[Suggestion] Spring joystick control.

[Super Grover]

37 minutes ago, kablamoman said:

With my spring-loaded stick, I can no more feel the P-51 control forces than I can those in the F-4E.  And yet the former module respects my control position, while the latter does not.

I'm sorry you feel this way. We, also on behalf of the F-4E, respect you and your position in every possible way.

[kablamoman]

2 minutes ago, Super Grover said:

[...] We, also on behalf of the F-4E, respect you and your position in every possible way.

And I too!

Again, I really appreciate all the hard work. I completely understand the impetus to pursue some of these kinds of things as well. I fly for a living and understand all too well that some of the visceral feel involved in the act is all too often just out of reach of simulation. Fancy equipment and VR gets us close.

Talented developers also help bridge the gap in many innovative ways, as you have all shown you are capable, both in the Tomcat before, and now again with the F-4E. I am sorry to focus on this one aspect so acutely, but it's only because it's the one thing that stands out to me in an otherwise monumental and beautiful offering.

 

 

[IronMike]

One of the big disconnects everyone sees, is rather visual. I added that to the "issues you may encounter" thread. This will go away eventually once you get used to it. But the feel is overall closer and the system, realistic, spring loaded or not. If you focus on what the stick in game does predominantly, you're not giving yourself a chance to get used to it, to just fly it and feel it. You are looking at the stick because you saw "ah there is a lag in the animation." Ask yourselves how much you actually look at what the stick does in other modules.

If you want to help yourselves getting used to it, you can hide the stick, and fly around for a while. Connect the feel for the FM and what your actual stick is doing, leave force blending on and stick force at default 60lbs. I am not saying you have to hide the stick to get used to it, or forever fly with a hidden stick. But if "unseeing" it is difficult, you can try if it helps.

And no, our SMEs did not just give up arguing about it because they grew tired of it, as one of you suggested. We discussed FM with them on a daily basis for well over a year, several times a day. They grew fond of it. They gave it a chance. And according to them, it now flies and feels as close to real life and their memory as possible, though minor improvements are still planned - but these are not related to this topic. Only one out of our 12 SMEs uses a force feedback stick.

I ask you all kindly again, give it some time please. It threw me off in the beginning, too, after a while one does not notice it anymore, unless comparing on purpose. Your inputs in the FM are not delayed, except through occurring inertia in the FM, as in the real jet.

Whether or not we change it, is not up to debate, and won't be, please be so kind and understand that. As much as we like to cater to you all. Whether or not you give it an open minded chance, is up to you on the other hand.

The movement of the 3d model of the stick in the cockpit goes through a type of low-pass-like filter. Otherwise, it might look erratic and actually even less realistic. However, it doesn't represent the actual physical simulation of the entire stick simulation, and that means that the change of your input is instantaneously transferred to the controls system. Yet, that system runs a complex simulation that adds realistic inertia to the mechanisms that transfer your input to the power actuators.

To summarise, the lag you may see when observing the 3D controls doesn't represent the system's delay; there might be some realistic inertia reducing the controls' responsiveness.

 

Edited May 24, 2024 by IronMike

[Aero4000]

If you give us a flexible setup, you won't lose anything. Everyone will get a visual demonstration that you are right that hard-wired sticks make flying terribly unrealistic. Please understand that in the community of virtual pilots, a visual demonstration of FM features is much more important than anything else in an airplane flight simulation. So let us touch both. But you just say NO. No one will accept this.

[Super Grover]

Actually, everyone would lose, and I'll explain it in detail soon.

[Baldrick33]

1 hour ago, IronMike said:

One of the big disconnects everyone sees, is rather visual. I added that to the "issues you may encounter" thread. This will go away eventually once you get used to it. But the feel is overall closer and the system, realistic, spring loaded or not. If you focus on what the stick in game does predominantly, you're not giving yourself a chance to get used to it, to just fly it and feel it. You are looking at the stick because you saw "ah there is a lag in the animation." Ask yourselves how much you actually look at what the stick does in other modules.

For me it is totally visual, it doesn’t change how I feel the aircraft react to my inputs it is purely an immersion factor. Much like a steering wheel where the virtual and physical wheel rotation doesn’t match, or there is a lag between them, it doesn’t make it harder to drive just an annoying immersion breaker.

I agree the more I fly the less I notice it as typically we don’t look at the stick, just that when it catches your eye doing something different to what your hands are doing it kind if gets hard to unsee!

Regardless the Phantom is fantastic and this is just a minor request from my perspective.

Edited May 24, 2024 by Baldrick33

[IronMike]

2 minutes ago, Aero4000 said:

If you give us a flexible setup, you won't lose anything. Everyone will get a visual demonstration that you are right that hard-wired sticks make flying terribly unrealistic. Please understand that in the community of virtual pilots, a visual demonstration of FM features is much more important than anything else in an airplane flight simulation. So let us touch both. But you just say NO. No one will accept this.

Please understand that what we cannot accept, is that folks want to tell us that the FM and stick input implementation is wrong, and to change it, after flying it for precisely 1 day or much less even in hours. and without having flown the real jet.

The reactions we get from SMEs (12 of them who flew the jet for thousands of combined hours and one of them who still flies it today) - are a destinct testimony to the opposite. The FM has been build based on heavily including their input, every step of the way. Again, whether you give it an open minded chance, or not, is up to you.

Fly it for a couple weeks, a month even - then tell us what you think. Of course that won't help much, if you never give it a real chance. If you are a real pilot, you will know that judging an FM and getting a good feel for it after a few hours of flight is not possible. I invited you to give it an honest try. You keep coming back with "no". After spending likely less than a full day with it. Please be so kind and understand that that won't just change our mind.

Don't get me wrong. I know where you are coming from. I felt very similar initially. And all I can say, I am so, so happy that I didn't convice Grover to go the classic route, as I initially wanted to, when I first tried it. I truly and fully know where you are coming from. But so many things that we were able to do that work as in the real aircraft, are made possible through this. And the rabbit hole goes deep. From bellow intakes freezing at high alt, to flipping backwards with pitch down stick in a clean tailslide to so many quirks and peculiarities in between, to just name a random few, which occur now naturally, most importantly - made it possible for us to take the FM to the next level, and we will continue improving on it, too. But you can't explore and experience it, if you don't give it an honest chance. I can only kindly invite you again to do so.

 

[unlikely_spider]

Thank you, members of HB, for the detailed explanation and for your presence on these boards.

I have to admit that last night I flew around a bit, ignoring the visuals of what the stick was doing, and it 'felt' better than the earlier flight when I was positively confused about the flight stick movements. ('Felt' is in quotes because I've only got experience in single engine props IRL and my only jet experience is limited to computer sims)

I trust that you were meticulous in your testing and positive notes from SMEs trump my own due to my aforementioned limited experience.

Perhaps it would have been best to have these forces acting 'behind the scenes' as it were, and not visually represented in the stick movement, especially when it's counter to what were actively doing with our hardware. That way we can be affected by all appropriate forces and can enjoy the flight model all the same, but without the confusing aspect of the stick seemingly having a mind of its own. Unfortunately I believe you will continue to receive messages from confused users throughout the life of this module otherwise. In the beginning this veteran flight simmer was pulling my hair out believing that I had double-bound an axis or had something else incorrectly configured.

Edited May 24, 2024 by unlikely_spider

[IronMike]

1 minute ago, unlikely_spider said:

Thank you, members of HB, for the detailed explanation and for your presence on these boards.

I have to admit that last night I flew around a bit, ignoring the visuals of what the stick was doing, and it 'felt' better than the earlier flight when I was positively confused about the flight stick movements. ('Felt' is in quotes because I've only got experience in single engine props IRL and my only jet experience is limited to computer sims)

I trust that you were meticulous in your testing and positive notes from SMEs trump my own due to my aforementioned limited experience.

Perhaps it would have been best to have these forces acting 'behind the scenes' as it were, and not visually represented in the stick movement, especially when it's counter to what were actively doing with our hardware. That way we can be affected by all appropriate forces and can enjoy the flight model all the same, but without the confusing aspect of the stick seemingly having a mind of its own. Unfortunately I believe you will continue to receive messages from confused users throughout the life of this module otherwise.

That is totally ok, and that is what we are here for. It will become easier to understand, once Grover puts up an explanation and how things work. Unfortunately without the low pass filter, it would look even less realistic, but we will explain that all in depth. We are currently just busy preparing a hotfix for you all, so the initial bugs get squashed.

[kablamoman]

48 minutes ago, IronMike said:

Please understand that what we cannot accept, is that folks want to tell us that the FM and stick input implementation is wrong, and to change it, after flying it for precisely 1 day or much less even in hours. and without having flown the real jet.

It's not that I don't think I can get used to it. It's that I understand but disagree with the entire premise of how you've chosen to model the stick.

Fundamental to flying in a non-FBW aircraft is the relationship of stick position to the deflection of control surfaces. The real F-4, despite the control feel mechanisms, still maintains the relationship of stab position to stick position regardless of how the control forces may be being modulated by the system.

Control feel with non-FFB hardware is an old problem in sims that's been tackled many different ways. The reason I fly DCS and not IL-2 is because DCS generally retains the key relationship between the virtual stick and your control peripheral, and therefore control surface deflection itself, foregoing the conceit of attempting to emulate feel in software. This module is now breaking that tradition, and despite the beauty of the entire package it really is a deal breaker for me.

It's not just the stick moving, and so it's not just a visual problem: The stab is moving along with it.

As such, the relationship between my stick at home and the stab in the game is broken. 

Feel is irrelevant to this critical point, and it can and should only be added as a layer on top of this with proper FFB support.

By abstracting this function, you have not-so-arbitrarily broken the critical link between the pilot and the sim. I say "not-so-arbitrarily" because there is a reasoning behind you doing this, and it is based on actual simulated control forces. But again, the fact that you've modeled the bobweight and bellows system is not at issue, despite the fact that you seem to be implying that's what some are complaining about or needing help to understand.

I believe it is also invalid to say that changing the stick simulation would somehow invalidate all the other good work you've done: One need only look at any other non-FBW module in the sim. All of those aircraft in real life were flown by pilots subjected to the same types of control forces and pressures as the F-4, and often have their own distinct systems just as the F-4 with its bobweights and bellows, and yet each one of them has been fully realized in the sim without having this particular issue.

You can have all the same control forces and reactions included in your model -- just as the warbirds and other non-FBW aircraft do -- but without decoupling the stick as you've done.

Edited May 24, 2024 by kablamoman

[Rudel_chw]

1 hour ago, IronMike said:

If you want to help yourselves getting used to it, you can hide the stick, and fly around for a while.

 

Thanks for the tip, I'm hiding it and still getting used to the aircraft 

Edited May 24, 2024 by Rudel_chw

[lengro]

21 minutes ago, IronMike said:

Please understand that what we cannot accept, is that folks want to tell us that the FM and stick input implementation is wrong

I don't think anyone is arguing that the FM or input implementation is wrong.

Visually it just looks odd that the 3D stick moves around when the stick on my desktop doesn't.

Treat the input send to the sim exactly as you do now, but allow an option to let the - visual, and only visual - 3D position of the stick to be directly linked to the physical stick.

Thanks for your hard work, it's a wonderful module!

Edited May 24, 2024 by lengro

[Aero4000]

I really like your Phantom. I am very happy to fly it. I have no serious comments on how it feels on the handle of an ordinary joystick. Because it's very similar to the jets I've flown and still fly. I'm just confused by the animation of the virtual control stick and stabilator. Maybe it's really better to leave it behind a screen? Good idea at current point of time!

[FusRoPotato]

I think you're addressing a different issue than what has been discussed here entirely, as there are different ideas bouncing around that lead the discussion in a different direction. A concern is not just visual lag but the oscillations and difficulties in gradual maneuvers due to the lack of tactile feedback in non-force feedback joysticks.

Whether you choose to debate or change it is beside the point (seems like you are choosing to debate it anyways). If the current system feels unrealistic and users are unhappy, it will inevitably be a topic of discussion, potentially becoming heated. Attempts to dismiss these concerns with straw man arguments, appeals to authority, or by conflating them with other issues will only amplify any dissatisfaction.

I personally don't feel too bothered by it, but I am seeing a lot of people express frustration about it and I understand why:

• An attempt was made to simulate force responses into sticks that don't provide force feedback.
• To make that work, assumptions were made as to how a pilot would react (or not react well) to said forces. In other words, This isn't because of how the stick is modeled, it's because of how the virtual pilot's reaction is or is not assumed. This creates unexpected control inputs leading to oscillations and control annoyances that are much less likely to exist with FFB where that tactile feedback is present.
• There seems to be an attempt to shift the topic and avoid responsibility in reluctance to engage with these issues or re-frame them as stick modeling. Failure to recognize how much that weakens claims of SME support can render a lot of those statements hard to trust.
• You are right, it could be easy for many to get used to, but that's aside the point. If flying a sim with precision using FFB is easier than without, there is work to be done with the non-FFB counterpart. Most of your consumers won't have FFB. It is very important to understand this.

Things are straight forward and easy once you have all your moments, leverage, and dampening sorted from tail to stick. You just send those forces happily to an FFB stick and you're good to go.

However, there is a general practice in MS&T: If the trainee can't sense it, either offer an preventative substitute cue, or inhibit/minimize the effect. I cannot reiterate this strongly enough: The disconnect between a spring-stick PC-sim and an aircraft disallows a certain level of realism, like some of the trim options for Helos in DCS. Things can go very badly for a products image if care isn't taken to respect that disconnect. Some things need to be accounted for or hidden in such a way that usage of a spring stick mimics an aircraft being expertly flown by someone with FFB or realistic recreation of a simpit. Otherwise, you will be expecting a pilot to learn and practice skills that are not typical because that feel can't be translated.

Edited May 24, 2024 by FusRoPotato

[Baldrick33]

9 minutes ago, lengro said:

Treat the input send to the sim exactly as you do now, but allow an option to let the - visual, and only visual - 3D position of the stick to be directly linked to the physical stick.

That would seem ideal to me.

[Super Grover]

30 minutes ago, lengro said:

I don't think anyone is arguing that the FM or input implementation is wrong.

Visually it just looks odd that the 3D stick moves around when the stick on my desktop doesn't.

Treat the input send to the sim exactly as you do now, but allow an option to let the - visual, and only visual - 3D position of the stick to be directly linked to the physical stick.

Thanks for your hard work, it's a wonderful module!

 

Yes, that's a feature we can and will add - just a 1:1 animation without any smoothing. Similarly, we want to add it to the pilot body, which, when smoothened, may cause unpleasant sensations in VR for some. I can't promise the timeline, but such an option will undoubtedly appear.

[IronMike]

34 minutes ago, Aero4000 said:

I really like your Phantom. I am very happy to fly it. I have no serious comments on how it feels on the handle of an ordinary joystick. Because it's very similar to the jets I've flown and still fly. I'm just confused by the animation of the virtual control stick and stabilator. Maybe it's really better to leave it behind a screen? Good idea at current point of time!

Like I said, I totally get you, 100%. It really threw me off for the first time with it, too. But once I wrapped my head around it, and rolled with it - I started realizing that it just opens up so much more we can do overall. And I couldn't imagine going back at all now. I really, honestly could not tell, when it was the last time that I noticed the stick animation throwing me, and I do like to glare at my stick every now and then. But I really can't remember when the last was exactly. Given, I have had well over a year with it and saw it progress every step of the way. But the coolest thing is, that so many things fall into place naturally. You model it, as was, and there it is. That is the idea underlying the entire components system. Things behaving naturally, not scripted anymore. Having a full blown stick simulation, not just a fake trim axis, etc. Btw, this is not meant demeaning in any kind of way, it's a valid, and tried method for sure. We apply it ourselves in our other modules. But this just lets us go a bit further.

There is always blowback when you break with a tradition. It's not really something we will allow to deterr us. I shall leave it at that for now, and kindly ask you all to wait till @Super Grover presents the system to you all more in depth, then we will be happy to discuss it further.

Just to put it into perspective in reply to some other comments above: from all the issues we gathered over the past day, this was the least mentioned one. It is only being debated by some in a heated manner, because they feel strongly about it. And that is ok. Please also do not misunderstand us, we value such input, we listen to it, and we debate it lively. But at the same time you cannot expect us to change a fundamental part of our module, because so far a few dislike it and want us to change it after flying with it for a day. I have answered hundreds of questions over the past two days, on forums, social media, reddit, discord - not once was this issue brought up to me outside of the two threads on these forums. That said, I am sure there are more who don't like it. But let us please not make it out to be, as if the vast majority of the community is rejecting it. It is not.
 

Edited May 24, 2024 by IronMike

[Maksim Savelev]

1 час назад, kablamoman сказал:

It's not that I don't think I can get used to it. It's that I understand but disagree with the entire premise of how you've chosen to model the stick.

Fundamental to flying in a non-FBW aircraft is the relationship of stick position to the deflection of control surfaces. The real F-4, despite the control feel mechanisms, still maintains the relationship of stab position to stick position regardless of how the control forces may be being modulated by the system.

Control feel with non-FFB hardware is an old problem in sims that's been tackled many different ways. The reason I fly DCS and not IL-2 is because DCS generally retains the key relationship between the virtual stick and your control peripheral, and therefore control surface deflection itself, foregoing the conceit of attempting to emulate feel in software. This module is now breaking that tradition, and despite the beauty of the entire package it really is a deal breaker for me.

It's not just the stick moving, and so it's not just a visual problem: The stab is moving along with it.

As such, the relationship between my stick at home and the stab in the game is broken. 

Feel is irrelevant to this critical point, and it can and should only be added as a layer on top of this with proper FFB support.

By abstracting this function, you have not-so-arbitrarily broken the critical link between the pilot and the sim. I say "not-so-arbitrarily" because there is a reasoning behind you doing this, and it is based on actual simulated control forces. But again, the fact that you've modeled the bobweight and bellows system is not at issue, despite the fact that you seem to be implying that's what some are complaining about or needing help to understand.

I believe it is also invalid to say that changing the stick simulation would somehow invalidate all the other good work you've done: One need only look at any other non-FBW module in the sim. All of those aircraft in real life were flown by pilots subjected to the same types of control forces and pressures as the F-4, and often have their own distinct systems just as the F-4 with its bobweights and bellows, and yet each one of them has been fully realized in the sim without having this particular issue.

You can have all the same control forces and reactions included in your model -- just as the warbirds and other non-FBW aircraft do -- but without decoupling the stick as you've done.

 

Interestingly enough is that when you connect FFB decoupling is gone, stick is linked to my FFB. What on earth is preventing you to make it for spring joystick. Just fake it, like its  FFB connected to the system and thats it. No trim - its not that a big deal, decoupling is a HUGE deal

Edited May 24, 2024 by Maksim Savelev

[lengro]

27 minutes ago, Super Grover said:

Yes, that's a feature we can and will add - just a 1:1 animation without any smoothing.

So much appreciated, thanks!

[funkyfranky]

13 minutes ago, Maksim Savelev said:

Interestingly enough is that when you connect FFB decoupling is gone, stick is linked to my FFB. What on earth is preventing you to make it for spring joystick. Just fake it, like its  FFB connected to the system and thats it. No trim - its not that a big deal, decoupling is a HUGE deal

 

That was indeed also my initial expectation, i.e. that with a non-FFB joystick, you are essentially pulling/pushing with an infinite force (therefore, not being desynced between physical and in-game representation of the stick). 

Of course, with FFB, it is at least possible to emulate the forces on the stick. But if you are able to synchronize an FFB stick (e.g. by pulling it back to the center position whilst on ground), it seems strange that a non-FFB stick is necessarily de-synchronized.

Got to dig out my good old Sidewinder to check for myself.

[kablamoman]

2 hours ago, IronMike said:

Things behaving naturally, not scripted anymore. Having a full blown stick simulation, not just a fake trim axis, etc. Btw, this is not meant demeaning in any kind of way, it's a valid, and tried method for sure. We apply it ourselves in our other modules. But this just lets us go a bit further.

I'm sorry but you don't have a full-blown simulation of the stick: This weird behavior is only enabled in the pitch channel with a non-FFB stick! Aerodynamic loads in roll (that are just as valid as the control pressures generated by the bobweight and bellows system in pitch) are not modelled in this fashion at all (nor should they be), but in the traditional way every other non-FBW module does it -- which makes it even more of an inconsistency.

I suspect that at the core of this is the fundamental mistake of conflating the control pressures under load with stick position, they cannot be treated the same. By conflating it in this way you have decoupled and broken the pitch axis and induced uncommanded oscillations of not just the visual representation of the stick -- but of the horizontal stab and thus the aircraft itself. I repeat, it's not just the visualization of the in-game stick: It's the stab and the aircraft itself, as well.

The neutral trim point (small diamond on the controls indicator) acts as a re-scaling reference point for our joystick input pitch axis in our non-FFB sticks. It's our fixed spring and cam loads in real life that provide us our approximation of control pressure, and the in-game trim is the means by which we re-align the in-game neutral trim position with that center. 

Edit: Struck out the part about the roll channel as the F-4E roll control is not actuated directly but via hydraulics.

Edited May 24, 2024 by kablamoman

[Super Grover]

Please note that the roll channel in the F-4 doesn't "emulate" any aerodynamic loads, and it's centred by a pretty simple spring.

Edited May 24, 2024 by Super Grover

[kablamoman]

20 minutes ago, Super Grover said:

Please note that the roll channel in the F-4 doesn't "emulate" any aerodynamic loads, and it's centred by a pretty simple spring.

A spring attached to a stick, which is also attached to a control linkage attached to control surfaces experiencing real loads!!!

From a simulation standpoint those loads are no different and exactly what the bobweight and bellows system is designed to emulate!!

Edit: I should clarify I don't know if the linkage to the aileron control is hydraulically boosted or not, but the point I am trying to make is that the bobweight/bellows provides a similar control feel to a traditional directly linked control surface. In a traditional warbird, for example, you would experience increase aileron loads at higher dynamic pressures (speed/deflection). We don't oscillate the input axis trim neutral point for those every time the load changes there in sims without a FFB stick for good reason

Edited May 24, 2024 by kablamoman

[BMS]

It looks like the developers are standing their ground. It is a pity that reasonable arguments cannot outweigh the reluctance to correct the original mistake. These are expenses, and the project is already being sold. However, I bought the module in anticipation of DСS standards in aircraft control.

If the DCS community continues to rely on beautiful videos from paid bloggers, commercialization will inexorably push real aviation enthusiasts out of the project, giving way to grateful buyers of impressive promos. However, this process is already underway.

I hope the developers will listen to people who understand the problem and change their position.

[Super Grover]

2 minutes ago, kablamoman said:

A spring attached to a stick, which is also attached to a control linkage attached to control surfaces experiencing real loads!!!

From a simulation standpoint those loads are no different and exactly what the bobweight and bellows system is designed to emulate!!

That's a negative, Maverick. The control linkages are not attached to the control surfaces, so there is no feedback from the control surfaces to the stick as the control surfaces are moved by hydraulic power actuators.