Smooth movement of aircraft controls

As requested, here's the revised version of my poorly worded post.

Summary: In-game movement of flight controls is not smooth which looks bad. All of the below is just my speculation on possible cause.

It looks like really bad right now, as if controls animation steps are set to a very low value, like 50 for the entire range. In legacy cockpits it was not so much out of place but now... I haven't posted on it as a bug report because I don't think anyone will consider it a bug. Here's a video as an example.

https://www.youtube.com/watch?v=yrh3-OjfT7o

It should be evident so if you don't see it I won't be able to clarify.

I'm posting on the outcome, unable to tell what's the reason behind it. I think there are two potential culprits:

Case 1: in-game 3D model

As mentioned, the animation steps value may be too low. Note: I have no idea on 3D animation.

Case 2: game controller input

For example it could be that the limited steps (resolution) is on the side of game controllers and DCS takes the input to drive the animation 1:1, without neither interpolation, nor smoothing. It shouldn't.

EDIT:

Case 2 variant 2: DCS allows impossible motion rates of pilot flight controls.

In-game flight controls movement should be realistic regardless of input device.

My stick moves perfectly smoothly, the game detects it as moving perfectly smoothly and the cockpit stick moves perfectly smoothly. You've got a bug!

Or your joystick is choppy.

+1 on the smooth controls. I don't see any stepping or jumping types of effects on the control animations.

That last vid looks a lot like spiking pots/crappy hardware to me.

That last vid looks a lot like spiking pots/crappy hardware to me.

Agreed. That is not the sim, IMHO. I have seen similar, and worse, if my Warthog loses calibration (for reasons unknown, it used to do so).

I would consider the input resolution of DCS an acceptable compromise.

I'm going to search for some vids made by warthogg users. Please post a link if you know such guys on YT.

In the meantime please note that even with it being controllers' fault IMO DCS still shouldn't take the input 1:1. See Case 2.

I would consider the input resolution of DCS an acceptable compromise.

You mean DCS has its own resolution irrespective of directinput parameters?

What is your joystick?

What is your joystick?

I don't have DCS installed right now. But it doesn't matter. I've been observing the issue for years now. Applies to vast majority of videos. This I not a matter of some isolated cases. The problem is that you guys see a video, the stick jumps and you go 'pot spiking!' disregarding the clunky moment of controls when the stick is not jumpy.

BTW, most of joystick controllers I've canibalized had a resolution of 1024, sometimes 512 on some axes. I seriously doubt DCS utilizes this many steps for flight controls animation.

EDIT:

I've just came up with an idea on how to isolate the cause: low input resolution vs low animation resolution. Some A/C have a trim system which moves the flight stick. So if that animation is smooth, the problem lies in: how DCS handles game controller input to drive flight controls animation.

BTW, most of joystick controllers I've canibalized had a resolution of 1024, sometimes 512 on some axes.

I am not sure what point you are trying to make here, but neither a resolution of 1024 nor 512 is very good.

The Warthog for example, has a 16-bit ADU which results in a resolution of 65536.

Back to topic though, I personally have not seen any of that stick-jerking in DCS. My stick here moves smoothly and is far of from being like in the videos you gave as an example.

However, the stick and especially throttle movement is indeed delayed by a tiny bit. This could have various reasons, but I assume that it has to do with either of both:

- Simulating a slow movement because of the control forces the pilot had to overcome.

- Using filters on the axis to prevent any control spikes. When filtering movement axis inputs a delay will be created that causes the simulated axis lag behind the real one.

That's like silk in comparison to IL2 COD.

No jumping about here. Its perfectly smoothly animated. Even if I let the stick oscilate (FFB + extension and no damping. deflect stick a little and release and it will oscilate) it jumps about, but jumps about smoothly, or at least as smoothly as my framerate allows

Edit: bear in mind that most videos are recorded at 30fps, which is slow enough for some people to detect as a little bit of skipping/freeze framing.

I don't have DCS installed right now. But it doesn't matter. I've been observing the issue for years now. Applies to vast majority of videos. This I not a matter of some isolated cases. The problem is that you guys see a video, the stick jumps and you go 'pot spiking!' disregarding the clunky moment of controls when the stick is not jumpy.

 

BTW, most of joystick controllers I've canibalized had a resolution of 1024, sometimes 512 on some axes. I seriously doubt DCS utilizes this many steps for flight controls animation.

 

~

in a way, yes... it does matter

this end, my warthog/ animation is fine...

In the meantime please note that even with it being controllers' fault IMO DCS still shouldn't take the input 1:1. See Case 2.

It is a design choice that the stick in the virtual pit and the (commanded) position are identical. Interpolation would be mostly useless because it doesn't improve the steady state resolution. Smoothing should, at the very most, be optional but IMHO it should be the stick manufacturer's duty to cover for crappy hardware if he choses to use such.

However all of this is sort of moot, because stick movement doesn't translate to control surface movement. Those are saturated to the max rad/sec the actuators can handle. In case of direct linkage between control surface and stick, the stick becomes saturated as well.

I am not sure what point you are trying to make here, but neither a resolution of 1024 nor 512 is very good.

The Warthog for example, has a 16-bit ADU which results in a resolution of 65536.

 

Back to topic though, I personally have not seen any of that stick-jerking in DCS. My stick here moves smoothly and is far of from being like in the videos you gave as an example.

 

However, the stick and especially throttle movement is indeed delayed by a tiny bit. This could have various reasons, but I assume that it has to do with either of both:

- Simulating a slow movement because of the control forces the pilot had to overcome.

- Using filters on the axis to prevent any control spikes. When filtering movement axis inputs a delay will be created that causes the simulated axis lag behind the real one.

My point then was:

even the cheapest joysticks provide more steps than the animation displayed in DCS.

I'll have to investigate this one further though...

Re text in bold:

It's weird that you wrote that since it's pretty evident that neither has been applied in the videos I've linked (again, these are just examples of a common behavior). You can clearly see movement of the flight control stick that is literally impossible in reality, even if the stick wasn't attached to enything except the gimbal.

Edit: bear in mind that most videos are recorded at 30fps, which is slow enough for some people to detect as a little bit of skipping/freeze framing.

30 FPS is enough for the issue to manifest in a clear manner.

in a way, yes... it does matter

 

 

this end, my warthog/ animation is fine...

Could you post a link to your YT profile?

Smoothing should, at the very most be optional,

I agree.

but IMHO it should be the stick manufacturer's duty to cover for crappy hardware if he choses to use such.

Couldn't disagree more. Vast majority of DCS players use controllers with potentiometers. I'd even dare to say that the issue doesn't exist only to those using Thrustmaster T.16000, Warthog or a DIY set of 12-bit controls.

However all of this is sort of moot, because stick movement doesn't translate to control surface movement. Those are saturated to the max rad/sec the actuators can handle. In case of direct linkage between control surface and stick, the stick becomes saturated as well.

Your argument is invalid. I specifically posted on the issue with how it looks in the cockpit. Nothing about control surfaces or how it influences aircraft handling.

Those are saturated to the max rad/sec the actuators can handle. In case of direct linkage between control surface and stick, the stick becomes saturated as well.

This at least DCS does well, unlike X-Plane for example. But it's OT.

Bucic i dont think Videos of other people are a good reference for yourselfs thoughts because you cant know which equipment is used or how the input to the controler looks like.

Bucic i dont think Videos of other people are a good reference for yourselfs thoughts because you cant know which equipment is used or how the input to the controler looks like.

My point is, the issue should be solved taking into account potentiometer-based controllers. Which happen to constitute the vast majority of controllers used. That's fact 1. Or to put it in other words, DCS should not display impossible flight control movements. Smoothing is not OpenCL support. That's fact 2. I'm fine with 'it won't be changed' but let's keep it eagle.ru here :)

The issue is visible in pretty much any video

so the ball is rather on the side of those who claim they've seen smoothly animated controls in DCS :) I'll try to find some myself.

I don't have DCS installed right now. But it doesn't matter. I've been observing the issue for years now. Applies to vast majority of videos.

You should be aware alot of videoes are from users only using keyboard to fly with.:music_whistling:

You should be aware alot of videoes are from users only using keyboard to fly with.:music_whistling:

Oh c'mon, you're desperately grasping the straws here. You won't find many keyboard "flyers" even amongst WarThunder crowd, let alone any proper sim pilots.

But back to the topic, the animation issue seems to be mod-dependent as well. It's smooth in my Mustang and Dora, but not so much in MiG-21. On the other hand, input curve setting doesn't seem to have any effect on it. TH Warthog here.

You mean DCS has its own resolution irrespective of directinput parameters?

Not doing this could really screw up some of the models, so yes.

Oh c'mon, you're desperately grasping the straws here.

Yeah, same with 30 FPS etc.

I know DCS well. It's vices and virtues... I've played FC and DCS for hundreds of hours, watched as many videos and I closely follow its development. So please rather focus on what I wrote and address the points.

Not doing this could really screw up some of the models, so yes.

I see.

This does not dismiss the issue however, in any way I look at it.

 

Couldn't disagree more. Vast majority of DCS players use controllers with potentiometers. I'd even dare to say that the issue doesn't exist only to those using Thrustmaster T.16000, Warthog or a DIY set of 12-bit controls.

 

looks like it does make a difference :huh:

Who is affected by this issue and who isn't? Any clarity?

Who is affected by this issue and who isn't? Any clarity?

My take is, everyone not using high-resolution hall-effect controls (e.g. Thrustmaster HOTAS Warthog) are affected, i.e. can see the jerky in-game flight controls.

I've finally found a video showing Su-27 practicaly without any jerking on flight controls*

https://www.youtube.com/watch?feature=player_detailpage&v=u_8CBlFxPgw#t=220

Controls: HOTAS WArthog with extended arm.

Also, take a look at this (War Thunder)

Perfectly smooth flight controls movement, even during +/- 10 mm corrections.

*too bad it introduces the unwanted factor of SFM vs PFM (the change in flight model)

For later:

Confirmed: Logitech Extreme 3D Pro