Star Citizen: Chris Robert`s new space sim (the Wing Commander guy)

Page 152 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Worthington

Golden Member
Apr 29, 2005
1,433
17
81
One thing I'm not liking at the moment, and it may be that they just haven't gotten it refined yet, is that ship physics in the game just isn't right. Ship movement is currently very, very raw and chunky. In fact it seems to have regressed a good bit from a year ago.

Also ship masses don't seem to be reflected in the game at all. I was watching the press demo video the other day, and for example the constellation - it just wooshed right off the landing pad and turned on a dime... just like the hornet did. Again, I'd expect this is just one of those polish pass refinement things that we're going to have to wait for, but right now it really doesn't feel good. I thought they'd have this farther along at this point than they do.

New flight mechanics are included with 2.0 which is (hopefully) very close.
 

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
One thing I'm not liking at the moment, and it may be that they just haven't gotten it refined yet, is that ship physics in the game just isn't right. Ship movement is currently very, very raw and chunky. In fact it seems to have regressed a good bit from a year ago.

Also ship masses don't seem to be reflected in the game at all. I was watching the press demo video the other day, and for example the constellation - it just wooshed right off the landing pad and turned on a dime... just like the hornet did. Again, I'd expect this is just one of those polish pass refinement things that we're going to have to wait for, but right now it really doesn't feel good. I thought they'd have this farther along at this point than they do.


As Worthington noted, this exact issue is being addressed in the new flight model being released in 2.0.

Here is a rather lengthy discourse on the changes:
https://robertsspaceindustries.com/comm-link/transmission/15031-Star-Citizen-Alpha-20

The switch to third-order motion control will be the biggest change and will smooth out the mechanical "jerky" motion that you're referring to that is a result of the current 2nd order motion, along with a modification of current jerk values and maneuvering thruster output.

Expect a very different flight experience in 2.0.
 

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
Well, most of the parts I need for my custom joystick have shipped. I need to pickup a new multimeter, but other than that, I think I have everything I need.


I'm thinking of picking up one of these:

ch.png


It's a mini Hall-effect thumbstick made by CH for industrial control applications, but it is available with native USB output - so I don't have to be smart to get it to work. ;)

I would like to stick it somewhere on my strafe stick to be used for free-look.


EDIT: Sooooo I got a quote back from CH on this...$185. Guess not. ;)
 
Last edited:

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
that's pretty cool! I might want one myself.


Here's another option I'm considering (even easier to install - and perhaps even better control).

Micro usb touchpad. Just stick it on with velcro. ;)

buy_usb_mini.jpg



Aside from free-look, I can think of other large-ship functions that it might be nice to have something like this - like controlling scanners or the "science" turret on my Aquila. Possibly even gimbals / weapons turrets - since on large ships it will be almost necessary to use them if you don't have NPC control over them.

EDIT: I ordered one. I'll report back with my findings.
 
Last edited:

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
Quick question... is ship controls more involved to the point I can't just use an XBOX controller?


I've not tried, but I can't see how you would have enough control over all the axes and necessary controls.

With two analog thumbsticks you could do pitch/yaw and vertical / horizontal strafe.

That means to control roll you would probably have to use upper triggers, and lower triggers for fire group 1 or 2.

That leaves you without any fingers left to control throttle, let alone reverse throttle and boost.

I think it would be extremely difficult to make it work.
 
Feb 4, 2009
34,506
15,731
136
I bet an Xbox controller would work it just may be awkward. I doubt the final game will be designed to require a high end flight stick.
 

Fallen Kell

Diamond Member
Oct 9, 1999
6,009
417
126
I bet an Xbox controller would work it just may be awkward. I doubt the final game will be designed to require a high end flight stick.

It is why the game works with keyboard+mouse. Grant it, they cheated to get it to work with the mouse controlling both the gimbal weapons and ship orientation at the same time, but it works. It just won't give you the immersion and some of the precision that a good flightstick setup will be able to do.


On that note, my controller boards arrived. Realized I forgot to order a USB connector. In a way I am glad since I just found these:

images


Which will let me cut out a small hole in the base of my joystick and essentially mount a proper USB port that I can then connect to the PC (the down side is that my controller will then have 2 USB connections, but I havn't found a small USB hub that will pass the power required by each board from a single USB connection, and would otherwise need a second wire anyway for a power supply connection).

I might add a pinky button or two to the 16000.M while I am at this, possibly just connecting them internally to one of the existing buttons on the base of the controller which would be hard to hit due to my second hand being on the mouse.
 

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
I bet an Xbox controller would work it just may be awkward. I doubt the final game will be designed to require a high end flight stick.


Here's the default mapping. It looks like they're using LT as a shift key to get a lot of functions. I think shifting to roll would be pretty awkward, but I haven't tried it. The biggest issue I see here is the lack of strafe control - but that could be assigned to the right stick instead of free look.

2Controller.png
 
Last edited:

Fallen Kell

Diamond Member
Oct 9, 1999
6,009
417
126
Yeah, I would think gimbal weapons aim control or strafe would be much more important than free look for default on the right stick. It might simple be that there are not many people truly using it and the defaults are simply bad as a result.
 

SunnyD

Belgian Waffler
Jan 2, 2001
32,674
145
106
www.neftastic.com
As Worthington noted, this exact issue is being addressed in the new flight model being released in 2.0.

Here is a rather lengthy discourse on the changes:
https://robertsspaceindustries.com/comm-link/transmission/15031-Star-Citizen-Alpha-20

The switch to third-order motion control will be the biggest change and will smooth out the mechanical "jerky" motion that you're referring to that is a result of the current 2nd order motion, along with a modification of current jerk values and maneuvering thruster output.

Expect a very different flight experience in 2.0.

After reading/watching all of that, they still didn't mention anything much about the exact ship physics being incorporated into the flight models other than thrust errors. It look to me like they're just refining their current algorithms and adding a refinement layer on them to smooth things out, but not adding actual physics to them... yet. They didn't specify anything about things like actual thrust/mass values being worked into those algorithms (or I missed it). Those things are kind of important.

Yes, the acceleration curves and whatnot will be much nicer and realistic, but it still won't mesh well if, say, a Connie can corner just as fast as a Mustang. And more importantly it will fundamentally change combat.
 

Sabrewings

Golden Member
Jun 27, 2015
1,942
35
51
After reading/watching all of that, they still didn't mention anything much about the exact ship physics being incorporated into the flight models other than thrust errors. It look to me like they're just refining their current algorithms and adding a refinement layer on them to smooth things out, but not adding actual physics to them... yet. They didn't specify anything about things like actual thrust/mass values being worked into those algorithms (or I missed it). Those things are kind of important.

Yes, the acceleration curves and whatnot will be much nicer and realistic, but it still won't mesh well if, say, a Connie can corner just as fast as a Mustang. And more importantly it will fundamentally change combat.

I understand your sentiment and I share the annoyance with how it makes takeoffs appear, but I don't think you are on the right track as to what the problem has been up to this point.

The ships already have and have always had mass incorporated into the flight model. Every ship has a specific mass that is acted upon by thrusts at various points around its center of mass. The issue is that the thrusters are too powerful and ramped up instantly resulting in jerky and unrealistic motion. Take an Avenger, for example. It is capable of blacking out its pilot, so that means the four thrusters on the bottom of the craft can generate in excess of 5 or 6 G's depending on the individual's blackout tolerance. Now, whenever that same craft has taken off from a landing pad, those thrusters instantly ramped up to full and it took off at that same 5 or 6 G's (50-60m/s/s) up to the ship's predetermined top speed (something that is changing, thank god). Anyone flying around in AC can tell after the first few turns that ships have mass due to the amount of drift. It takes time for the maneuvering thrusters to cancel out the undesired vectors in the realms of 200+m/s. The problem is that when we're near platforms we have that same amount of maneuvering thrust available instantly when we should only realistically be moving 30-40m/s max in a docking situation (much less when coming in for contact, preferably less than 10m/s).

Even the Connie, for example, has pretty powerful thrusters. It has to be able to VTOL, so for reasonable overhead for vertical acceleration and maneuvering, at least 2 G's is required but it can probably do closer to 3-4 G's on thrusters vertically. It's not going to black out a pilot, but it's quite a spritely amount of thrust. Also keep in mind it has that same amount of thrust available on the dorsal side, so that's why it jerks to a stop when taking off vertically as well.

The fix is that when docking we'll have the precision maneuvering mode which cuts down on this high amount of thrust considerably. Instead of an instant on/off digital discrete of several G's, we'll have a much more subdued access to the thrusters. Also, the addition of third degree movement will mean that the first moments of acceleration are reduced compared to power output after the thruster has had time to ramp up.

Thus far what we've been seeing is similar to everyone on the road having to choose between accelerator the floor, no input, or brake pedal to the floor. Things would look a lot different if that were the case, so expect a similar degree of change when we finally have precision maneuvering and third degree motion.
 
Last edited:

SunnyD

Belgian Waffler
Jan 2, 2001
32,674
145
106
www.neftastic.com
I understand your sentiment and I share the annoyance with how it makes takeoffs appear, but I don't think you are on the right track as to what the problem has been up to this point.

The ships already have and have always had mass incorporated into the flight model. Every ship has a specific mass that is acted upon by thrusts at various points around its center of mass. The issue is that the thrusters are too powerful and ramped up instantly resulting in jerky and unrealistic motion. Take an Avenger, for example. It is capable of blacking out its pilot, so that means the four thrusters on the bottom of the craft can generate in excess of 5 or 6 G's depending on the individual's blackout tolerance. Now, whenever that same craft has taken off from a landing pad, those thrusters instantly ramped up to full and it took off at that same 5 or 6 G's (50-60m/s/s) up to the ship's predetermined top speed (something that is changing, thank god). Anyone flying around in AC can tell after the first few turns that ships have mass due to the amount of drift. It takes time for the maneuvering thrusters to cancel out the undesired vectors in the realms of 200+m/s. The problem is that when we're near platforms we have that same amount of maneuvering thrust available instantly when we should only realistically be moving 30-40m/s max in a docking situation (much less when coming in for contact, preferably less than 10m/s).

Even the Connie, for example, has pretty powerful thrusters. It has to be able to VTOL, so for reasonable overhead for vertical acceleration and maneuvering, at least 2 G's is required but it can probably do closer to 3-4 G's on thrusters vertically. It's not going to black out a pilot, but it's quite a spritely amount of thrust. Also keep in mind it has that same amount of thrust available on the dorsal side, so that's why it jerks to a stop when taking off vertically as well.

The fix is that when docking we'll have the precision maneuvering mode which cuts down on this high amount of thrust considerably. Instead of an instant on/off digital discrete of several G's, we'll have a much more subdued access to the thrusters. Also, the addition of third degree movement will mean that the first moments of acceleration are reduced compared to power output after the thruster has had time to ramp up.

Thus far what we've been seeing is similar to everyone on the road having to choose between accelerator the floor, no input, or brake pedal to the floor. Things would look a lot different if that were the case, so expect a similar degree of change when we finally have precision maneuvering and third degree motion.

I understand everything you're saying here, but what I'm saying is that the mass/thrust models currently utilized don't appear correct. I don't know if that's just an artifact of the current movement system in general or whether they're simply not being used - it almost feels like they're just using a standard constant across the board right now regardless of the ship. It's not that the mechanics are all or nothing in the current model, it's that in addition to the current formulas there just feels like there's no mass/thrust (or rather a constant mass/thrust) to each ship giving every single ship the exact same performance metric.
 
Last edited:

rivethead

Platinum Member
Jan 16, 2005
2,635
106
106
SunnyD - you are correct. I don't think they've factor in mass changes yet into the calculations.

I did a test several months ago: raced my Aurora LN through OV fully equipped. The stripped off all the weapons and missiles (making it lighter, in theory) and raced again.

There was no difference in flight. The ship felt exactly the same. I realize weapons and missiles aren't a lot of mass, but it's still enough (in my opinion) to make a tangible different in drift.

I don't think they've built that in yet.
 

rivethead

Platinum Member
Jan 16, 2005
2,635
106
106
So, when I get tired of Fallout 4 in a year, will this be ready for me?

It depends on your definition of "ready".......


Significant new content is going to be available to backers within the next few weeks (most likely).

Will it be fun? Hell yes.

Will it be unfinished and buggy? Hell yes.
 

Fallen Kell

Diamond Member
Oct 9, 1999
6,009
417
126
Well, the one thumbstick arrived the other day, and last night I got my new multimeter so I could get some readings off it. I don't think I will be using it, or if I do, I need to make a HUGE center deadzone. Basically in one of the directions, if I move it all the way to the limit and let it zero, it gets pretty close (according to the multimeter), but if I then move it all the way to the other limit, it is really off. It probably won't matter as much for my intended use for strafe, so I still might use this one. I've been doing the math and if I put this on a 10bit counter (range would then be 0 - 1023), I believe the one direction if I moved it to the limit would give me a reading of about 2-3 and the return to center of about 450, but the other direction the limit would be about 1020 and the center that is moves back to would be around 800!

I have a PS2/3 style thumbstick coming which supposedly much, much better, but it is also a lot larger (I really don't know if I can get it to fit. I think I will be able to, but I really don't know for a fact. The nice thing about the PSP 1000 stick is it's size is so compact. I'm not looking forward to soldering to it's contacts (the "board" that it uses is as thin as business card stock, which means it will be VERY susceptible to heat damage with the contacts/traces very likely to melt off).
 

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
I have a PS2/3 style thumbstick coming which supposedly much, much better, but it is also a lot larger (I really don't know if I can get it to fit. I think I will be able to, but I really don't know for a fact. The nice thing about the PSP 1000 stick is it's size is so compact. I'm not looking forward to soldering to it's contacts (the "board" that it uses is as thin as business card stock, which means it will be VERY susceptible to heat damage with the contacts/traces very likely to melt off).


That sucks. More waiting for more parts...hopefully the new one works out better.
 

Fallen Kell

Diamond Member
Oct 9, 1999
6,009
417
126
New one showed up today, but it is like I thought, big. I am not sure it will fit either. It definitely doesn't fit on the top front like I was hoping that it "might"...

Basically I am screwed by how they did the Z-axis twist censor. Here is an idea of what it looks like internally:
atWmAfP.jpg


The grey piece that was hand added is where the main shaft is located. The green piece with the red dot on top of it is the Z-axis twist sensor. I can't tell if it is a hall-effect sensor or some type of pot (I am leaning towards hall-effect due to its small size). The shaft doesn't actually twist, just the grip piece twists around the shaft (there is a piece that I did not draw in there which is about 1/2 way up the shaft that is sandwitched between two of the pieces of support plastic on the inside of the grip to keep it from moving up or down. That piece also has a metal spring wire which sticks out the end (it is sort of "V" shaped with the two ends sticking out some and coming in contact with more plastic support on the inside of the grip such that when the handle twists, it places a load on the spring, compressing the ends towards each other). When the grip twists, it twists the green part of the sensor with the base of the sensor staying attached to the rod.

Anyway, there is simply no room for anything on the front of the joystick as a result of the internal structure due to that sensor setup. On the top, I drew in the internal PCB board that exists to give you an idea of what else is in the way (you are looking at the board edge on in that view).

So that really leaves 2 locations:
I could potentially remove the hat switch and place the thumbstick in that location.

Or I could place it on the backside of the grip, below the trigger.

That hat switch location is probably the most ideal, but then I obviously lose the hat. In my situation, I don't think that is too bad of a loss as I also have VoiceAttack to make up for the lost actions that could be tied to the hat.

On the back of the grip behind the trigger is more out of the way, but much harder to operate as the fingers that can access the stick in that location have very limited vertical movement, making it very difficult to have any kind of precise control of movement in that axis.
 
Last edited:

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
This week's AtV had some update art for the Freelancer interior pass. Looking pretty awesome.

I have a feeling I will be spending a lot of time in my DUR. It supposedly is very efficient to operate, has an explorer scanner package, and a nice big cargo area. Seems like it will probably be able to fill a few different roles really well and make some cash.

Kinda sad it won't be flyable in 2.0, but ah well.

7RDwwjD.jpg


3WjhleV.jpg
 

Sabrewings

Golden Member
Jun 27, 2015
1,942
35
51
Yeah, there was that... then there was also the 20 minute Connie rework video!!!

feature-pun-dog-370x297.jpg


Things I liked: Awesome exterior details, awesome interior details, cockpit bulkhead, rear facing turret entry, Tali style missile magazines (can you say missile spam?), and the new Merlin dock.

Things I didn't like: Lower turret still not inverted (willing to try it with the pivot seat), airlock hatches no longer have windows
 

Stringjam

Golden Member
Jun 30, 2011
1,871
33
91
Yeah, there was that... then there was also the 20 minute Connie rework video!!!


Things I liked: Awesome exterior details, awesome interior details, cockpit bulkhead, rear facing turret entry, Tali style missile magazines (can you say missile spam?), and the new Merlin dock.

Things I didn't like: Lower turret still not inverted (willing to try it with the pivot seat), airlock hatches no longer have windows


Indeed! I can't wait to see the Connie in 2.0 - even though I have a feeling it will only be the Andromeda and not the Aquila yet. They did a fantastic job with both the interior and exterior rework.

And yeah.....MISSILES.