wip Uncommanded Catastrophic Rolls

[Zadren]

Please see the end of the attached tracks. On several occasions now, the heli seems to fall into an unrecoverable roll for reasons I can't fathom.

I think, in the first track, I'm struggling to get into a hover and it falls over. In the second, I'm just goofing around with the gun in normal flight, and it rolls into the water. There doesn't seem to be any strange inputs on review.

Wrt hovering, I know I'm a bit of a ham-fisted clot, but I could have sworn I was starting to get the hang of it until you changed something. Now, it's infuriating trying to get into a hover after forward flight.

Thanks.

ApacheRoll.trk ApacheRoll2.trk

[admiki]

I watched only 1st track. I couldn't be bothered to watch you inputing waypoints. Please, if you are trying to report something, make a habit of going for it instead of messing around.

Anyway, I did not see anything catastrophic in first track. I saw few rolls to the right, but those are just reaction to your somewhat big left pedal inputs. If there is anything else that happened to you, think about recording a video since tracks can get broken. 

[Zadren]

I said it was at the end of the track, why would you even consider sitting through the entire thing? I saved the track as it happened in normal play. I wasn't specifically looking for a bug, therefore the track - which we are constantly asked to submit - is as long as it is.

The second one is much shorter. Again, it falls out of the sky at the end.

I could send you a video, but maybe I can't be bothered either...

[admiki]

How do you get to the end of track?

[Floyd1212]

In the second track, it looks like you are trying to gun a target below you, and as you fly over it and try to get eyes on it again, you end up slipping sideways a bit.  I noticed last night when flying (and trying to keep my nose on some infantry on the ground) that just a little bit of pedal while side slipping can induce a rolling tendency.  This has always been the case, but it does feel more pronounced now after the recent patch.

7 minutes ago, admiki said:

How do you get to the end of track?

CTRL+Z to accelerate time

SHIFT+Z to return to 1x speed

(Not always recommended when a lot is going on, but just to skip past the data entry, and admiring your 1-foot hover in 3rd person, it's fine.)

Edited April 18, 2023 by Floyd1212

[Zadren]

23 minutes ago, Floyd1212 said:

....and admiring your 1-foot hover in 3rd person, it's fine.)

 

Alas, that isn't my pilotage doing that. I engaged hold mode before I lifted. I genuinely suck. 

[admiki]

1 hour ago, Floyd1212 said:

CTRL+Z to accelerate time

SHIFT+Z to return to 1x speed

I know time commands, but tracks are messed up as it is. Accelerating them just makes it worse.

[Abahji]

I'm not rated, so cannot fully explain the gyroscopic forces at play here. I can offer my observations though. The flight model from what I understand is a work in progress, and from what I've seen since release it has taken some pretty fantastic steps forward, and I doubt what you are seeing is a flight model bug? In fact, it seems like the condition went from almost instantaneous when the module dropped, to now taking some seconds to develop in the new flight model adjustments. Which I would guess is closer to RL.

Again I am no expert, but if you have a lot of power pulled in (blade pitch), and have a lot of forward cyclic input for a fast forward acceleration or just flying faster, and then suddenly drop the collective, without centering the cyclic in conjunction with flattening the pitch, the bird will go zero G or even negative. With a whole bunch of forward cyclic and the sudden unloading of the system, I guess it is the gyroscopic forces taking over, and rolling her right over?

You can readily duplicate this behavior, at will, by simply trying zero or negative G maneuvers with the collective bottomed. You can get some pretty wild roll rates going if you hold it fwd and allow the roll rate to gain momentum, giving yourself plenty of room for recovery (Pull in collective and center/correct for roll the cyclic). For some real fun, throw in some pedal deflection.
It is probably something to practice so you can quickly identify it developing, and correct your inputs. Much the same way you learn to quickly identify VRS or "settling with power." Pretty soon it's second nature for you, and you may even start using it to your advantage, just like you do with many other maneuvers that leverage the natural forces. Gravity, right hand turns, etc.  

To solve, as you lower the collective, simply make a corresponding adjustment of the cyclic, this helps the system stay loaded and avoids the condition entirely. 

[Hiob]

Just chiming in here....

I was flying the Apache for the first time after a long time yesterday - and had the same issue....

However, I quickly realised that my pedal input was all over the place.

What we did, was flying a couple of different Helis in SA. On its own and in comparison - the sensitivity of the Apache to the pedals seemed a bit off to me. Each small Pedal input would make the ball jump from left to right and vice versa, I was barely able to keep it centered. (15% Curve and 70% saturation on the input axis)

I don't claim for a second, that the problem can't be me though. Maybe I'm missing something.....

[NeedzWD40]

A friend saw this behavior a few days ago and I have a nagging suspicion it's related to the yaw SAS over-controlling things. Have you tried turning off the yaw SAS channel and seeing if the behavior is replicated?

[BIGNEWY]

On 4/18/2023 at 5:10 PM, Zadren said:

Please see the end of the attached tracks. On several occasions now, the heli seems to fall into an unrecoverable roll for reasons I can't fathom.

I think, in the first track, I'm struggling to get into a hover and it falls over. In the second, I'm just goofing around with the gun in normal flight, and it rolls into the water. There doesn't seem to be any strange inputs on review.

Wrt hovering, I know I'm a bit of a ham-fisted clot, but I could have sworn I was starting to get the hang of it until you changed something. Now, it's infuriating trying to get into a hover after forward flight.

Thanks.

ApacheRoll.trk 10.27 MB · 1 download ApacheRoll2.trk 6.52 MB · 2 downloads

Hello, 

the flight model is being tweaked and you will see changes in future patches. 

With that said, you need to fly smooth, ham fisted control inputs wont help. 

[macedk]

On 4/20/2023 at 1:54 PM, BIGNEWY said:

Hello, 

the flight model is being tweaked and you will see changes in future patches. 

With that said, you need to fly smooth, ham fisted control inputs wont help. 

Just another plea to include FM tweaks in the patch notes. People are chasing ghost everytime.:)
It doesn't have to be the hole 9 yards, but just a notice that it has changed
Ghostbuster out

[Raptor9]

1 hour ago, macedk said:

Just another plea to include FM tweaks in the patch notes. People are chasing ghost everytime.:)
It doesn't have to be the hole 9 yards, but just a notice that it has changed
Ghostbuster out

If changes are made to the actual flight model, we do. However, we have no control over subjective impressions posted by players. Following almost every update since release some have claimed something changed, and we have no control over that or what people choose to believe.

This is common in other areas of discussion too. Even when an SME states something factual about the AH-64 to help in understanding, there are those that choose to believe something else.

You can bring a horse to water, but you can't make it drink.

[lefuneste01]

33 minutes ago, Raptor9 said:

If changes are made to the actual flight model, we do. However, we have no control over subjective impressions posted by players. Following almost every update since release some have claimed something changed, and we have no control over that or what people choose to believe.

This is common in other areas of discussion too. Even when an SME states something factual about the AH-64 to help in understanding, there are those that choose to believe something else.

You can bring a horse to water, but you can't make it drink.

But at the end, there are these end user reports you should also take into account, even if finally the statement may be wrong : "it's because or this or of that" the result are there: the AH64 is behaving in a strange way for lot of players.
As other people reporting here, I stopped to use her some version ago because I feel in virtual danger by flying her, do to the "uncommanded roll" behaviour. As I'm playing only solo campaign and do not have so much time, I was too frustrating to fall to the ground in a middle of a mission because of these rolls that seems so strange and unpredicable.

So I moved to the Hind...That's not I'm not crashing less but at least I'm understanding the error I made .

So either it's normal to have this kind of things and then explanation should be given : "you should not fly her in this way, otherwise she will crash, that's proven and real pilot are trained to avoid it" or please try to understand and fix or explain the issue, whatever it will be. Saying it's a user subjective impression is a bit short, even if at the end you need factual elements (tracks, way to reproduce) to be factual. It's a complex balance beween taking all the misery of the world and negate end users feeling...

At the end, the horse brought to water and which won't drink will be your customer...Once you purchased a plane or helo that's seems finally bugged (from your non expert customer point of view), you may not purchase another one...

 

Edited April 22, 2023 by lefuneste01

[Boogdud]

28 minutes ago, Raptor9 said:

If changes are made to the actual flight model, we do. However, we have no control over subjective impressions posted by players.

You can bring a horse to water, but you can't make it drink.

At this point it might be easier to just say "we didn't make any changes to the flight model this patch", though honestly that probably wouldn't even help....

reminds me of a phrase my grandma always used to say: Buy 'em books, and buy 'em books... all they do is eat the pages.

[Zadren]

2 hours ago, Raptor9 said:

If changes are made to the actual flight model, we do. However, we have no control over subjective impressions posted by players. Following almost every update since release some have claimed something changed, and we have no control over that or what people choose to believe.

This is common in other areas of discussion too. Even when an SME states something factual about the AH-64 to help in understanding, there are those that choose to believe something else.

You can bring a horse to water, but you can't make it drink.

I'd happily accept that if mine was an isolated case. There's a lot of conversation out here about undesirable rolling behaviour regardless of whether you changed the flight model or not.

There has to be a reason for it, even if that reason is we're all crap pilots. Rather than just dismissing people's concerns, perhaps explain in clear, loud, unambiguous way what it is we're all doing wrong, and how to correct it.

[Raptor9]

2 hours ago, lefuneste01 said:

But at the end, there are these end user reports you should also take into account, even if finally the statement may be wrong

33 minutes ago, Zadren said:

Rather than just dismissing people's concerns, perhaps explain in clear, loud, unambiguous way what it is we're all doing wrong, and how to correct it.

The request that was made by macedk was to include flight model changes in the changelog, to which I confirmed that we do, and would. Every time that someone asks us if changes were made in the flight model, we will respond truthfully and accurately. This does not mean we are dismissing anyone's concerns or experiences, we are stating facts about what has occurred in the development. If someone claims something that is factually incorrect or inaccurate, countering that statement with our own does not constitute dismissal of their entire viewpoint or opinion, it is a correction of fact, nothing more. Some of you seem intent on turning this into something it is not.

If you read through the flight model discussion thread, you will see that no one is being dismissed; in fact @bradmick (one of our content creators and SMEs) even put together a specific instruction video to address some of these comments in recent weeks, and several of us have tried to address the issue with some of our replies.

[Toge]

 

31 minutes ago, Raptor9 said:

The request that was made by macedk was to include flight model changes in the changelog, to which I confirmed that we do, and would. Every time that someone asks us if changes were made in the flight model, we will respond truthfully and accurately. This does not mean we are dismissing anyone's concerns or experiences, we are stating facts about what has occurred in the development. If someone claims something that is factually incorrect or inaccurate, countering that statement with our own does not constitute dismissal of their entire viewpoint or opinion, it is a correction of fact, nothing more. Some of you seem intent on turning this into something it is not.

If you read through the flight model discussion thread, you will see that no one is being dismissed; in fact @bradmick (one of our content creators and SMEs) even put together a specific instruction video to address some of these comments in recent weeks, and several of us have tried to address the issue with some of our replies.

The thing you don't seem to realize is that for many people real and obvious changes did happen in the previous patch. I'm not talking about people not knowing how to take off, but all weird sudden rolls that are happening during e.g. slow, trimmed forward flight, that certainly didn't happen before the previous patch (not the latest one), or rolling related to collective changes (which is logical, as you explained, but still different than before the patch). We believe you didn't intend these changes and did not change the "core flight model" that you're so stuck on, but you yourself said that the FMC collective channel and center of mass did change in that patch. Why is it so hard to believe that perhaps these or some other unintended (or internally undocumented/uncommunicated?) change did cause this behavior for all these people, even if you don't see it in your own testing? I'd understand your point if you really only changed e.g. some textures or pilot models or such... Even Casmo said in his video a short while back that the apache behaved differently for him after the patch, that he found it doing unexpected things when he tried to repeat what he had done before. I'm inclined to think he knows a thing or two about helicopters, so it's not because he doesn't know how to fly it.

3 hours ago, Raptor9 said:

You can bring a horse to water, but you can't make it drink.

So here you're basically calling your customers stupid, that's just great. Really makes me want to pay you guys more money.

[Floyd1212]

26 minutes ago, Toge said:

weird sudden rolls that are happening during e.g. slow, trimmed forward flight

Do you have any examples of this?  I don't think I have seen anyone saying they are just flying forward in a straight line and the helo suddenly rolls on them for no reason.

[Floyd1212]

Here is an example of what can induce an uncommanded snap roll.  This is being way to aggressive on the inputs while the aircraft is in an unbalanced state.

I'm not a SME, so I can't say if this behavior is accurate, but with the current way the Apache is modeled in DCS, this will get you into trouble every time, so don't do it.

Notice I have quite a bit of left stick, trying to bank left toward the "target", but the other forces at play easily overcome my left stick and I end up making a full 360 roll to the right.

This is the track from the above clip as well.

Induced_Roll.trk

This can also happen when aggressively side-slipping while trying to keep the nose pointed on-target.  For example, I'm in a hover and my CPG has a missile in the air on a T90, and the T90 fires a sniper at me.  If I aggressively slip to one side or the other to dodge the sniper, while trying to keep the nose on target, the aircraft can easily be upset and roll over.

[Raptor9]

2 hours ago, Toge said:

The thing you don't seem to realize is that for many people real and obvious changes did happen in the previous patch. I'm not talking about people not knowing how to take off, but all weird sudden rolls that are happening during e.g. slow, trimmed forward flight, that certainly didn't happen before the previous patch (not the latest one)

I understand perfectly. You are diverting and misconstruing factual statements that have been made by myself and other members of the ED team about what went into the recent patches. We are stating what occurred in the update, we are not saying people are not experiencing what they are experiencing. These are two separate concepts, of which we have stated facts about the former, but you are inferring something else entirely about the latter. The facts will not change, nor is it cause for such emotional responses on the matter.

2 hours ago, Toge said:

So here you're basically calling your customers stupid, that's just great.

I said nothing of the sort. Absurd and unreasonable accusations like this are attempts to gaslight the conversation. If you cannot have a reasonable conversation on the matter and what was said, there is no point in continuing this discussion.

[NineLine]

Dear all, as stated the FM is still WIP, I have tagged the thread as such. Thanks for the report but we have commented and will close this now.