Not so stable?

[YoYo]

Hi guys, I saw the Casmo's video and one thing seems quite strange to me related to FM. The model in hover mode is very, very stable. I don't know if there is any wind here, but there should be some slight drift. When the missile is fired, it should also tremble and maybe even move back slightly due to the recoil, but nothing, it just stands in the air like a stump. Any comment for this? TY!

IRL: 

With ground effect:

For example, the DCS Apache has a slight drift even in auto hover mode.

Edited June 1, 2024 by YoYo

[Hiob]

IIRC from their Q&A, they opted for a limited Pilot-AI, to enable single players to use the armament from the co-pilots seat. Consciously trading realism for (single player) usability.

Not 100% sure if I relaying it correctly here and if it is related. Maybe take look at their Q&A yourself.

[Wostg]

Yeah they acknowledged that it might feel 'on rails' for the benefit of single players in the cpg seat vs all the hassles George brings, but also stated that they'd allow this to be turned off in mission/mp settings etc for those with objections

[YoYo]

19 minutes ago, Hiob said:

IIRC from their Q&A, they opted for a limited Pilot-AI, to enable single players to use the armament from the co-pilots seat. Consciously trading realism for (single player) usability.

Not 100% sure if I relaying it correctly here and if it is related. Maybe take look at their Q&A yourself.

Hmmm, ok, but they can add even a slight drift animation (even small), I never had a problem with Gazelle, even though I handled everything at the same time.

Edited May 27, 2024 by YoYo

[TheGhostOfDefi]

3 hours ago, YoYo said:

Hmmm, ok, but they can add even a slight drift animation (even small), I never had a problem with Gazelle, even though I handled everything at the same time.

 

Well you will have in Multicrew.

[ThorBrasil]

14 hours ago, YoYo said:

Hi guys, I saw the Casmo's video and one thing seems quite strange to me related to FM. The model in hover mode is very, very stable. I don't know if there is any wind here, but there should be some slight drift. When the missile is fired, it should also tremble and maybe even move back slightly due to the recoil, but nothing, it just stands in the air like a stump. Any comment for this? TY!

IRL: 

With ground effect:

For example, the DCS Apache has a slight drift even in auto hover mode.

 

 

By Polychop Sven

We made a statement about it in our Q&A. The AI is perfectly stable yes for gameplay purposes for single player. The Ai will counter any CG shifty immediatly, because it is AI. If you want it to fly with the full sense of motion, you have to trim it out with force trim, hover by yourself while you do all the work in left and right seat. Can be done, have done it and as far I remember even in the pilot flightschool you have a task where you fly and operate the aircraft from left seat, but my memory can serve me wrong on that one.
when you fly the aircraft without AI you will defenately feel the CG shift when a Hellfire goes of the rail, promised, even if you have it trimmed out as much as you can
same goes for the gun or any other weapon. lateral and longitudenal cg shift are a real thing in this dcs helicopter

 

[YoYo]

6 hours ago, ThorBrasil said:

 

By Polychop Sven

We made a statement about it in our Q&A. The AI is perfectly stable yes for gameplay purposes for single player. The Ai will counter any CG shifty immediatly, because it is AI. If you want it to fly with the full sense of motion, you have to trim it out with force trim, hover by yourself while you do all the work in left and right seat. Can be done, have done it and as far I remember even in the pilot flightschool you have a task where you fly and operate the aircraft from left seat, but my memory can serve me wrong on that one.
when you fly the aircraft without AI you will defenately feel the CG shift when a Hellfire goes of the rail, promised, even if you have it trimmed out as much as you can
same goes for the gun or any other weapon. lateral and longitudenal cg shift are a real thing in this dcs helicopter

 

Interesting and thanks, but I still think that even with this AI a slight drift could be achieved. People with HOTAS or collective levers can get by with having several POVs, small joysticks and can use it at the same time while piloting. I remember that Gazelle is not that stable, even in auto hover. We will see soon :).

[Kinkkujuustovoileipä]

40 minutes ago, YoYo said:

Interesting and thanks, but I still think that even with this AI a slight drift could be achieved. People with HOTAS or collective levers can get by with having several POVs, small joysticks and can use it at the same time while piloting. I remember that Gazelle is not that stable, even in auto hover. We will see soon :).

Autohover is not the same as AI hover.

[Dangerzone]

1 hour ago, YoYo said:

Interesting and thanks, but I still think that even with this AI a slight drift could be achieved.

It quite could be, but I think it's a matter of goals and priorities. There are 2 ways to achieve this. Start with the drift/realism as a higher priority and work the way to stability (aka, George) or start with stabilization as a higher priority and work the way towards realism. Whichever way is used, there will be an element of err, or 'not being far enough' on the side of caution. Given the issues I have faced with George over time - especially in the hover - I welcome the new approach of 'on rails'. While not perfect, I think it's the best of the too evil. As for not being perfect, what's the saying? "Perfection is the enemy of good".

That's not to say I wouldn't encourage additional work down the track towards this goal, but not something I would consider a day 1 holdup to get right first, even in a non early access model.

Also, if stability with just a slight drift was easy enough - I dare say we wouldn't have gone through the issues we faced with George either. There may be other factors at play that make this first more difficult than is perceived. In 30 years of professional coding I have learned that what appears, and should be 'simple' on the surface often has hidden obstacles that aren't revealed until part way through - and it can be very difficult (or impossible) to relay the reality of this to people who haven't seen those obstacles for themselves. As such, I often tend to give dev's the benefit of the doubt - at least until proof of concept is done by someone else where the code is visible. 

[YoYo]

6 hours ago, Kinkkujuustovoileipä said:

Autohover is not the same as AI hover.

I know, so is autohover present in Kiowa too? Can I use autohover and jump to the second station?

Edited May 28, 2024 by YoYo

[Ramsay]

3 hours ago, YoYo said:

... is autohover present in Kiowa ?

No.

Edited May 28, 2024 by Ramsay

[Hiob]

3 hours ago, YoYo said:

I know, so is autohover present in Kiowa too? Can I use autohover and jump to the second station?

 

depends on what you mean. In the sense of an actual simulated system of the real aircraft - no.

As part of the interface to make it accessible for single players - yes (it‘s called AI hover and works differently than what we originally had in the Gazelle).

Edit: sorry - should have read two or three posts further back…. anyway the answer above is still true

Edited May 28, 2024 by Hiob

[Schmidtfire]

The question is if Polychop would consider adding a very tiny bit of movement to the AI hover.
More in line with the Cockpit Shake found in Special Options tab. It's a small effect rather than
a complex movement that will mess up your targeting. It should not affect the reliability of AI auto-hover,
just making it less artificial. Would be great to see it as an option.

 

[YoYo]

21 hours ago, Schmidtfire said:

The question is if Polychop would consider adding a very tiny bit of movement to the AI hover.
More in line with the Cockpit Shake found in Special Options tab. It's a small effect rather than
a complex movement that will mess up your targeting. It should not affect the reliability of AI auto-hover,
just making it less artificial. Would be great to see it as an option.

 

+1.

[zerO_crash]

Conceptually, while it did work with drift in Sa-342, it was somewhat off. On one side, there will be slight corrections all the time. On the other, as a co-pilot, you had to constantly monitor your position, that, especially in concealed areas.

 

With the way KW does it, it might seem almost too stable, true (possibly some wobbling/drift can be added with time). However, you don't have to worry about the piloting-side of things. You can focus fully on the task of warfare.

[fjacobsen]

Maybe we should simply come down to the fact that in single player - one person has to do two persons job.

'So while a slight "drift" would be nice, I can also accept the current way they have done it.

[YoYo]

15 hours ago, fjacobsen said:

Maybe we should simply come down to the fact that in single player - one person has to do two persons job.

'So while a slight "drift" would be nice, I can also accept the current way they have done it.

You just have to remember that, in principle, no human can perfectly stay in one place in the air for a long time without even the small movement (even up and down) and even this situation does not reflect the fact that there would be two people on the crew (especially in multicrew). Something else is automatic auto-hover. So no, it doesn't simulate two persons in the cockpit as well. Its more like arcade / easy hover function, however now its only speculation, we need to check it after the release.

Edited May 31, 2024 by YoYo

[YoYo]

On 5/28/2024 at 8:33 AM, Kinkkujuustovoileipä said:

Autohover is not the same as AI hover.

Btw. If no autohover so please to think about adding some kind of human behavior before release and possibly the possibility of drifting or small up / down movement (maybe with wind direction?). If this is meant to simulate human behavior (second crew), we cannot assume that we have a crew member holding the Kiowa in the air as if it were standing on the ground (and this is what it looks like in the videos, apart from the slightly shaking cockpit). It is simply impossible to stay in the air by hand for a long time with the same result like on the ground. So logically, since KW does not have an autohover - you can't assume that a human crew member is a Terminator who doesn't make any mistakes and his hand doesn't shake if the entire cockpit shakes (this also affects the hover, after all).

This is good example (4:10).  Additionally, of course, it is a hover using the ground effect, which makes it even easier to maintain the helicopter. You can clearly see how the whole machine shifts slightly and the pilot keeps correcting the tilt, returning to the previous position. 

Generally, it should behave like this as default FM, and possibly the super stable AI option should be enabled in the Special tab, if someone wants less realistic behavior and a more fictional flight model in hands of AI.

Edited May 31, 2024 by YoYo

[YoYo]

The problem of FM of AI is clearly visible in the new Casomo's video. He just acts like he's on rails. Its not a PAUSE :

Like it was basically standing on the ground, not hanging in the air.

Edited May 31, 2024 by YoYo

[zerO_crash]

I was looking at it, and yes, it seems a little too "perfect". Wobbling/swaying, delay between issuing a command and fulfilling it (not instant) and generally a sensation of proper flight model, would be preferred

[Czar66]

On 5/27/2024 at 5:30 AM, YoYo said:

Hi guys, I saw the Casmo's video and one thing seems quite strange to me related to FM. The model in hover mode is very, very stable. I don't know if there is any wind here, but there should be some slight drift.

 

I've watched the video when it came out. It bothers me to no end showcase videos in DCS that are laboratory like with zero wind and zero turbulence. The actual prime time for showcase the module before release especially the flight dynamics.

There is nothing going on wind related in the video making the heli movements seems like a blender (or any other software like it) animation. Static and dull. It is not the blame on the module (perhaps) but whoever made the mission. The smoke from the destroyed vehicles goes straight up.

If only they would put a bit of wind of 2 to 5 knots and turbulence above 10. Some shots inside the cockpit could very well look like the helicopter was on a pad shooting missiles.

I can't comment about the missile shot recoil. In my understanding the missile propels itself forward unlike a cannon where the gun needs to react to the gas to push the shell out on the other end. I just miss a bit of wobble of the weight shift, maybe.
 

Edited May 31, 2024 by Czar66

[YoYo]

29 minutes ago, zerO_crash said:

I was looking at it, and yes, it seems a little too "perfect". Wobbling/swaying, delay between issuing a command and fulfilling it (not instant) and generally a sensation of proper flight model, would be preferred

Of course, we have to separate the flight model into the human one and the AI one, but here we are talking about the flight model for AI. The human one is probably correct and very accurate, from what I can see.

[zerO_crash]

18 minutes ago, YoYo said:

Of course, we have to separate the flight model into the human one and the AI one, but here we are talking about the flight model for AI. The human one is probably correct and very accurate, from what I can see.

That's what I'm talking about, hence "... sensation of proper flight model...". For the moment, AI cannot use human flight model, otherwise it would hit the performance. On the other side, AI model, should in this case give the sensation of being closer to the human one.

 

Actually, the best comparison here is the Mi-24P and AH-64D (same goes for fixed wing). There, the AI actually operates on the same FM the human pilot does. I'm demanding the same from KW, but it feels way too static currently.

 

That does not say anything about the FM the pilot has. As I hear, it's pretty good, and in line with IRL (except roll being too sensitive - needs some tuning).

[YoYo]

Of course, I'll take KW anyway in the 1st day and I'm looking forward to it, but I hope they will do something about it. I like the way George AI or Petrovich AI works, where we have a certain inertia and small kind of "instability" of the front of the copters if you use AI help.

[zerO_crash]

Exactly my point. It needs some tuning, especially when you compare the other modules with AI