Motion Support for IL-2

[hexpod]

3 hours ago, frenchfly said:

I'll test this when I get home from work and let you know. Hexpod seems to think it's off but I'll isolate those forces individually later to be sure.

 

 

Yes, open the tuning center and once in the air make your aircraft oscillate with the rudder. Observe which translational values are moving the most, surge or sway ?

 

Than, fire some bullets with the cannons and observe which translational values move the most, surge or sway ?

 

4 hours ago, value1 said:

I have matched

• Surge_Output with Acc_y

• Sway_Output with Acc_x

• Heave_Output with Acc_z

Not sure, if this is right – it looked somewhat right from the data I received from the UDP stream.

Should I switch Acc_x and Acc_y?

 

From what I see, yes, that needs to be inverted if you search for coherence in the tunning center.

 

It has to be confirmed.

Edited April 26, 2019 by hexpod

[hexpod]

9 hours ago, frenchfly said:

Can you post a screen shot of your simtools axis settings so I can see how they compare to mine.

 

Thanks for the info

 

F

 

If for your 2dof platform, you are using the first and second DOF column for the angular velocities, I believe, you can mix it with linear accelerations by putting them in the column 3 and 4. (Usually surge on pitch and sway on roll) 

 

Most probably you will get a better result doing so as if you would use the angular positions. 

 

You can try it if you like.

Edited April 26, 2019 by hexpod

[frenchfly]

Hexpod I think your right, I turned off all other forces but surge and it was way more active with the rudder than when I had on just sway.

 

Sway only had any real movement when crashing, that's with it set at 2 in tuning center.

 

 

[hexpod]

4 hours ago, value1 said:

I have matched

• Surge_Output with Acc_y

• Sway_Output with Acc_x

• Heave_Output with Acc_z

Not sure, if this is right – it looked somewhat right from the data I received from the UDP stream.

Should I switch Acc_x and Acc_y?

 

For the Extra1_Output you took Spin_x or Spin_y?

 

Those are correct.

Edited April 26, 2019 by hexpod

[TWHYata_PL]

I have matched this axis in my 2 dof ( front 100mm acu)

 

Edited April 26, 2019 by TWHYata_PL

[hexpod]

20 hours ago, TWHYata_PL said:

I have matched this axis in my 2 dof ( front 100mm acu)

 

 

You may need to try it differently if you want to mix the velocities with accelerations.

 

1.)

you could invert one axis in your DOF1 column because I think it’s actually sway.

2.)

you could replace roll with lateral angular velocity in your DOF2 (Extra2)

3.)

you could add sway (actually longitudinal accel. - plugin v.2.0.0.5) in DOF4 

Edited April 27, 2019 by hexpod

[hexpod]

Quick profile with angular velocities and linear accelerations mixed. No positions in use.

 

surge and sway inverted as I smell a small mistake there. ? - best

 

 

Edited April 27, 2019 by hexpod

[value1]

19 hours ago, hexpod said:

 

For the Extra1_Output you took Spin_x or Spin_y?

 

Those are correct.

                Extra1_Output = Spin_y
                Extra2_Output = Spin_x
                Extra3_Output = Spin_z

[hexpod]

Here the matrix in use in the vid above:

 

[hexpod]

1 hour ago, value1 said:

                Extra1_Output = Spin_y
                Extra2_Output = Spin_x
                Extra3_Output = Spin_z

 

If roll = x around longitudinal axis and pitch = y around lateral axis it’s logical surge is along x and sway along y.

 

don’t you think so?

 

The developer gave this as reference in the documentation.

 

 

Edited April 27, 2019 by hexpod

[hexpod]

On 4/26/2019 at 7:09 PM, value1 said:

I have matched

• Surge_Output with Acc_y

• Sway_Output with Acc_x

• Heave_Output with Acc_z

Not sure, if this is right – it looked somewhat right from the data I received from the UDP stream.

Should I switch Acc_x and Acc_y?

 

Yes please. 

Even if many are already happy with their false assignments I would advocate you change it as it will be difficult for beginners to get it right. ?

 

Thanks - Best

Edited April 27, 2019 by hexpod

[value1]

Thank you for your feedback!

I have swapped surge and sway. Please give it a try and let me know if this is correct now.

Thanks!

 

Il_2BoS_GamePlugin_V2.0.0.6.zip

1 hour ago, hexpod said:

 

If roll = x around longitudinal axis and pitch = y around lateral axis it’s logical surge is along x and sway along y.

 

don’t you think so?

 

The developer gave this as reference in the documentation.

 

 

True I didn't pay attention enough ? 

Edited April 27, 2019 by value1

[frenchfly]

Your right hexpod:)

Edited April 27, 2019 by frenchfly

[hexpod]

2 hours ago, value1 said:

Thank you for your feedback!

I have swapped surge and sway. Please give it a try and let me know if this is correct now.

Thanks!

 

Il_2BoS_GamePlugin_V2.0.0.6.zip 38.12 kB · 1 download

True I didn't pay attention enough ? 

 

That’s it. Works coherently. Thank you.

[rowdyb00t]

@hexpod what is the brand of the motion sim that you have there?

[hexpod]

5 hours ago, rowdyb00t said:

@hexpod what is the brand of the motion sim that you have there?

 

DiY

[hexpod]

On 3/30/2019 at 6:11 PM, Jason_Williams said:

Thank you Value1 for the plugin and Gforcefactory for your fast support as well! 

 

And special thanks to Sergey our lead programmer who made the support quickly once we committed to it. 

 

Feedback from either implementation is welcomed. 

 

Now I need to buy a motion rig! 

 

We have the best community and team in simming. 

 

Jason

 

@Jason_Williams Thank you again to give us access to the main motion telemetry.

 

Above you can see a small video with a 6-dof platform with mixed angular velocites and linear accelerations. It’s a fast first profile but already with a impressive result.

 

Now, as everything concerning the main telemetry got sorted out, I have a small question:

 

What is exactly the visual camera shaking effect which disappears in high RPM’s?

If it is an engine or propeller vibration exaggeration, I think we could take an big advantage out of it. 

 

We could have 3 free slots for additional telemetry output for a simtools’s experimental plugin.

 

Could the lead programmer give us a small expertise, what kind of effects (motor, propeller stall, turbulences) would make sense to output additionally for our platforms?

 

Thank you very much.

 

cheers

 

 

Edited April 28, 2019 by hexpod

[Alonzo]

Amazing video @hexpod — what is the approximate cost of your motion rig and do you have an instructions thread for it somewhere please?

[hexpod]

41 minutes ago, Alonzo said:

Amazing video @hexpod — what is the approximate cost of your motion rig and do you have an instructions thread for it somewhere please?

 

You should find all info you need about motion from 2 up to 6-dof in various forums at http://www.xsimulator.net

Edited May 4, 2019 by hexpod

[spitfirejoe]

Really nice.....a hexapod is of course much much better for flight simulation than my motion rig.

 

Good Job.

[hso]

I've made a plug-in for the CKAS 2/3/6-DOF systems. It all works fine as documented.

 

I want to thank the developers who, perhaps for the first time amongst the games I know, put out precisely what is needed for proper motion. That is

• Body inertial accelerations (decoupled from gravity)
• Gravity acceleration [direction], determined by pitch and roll angles
• Body angular velocities.

Also, the axis frame, although not a very orthodox one for aviation, is at least consistent in terms of sign conventions and is right-handed.

 

That said, this is a really minimalistic set. It would benefit to have more. One of the main reasons is that higher frequency effects (like vibrations and buffet) require different treatment, and may even go to different devices (as many people here already mentioned), such as shakers or fans.

 

If I may offer suggestions, I'd like to add:

•  On-the-ground flag. IL-2, being operated mostly on dirt fields, produces nauseatingly big movement when running or taxiing. This may be 'real' in some sense, but on limited-motion systems (which most people have) it clips often and feels unnatural. Yet if tuned down, it affects in-air effects negatively. A way to combat this is to treat ground movement separately from flying (on the plug-in side), but to do it we need the ground flag.
• Engine(s) RPM. This allows to produce engine vibration.
• Flaps angle (or setting). For higher-frequency buffet.
• Landing gear extension (or just a flag whether extended). Also for aerodynamic buffet, as well as for light bumps when fully extended/retracted (though the latter is less applicable to WWII aircraft).
• Other similar equipment deployment (air/dive brakes etc.)
• Stall condition. Some aircraft have relatively high-pitched buffet when stalled, which is better induced separately than filtered from normal accelerations.
• Gun(s) activation, or better the magnitude of recoil. IL-2 does produce it 'naturally' in the X acceleration, but again, it may be too high-frequency to treat it adequately as 'normal' motion.
• Many also ask for airspeed, although motion per se doesn't require that. Yet airspeed would be helpful to moderate other effects (such as buffets mentioned above).
• Current vehicle name/identification (i.e. the model). Different airplanes (let alone tanks ) may prefer different style of motion, and knowing the model is a pre-requisite for that.

It would be nice to keep backward compatibility, so that game updates wouldn't break motion for everyone. To do that, it makes sense only to append the new data to the existing structure. The last item, being a variable string, requires some attention to be future-proof, but that's up to the developers.

 

But all in all, thanks for a good job!

 

Edited May 22, 2019 by hso
one more point

[Gforcefactory]

Hi All,

 

Just came back from Flight Sim Expo 2019, Orlando, Florida. Thanks to Jason for the wonderful cooperation!

 

Here an impression of the 1C Game Studio booth.

 

Best,

 

Team Gforcefactory

 

.

 

[Alonzo]

10 hours ago, Gforcefactory said:

Just came back from Flight Sim Expo 2019, Orlando, Florida. Thanks to Jason for the wonderful cooperation!

 

Here an impression of the 1C Game Studio booth.

 

Great video! How does the motion-cancellation work on the Vive? Do you have a motion cancellation solution for other headsets, such as Rift S ?

[Gforcefactory]

Hi Alonzo,

 

Thanks for your kind feedback. Motion cancellation in the video works with OpenVR-InputEmulator. We have built solutions for Oculus earlier, but these were difficult to maintain due to Oculus software updates (a common problem). To further support motion cancellation our system now provides a UDP API by which software motion cancellation can be built by third parties (so, no tracker required). We have already an implementation for Vive in beta. One of our customers has used the API for the combination Oculus/Unreal. Another customer is currently using the API for a solution for Vive without tracker. If you happen to know other solutions, we are all ears!

 

Best regards,

 

Team Gforcefactory

[hexpod]

57 minutes ago, Gforcefactory said:

Hi Alonzo,

 

Thanks for your kind feedback. Motion cancellation in the video works with OpenVR-InputEmulator. We have built solutions for Oculus earlier, but these were difficult to maintain due to Oculus software updates (a common problem). To further support motion cancellation our system now provides a UDP API by which software motion cancellation can be built by third parties (so, no tracker required). We have already an implementation for Vive in beta. One of our customers has used the API for the combination Oculus/Unreal. Another customer is currently using the API for a solution for Vive without tracker. If you happen to know other solutions, we are all ears!

 

Best regards,

 

Team Gforcefactory

 

Can you please tell more about your UDP API ?

 

Are you sending the platform commanded pose through UDP and subtract it inside VR software?

 

What’s the idea behind using VIve without tracker ? 

 

It’s hard to imagine how the commanded pose can match considering a mismatch due to the platform inertia ?

 

i would be thankful if you could explain the concept in more detailed way.

 

Thanks

[Gforcefactory]

Hello Hexpod,

 

Congratulations with your system. Looks great!

 

The UDP-API can provide the platform positions at 200Hz. This information can be used by the VR headset driver to cancel out this position. There is no mismatch due to platform inertia, because we can read out the exact positions of the platform; no magic here ;).

 

Best regards,

 

Team Gforcefactory

[hexpod]

3 hours ago, Gforcefactory said:

Hello Hexpod,

 

Congratulations with your system. Looks great!

 

The UDP-API can provide the platform positions at 200Hz. This information can be used by the VR headset driver to cancel out this position. There is no mismatch due to platform inertia, because we can read out the exact positions of the platform; no magic here ;).

 

Best regards,

 

Team Gforcefactory

 

Thank you for the info.

 

If I understand the mechanism correctly, the most of 6-dof kinematics are based on inverse one which translates the pose of the platform into the actuators movement. Depending of the platform velocity, I can imagine the delay could be noticeable.

 

It can be you use “forward kinematics” to be perfectly in sync.

 

How many headset manufacturers are compatible with your UDP API ?

 

If there is any UDP array “standard” to be sent so the degrees and millimeters can be subtracted by the cancellation software ?

 

I would be thankful for additional info so we could follow the same standard.

 

best!

Edited June 19, 2019 by hexpod

[Gforcefactory]

Hello Hexpod,

 

Pleased to hear you know the theoretical underpinnings. If you are interested, the documentation of the UDP-API can be downloaded from our website, including the syntax. For now, we have only implemented a solution for HTC Vive.

 

If you have further questions, you can contact us directly of course.

 

Best regards,

 

Team Gforcefactory

[Swine08]

I think the Next Level Racing V3 supports Great Battles now.  Is that correct??  

 

I contacted Next Level Racing and this was they said:

 

“Thank you for contacting us.

The Motion Platform V3 is compatible with the below versions of IL-2. I believe all of these games are part of the Great Battles series, is this correct?

IL-2 Cliffs of Dover

IL-2 Battle of Bodenplatte

IL-2 Battle of Kuban

IL-2 Battle of Moscow

IL-2 Battle of Stalingrad

IL-2 Cliffs of Dover Blitz

If you have any further queries we are happy to assist you. 

Kind Regards,”

[AndyJWest]

Someone should tell them that Cliffs of Dover (inc Blitz) isn't part of the Great Battles series.

[Swine08]

I did.  They already supported COD since the beginning.  I advised Great Battles and COD were different. 

I explained it in detail.

[Taurus]

Thank you so much Jason and team for your efforts in making motion output possible! IMHO this has just cemented the IL-2 titles as the unsurpassed leader in WW1 and WW2 air combat SIMULATION software, ever!

 

I have been in the workshop for months designing my g-Seat-II and have just emerged to start testing the software interfacing to find this absolutely stunning news. Value1 over at XSimulator has a Simtools interface, and knowing Value1's efforts, I cannot wait to try it with your software.

 

These are really great days! And thanks again.

 

EDIT 18/08/2019: Confirmed that Value1's plugin for XSimulator Simtools does work well with the Steam version of BOS for the Thanos AMC-AASD15A Servo Controller with SFX-100 actuators. Thanks to Value1's help yet again the 1CGS version of BOM for the Thanos AMC-AASD15A Servo Controller with SFX-100 actuators is now working. The RoboClaw and g-Seat-II to be confirmed. (BOK, Flying Circus and Tank Crew components have motion output as well! Tank Crew will be very interesting!)

 

EDIT 04/12/2019: Seat working and flying with VR. It doesn't get any better than this! Only thing missing now is the smell. The pilot g effects are brilliant and very effective. Tank Crew didn't disappoint either. Feeling the tank lurching and jolting underneath, the recoil of the main gun, is a huge jump in the immersion level. VERY VERY HAPPY!

 

EDIT 26/01/2020: Add Video of seat in action. 

 

Youtube video

 

 

 

 

 

Edited January 26, 2020 by Taurus
add video link 26/01/2020

[FlyingNutcase]

On 4/28/2019 at 5:03 AM, hexpod said:

Quick profile with angular velocities and linear accelerations mixed. No positions in use.

 

surge and sway inverted as I smell a small mistake there. ? - best

 

 

 

Damn Diggity Dog, that is some impressive rig! And cleary stronger than it looks. Waiting for this sort of stuff go in to (relatively) mass production...

Edited September 22, 2019 by FlyingNutcase

[SCG_motoadve]

Does anyone has Motion Platform V3 working with IL2?

[SkydiversElliot-Priski]

Big props to 1 c game studios for making this happen, this is huge! I bought my DOF P3 motion chair pretty much exclusively for this game a few months ago. Now finally SRS Studio has released an update to support it!

 

I have been playing around for quite a while with a Valve Index headset and warthog hotas and I have a question and also tips.

 

First of all tips. I thought at first that the forces felt when turning were a little off. I thought a bit about how to solve this, and read a like mind who had the same idea on these forums. French fly advised doing to complete opposite with the chair as the plane.

 

Aka roll/bank to the left, the chair should move to the right. This really works, and for those using SRS studio just set sway to 20 and put roll to 0. Maybe this only works in VR, but the effect is fantastic. You bank left, your body feels the force as your legs and head etc move right (caused in reality by gravity as your chair is banking right). This ups the realism by 10x

 

I have included an image of my saved settings in SRS as I find them an order of magnitude better than the default settings.

 

 

Now for my question... for those kind enough to answer.

 

When I do any action in the game (bank etc) it takes about 0.8 seconds roughly for my chair to move. This does not affect other games with my software and chair. This lowers immersion. Is this something to do with IL2, my computer or SRS studio? Does anyone else have this issue in SRS or other software?

 

Thank you again to the devs, for all your amazing work with this game... really loving the series and love the extra mile you guys are going for us 'hardcore' folk ?

Edited October 18, 2019 by SkydiversElliot-Priski

[SR-F_Winger]

oh i really want such a motion seat.
I am sooo jealous. Starting to save money. The DOF Reality products really do look very nice.
I hooe you get your problem sorted becaus this amount of latency must be a PITA for immersion.

[SkydiversElliot-Priski]

On 10/24/2019 at 3:42 PM, Winger said:

oh i really want such a motion seat.
I am sooo jealous. Starting to save money. The DOF Reality products really do look very nice.
I hooe you get your problem sorted becaus this amount of latency must be a PITA for immersion.

I hope you can afford one one day. I think my chair is amazing and it adds so much to immersion, even with il2 with this delay I have. It still really feels like I am flying a place and it makes the game so much more of a pleasure to play.

[Lombra]

I've just seen that DCS lets server hosts disable telemetry output, so that could be an option if cheating is a concern with potential future telemetry.

[hexpod]

On 3/29/2019 at 10:41 AM, AndyJWest said:

I think it might be better if people stopped guessing at what they think the API might provide, and took a look at what it actually does. It provides values for orientation over three axes (pitch, roll and yaw), turn rates around these three axes, and acceleration data for the same three axes. Via UDP, at a maximum data rate of 50 Hz. That is all it does. Because that is what you need for its required purpose - to feed data to motion simulators. It doesn't provide high-frequency vibration data, because that isn't required for motion simulators. I don't know exactly how these seat-vibrator devices work, but I'd have thought they wanted something closer to a continuous audio-frequency signal, rather than UDP data blocks. From what I can figure out from the DCS forum, that is how it works with their sim.

 

If people want such data, I suggest they get together and put forward a proper request directly to Jason, making clear what exactly it is they are asking for, rather than making repeated posts in a thread about an API which has nothing to do with what they want, and isn't suited to providing it.

 

Hey @AndyJWest

Hey @Jason_Williams

 

After several tests we can confirm that positions and rates are consistent.

 

Nevertheless we have an issue with accelerations which I try to describe in the video.

 

As you can see, the coordinated turn outputs for some reason a bizarre lateral acceleration and the turn indicator is doing strange oscillations while doing barrel rolls.

 

Would you have for us another set of translational accelerations which are proving the local forces in more natural way ?

 

Thank you very much.

 

 

Edited February 4, 2020 by hexpod

[hexpod]

On 3/30/2019 at 6:11 PM, Jason_Williams said:

Thank you Value1 for the plugin and Gforcefactory for your fast support as well! 

 

And special thanks to Sergey our lead programmer who made the support quickly once we committed to it. 

 

Feedback from either implementation is welcomed. 

 

Now I need to buy a motion rig! 

 

We have the best community and team in simming. 

 

Jason

 

Dear Jason,

 

Could you please add just one more variable to the existing ones ?

I have in mind the raw data from the slip  indicator (the small ball).

 

Thats the only thing which is cruelly missing which could make complete our motion cueing algorithm.

 

For the moment the data for sway (lateral acceleration) you provided, doesn’t include the flight symmetry which is crucial for a good feel.

 

Could you please forward my request to Sergey or PM me his contact so I could directly write to him ?

 

Thank you for your concern,

 

Best regards - HeXpod.

Edited February 11, 2020 by hexpod