one higher into the thing called jerk. What jerk is … we have a constant jerk and we're having velocity and acceleration change over time. So what that gives you is instead of your velocity being like this, it goes that little curve. So, if we use jerk to figure out how fast we're going to go at any given point, we can get rid of that binary system and we just choose a position and we know the maximum speed of the drive and we know the maximum … the minimum speed. We could calculate the min acceleration, the max acceleration based off the jerk, and ramp us up.
So what that gave us is no matter where we choose on the Star Map … whatever location, we'd get that nice little curve. So if I was going to make Mars, I would accelerate and hit this click … max … the speed and then decelerate again, but if I was going to Pluto I'd go all the way up, even higher speed, cruise a little bit and then go down. So it gives you that more short term long jump hybrid sort of experience, and because it's no longer a binary or system where either short or either long it's based off of the distance we could tweak all of our effects to be based off speed. So, when you're accelerating up if you don't hit a certain speed, you don't get the tunnel effect. You just get these little sparkles, cause you're not hitting that … I don't know … I guess sound barrier of the quantum. You're not hitting the quantum barrier. You're going just right before it. You are going faster, but you're not hitting the barrier to break the sound. So you had the little effects, but then you dipped out of it, but if you go a little further out you hit the barrier, so we trigger on the effects, and it makes it much a more dramatic thing. So it gives you that … alright, I get to a system. There's a planet. I select one of the points. I could get closer to it, but I don't see the tunneling effect, but if I go pick out something further out … go bam! Go out and see the tunnel. It's quite a cool little interesting thing, and while doing this since the math was kind of involved we actually prototyped it out in a completely different executable so we could give designers saying, “Hey, plug these in. See the graphs. See the velocity. See the accelerations. It's pretty cool.”
MD: When I came on my first roll here was starting quantum travel conversion to 2.0 At the time, yeah. I mean it did obviously fly at the time, but it wasn't able to integrate well and it would be very … it'd be a totally different set of code basically from how everything else works. So bringing it over brings all that together. We get to use all the connections and stuff we natively get being Item 2.0, plus the old drive while it worked had been kind of built alongside the design so there was a lot of code in there that while it worked wasn't necessarily needed to go a certain way.
It was really more getting it hooked where we could attach it to a 2.0 ship but also getting it connected to the pipe system for power, heat… for things of that nature, for the fuel system. The old drive didn’t use any of that mechanics because it wasn’t compatible. So with the new system as a… Max is building out pipes for fuel and power, I was able to start attaching those as well.
Aran Anderson (AA): The effects for the quantum travel consisted of this tunneling effect originally. So Chris has asked for more of an adaption to this so we had… he wants more of a Hollywood feel, Star Treky sort of effect where it’s residue appears from the front of the ship and moves to the back of the ship. Then it just starts with the spooling stage and hits acceleration rate which is as Chris puts ‘reaching 88 mph, like in Back to the Future’. So once you get through the spooling stage, it’s a mixture of residue that starts to build up from the front and it just builds up more and more and then it just flares up and then hit a big bang which is like the entry effect into the quantum travel. The starfield effect was this effect with stars appearing, whizzing past the planets and sun all the way to your destination as you travel. This new effect we added in is… adds more of a residue from the front of the ship, cones around the front of the nose of the ship and then leaves more like whizzing flares kind of what couldn’t be done with the original CPU system. So had to use this new particle GPU system add in thousands of dots that throw off residue.
What we originally had was this blue sort of plasma effect but then Chris wanted to add in more colour so we’ve brought in some of the colours from the engine trails which we’ve recently added. So we’ve got like for the Mustang for example or Freelancer, we’ve got this like tealy blue effect that mixed in with the blue plasma, the M50 had more of a yellowy orange and blue effect mixed in.
So the game code provide us these tags and triggers which are like spooling, the entry, the traveling effect and the flash effect. So these all trigger at certain parts of the stages so the spooling would trigger between the acceleration phase and when you enter. Then the enter effect would trigger when you enter to traveling, then the traveling effect would as you’re traveling from your destination to your other destination. Then once you’ve hit your destination it hits a flash entry point… flash exit effect and then it the exit effect is this basically a big flash lense flair with a big explosion sort of whispy smoke trails and traveling from where you exit basically.
Mark Abent (MA): Coming back to what quantum drive is, we have two parts to that specifically. One, design aspect… what does design want? What does CR want? That was the slow ramping up and giving that little curve and coming out of that, you know, we play with the accel we play with the program to figure out the math involved. The second part is, ok we have this math it’s Jerk there’s formulas for it. We have pages and pages of how we can calculate it which is cool but how do we translate that into the game. That’s always the tricky part, now fortunately for us Jerk, constant Jerk with acceleration and velocity, we know certain parts of the formula so we can adapt that into programming into different parts. How we did that was actually five phases of quantum drive… there’s the initial ramp which is you’re going from zero to that 88 mph, we’re actually going to replace that to going from your current speed to afterburner, cause we finally have afterburner. So you can think of that as going from zero to afterburner or whatever your speed it. That’s just a linear velocity so that’s one block where we have to calculate the distance and the speed, so once we calculate that we put it here.
Then we have the ramp up phase which is where introduce the Jerk of your curve to figure out where your maximum speed is and we calculate that stuff, we put it here. Then we have the middle part which is the cruise, so if you’re actually to hit the maximum speed of your drive, we have to stay in this top speed for awhile and then kick out so we put that here. Then we have the ramp down and the speed down, which are the same as the speed up and ramp up but just inverted to go from your top down. So we have put all those on the side and we go, ok what do we need to get this functional.
We have all the math, we have all the time, that’s another fun thing. We need this system to be predictable, so if I knew the start time and the end time I need to know where I am at any given point. We needed an analytical solution, lucky for us those have all been derived off of Newtonian physics for years which is awesome. So knowing those key five parts, we broke it down into distance, so we know… so if we know the distance, we could do a couple thing. We could figure out the speed up phase. We know the velocity, we know the current velocity… we can figure out the distance will be during in there. We could figure the distance it would take to hit max and then we do that on both sides, so we leave the cruise out and that gives us the ability to do two things. One, if we’re able to hit top speed it… if we’re able to hit top speed it makes the formula so beautiful, we can do the easy task. If we can’t do the top speed, this is where it gets complicated because we have to solve, I think it was a fourth order polynomial and that math just bonkers, I’ll spare you the details. We had to solve that fun guy using that very quick and dirty solution to make sure that it’s performance… none of it is performace heavy. So we calculate all the distances that you’re going to be in in each one and from there we can calculate the time and then we split it out and save the information and it’s called a quantum snapshot. So we figure out the distance you’ll be, we figure where you’ll be at any point in time and from there when we pass in the time we could go, all right at this time we’re going to be at phase ramp down. So ramp down phase and then we can go, oh no we’re going to wind back we’re actually here. So we have this time which allows us to go kind of backwards and forwards if we wanted to. One, this is for networking. Two, if we wanted to for trackview for a cinematic sequences we can wind back in time. So it makes it everything nice and easy and beautiful but it took quite a long time to figure out the math, figure out how to break down everything so it works in networking, works in multiple players, works in trackview, works in cinematics. One solution that works in all the fun details but because it’s timeline based we can rewind back and forward, we can play effects depending on where we are in the timeline and it just works beautifully. Lots and lots of coffee to get here.
CR: So as you saw it took a lot of tweaking and complex math to get the quantum travel system working the way we wanted it. Now we have a system that provides a more dynamic and varied gameplay experience for Alpha 3.0 and beyond.
SG: And that’s all for today’s episode, thanks to all of our subscribers for helping make Around the Verse and our other shows like Bugsmashers and Citizens of the Stars. This month’s subscriber flair will also be released on Friday so be sure to keep an eye out for it.
CR: Yup, definitely do that and a big thanks to all our backers for your support. This game is only possible because of your dedication and enthusiasm allow us to make a first person universe like any other. So thank you very much.
SG: Finally, be sure to tune in to Happy Hour game deve tomorrow at 8 am Pacific to watch Level Designers, Andreas Johannson and Tobias Johannson construct a surface outpost live.
CR: Yes, they’re not even related.
SG: They are not.
CR: Same last name.
SG: They sit next to each other and yup…
CR: Until next week we’ll see you…
CR/SG: Around the Verse.