Advanced Helicopter FDM Feedback

[x3kj 1247]

One slightly unrelated question: Are the Dials you get with AFM a regular Rsc display? If so, where can we find them to make our own displays for different vehicles?

[Garuda 10]

Hello, I am new to the BI forum. After spending a good amount of time using the AFM, I have decided to register to give some feedback of my own. I've read a lot of this thread, but not all of it, so I am sorry if this is simply rehashing topics that have already been brought up.

In its current state, I find the AFM to be very good! I use saitek x-55, saitek pedals, and track ir to control the helicopters. However, there are two things that I have noticed that I find strange.

1. The "power" gauge seems to measure how much collective pitch is used rather than torque percent. This is perfectly fine, as there are warning lights which alert you when you about to over torque. However, the amount of collective pitch needed to induce an over torque situation seems to change on the load of the helicopter. The MH-9, for example, can use about 95% collective pitch (as per the gauge) before the orange torque light illuminates if only the pilot and co-pilot are inside. Once the helicopter is fully loaded, a collective pitch closer to 80% causes the orange torque light to illuminate.

Understandably, the heavier the load of the helicopter the more collective pitch will be required for it to fly. The weight of the helicopter will not change the effectiveness of the engine to spin the rotors though. Under the same conditions, a minimally loaded helicopter should begin to over torque at the same collective pitch as a heavily loaded helicopter.

2. The torque effect that the tail rotors are designed to counter does not seem to be related to torque in the game, but rather the collective pitch setting. When the engine is running, raising the collective causes the helicopter to yaw, which is good. However, if you turn the engine off, raising the collective will cause the same amount of yaw even though the engine is no longer providing torque. I was able to demonstrate this on all single main rotor helicopters.

Bringing the MH-9 to an approximate 6 foot hover, I turned to engine off and made no control inputs. The helicopter descended straight down without any yaw. Realistically, the helicopter should begin to yaw to the left once the engine is turned off, required right pedal input to maintain heading. On second test, I turned the engine off during the same 6 foot hover, only this time I slowly raised the collective in order to slow the descent rate for a softer landing. As soon as I began raising the collective, the helicopter began to yaw to the right.

Ideally, turning the engine off in a hover should cause the MH-9 to yaw to the left (assuming the helicopter was holding a constant heading beforehand), requiring right pedal input to stop. Raising the collective should not cause any yaw to occur. I find this to be a problem mainly on larger helicopters when performing auto rotation landings, as raising the collective to cushion the landing at the end of the flare will cause the helicopter to start yawing to the right. This makes it more likely to roll over and explode during touchdown.

[Gruman 123]

Question to BIS:

Is there a way to define a GroundContact in the XML as a long flat plane? Because when you are defining ground contacts for a MD500 with skids, you would have to have so many points in between to cover the whole skid. Otherwise you will likely clip into objects and explode if you try to land on them...

The A/MH-9 just has 4 GroudContacts defined, but the rest of the skids will sink into objects when the landing area is not flat. And eventually blow up because of the "collision" with the airframe. (One example)

Is there a solution other than using a x-number of groundcontatcs until the whole skid is covered? The FGGroundContact200 Model looks more intended for wheels?!

Thank you

[Gruman 123]

Doublepost, but a month apart...

Is there a bug report about the Input Lag on the Tracker? I wasnt able to find one...

[oukej 2911]

FYI - there was a little tweak to the interpolation / input lag recently (now it should be consistent 1/15s).

27-08-2015

EXE rev. 132000 (game)

Fixed: Input problems after Time Trials restarting

[Airwolf 50]

I've been doing a lot of crash landings lately with the AFM, and my biggest issue is that a succesfull auto rotation is often ruined by the helicopter flipping over and exploding. Helicopters should not bounce when they hit the ground, or flip over so often. Even if they do flip over they shouldn't just explode.

 

Please make it so the helicopters do not explode when they roll. The pilot should be able to get out with minor injuries.

[en3x 209]

that was asked on livestream when afm was showcased.They know about, happens when geometry collides with a ground,

deep rooted not fixxable in short term I believe.

[Airwolf 50]

Then the weight of the helicopters should be increased so they don't bounce and roll so often. I feel that they are far too lightweight.

[en3x 209]

Then the weight of the helicopters should be increased so they don't bounce and roll so often. I feel that they are far too lightweight.

Why does explosions happen in first place?According to pettka

engine computes force necessary to put something off the geometry and if it is too high, the object simply explodes.

Problem is deeper.

Increasing weight would mess with sling loading and would be band aid on issue and cause only more issues down the road.

Issue needs to adressed at its core ;)

[pettka 694]

Why does explosions happen in first place?According to pettka

engine computes force necessary to put something off the geometry and if it is too high, the object simply explodes.

Problem is deeper.

Increasing weight would mess with sling loading and would be band aid on issue and cause only more issues down the road.

Issue needs to adressed at its core ;)

Just to be sure, this is a simplified reason, it depends on many other aspects. But generally speaking, this is the most common issue - as engine calculates states vehicles each frame, it's kind of a quantum physics - the helicopter may be above ground in one frame and under it in the next, that depends on time between frames. And as the helicopter "appears" under the ground, engine calculates force necessary to push it out (as a inverse force to force necessary to put it that deep). You may possibly imagine that results may be extreme with low fps.

[Airwolf 50]

Pettka, is this something that is on the radar to fix or will heli's always explode when they roll, and the rolling will continue?

[machineabuse 11]

Just to be sure, this is a simplified reason, it depends on many other aspects. But generally speaking, this is the most common issue - as engine calculates states vehicles each frame, it's kind of a quantum physics - the helicopter may be above ground in one frame and under it in the next, that depends on time between frames. And as the helicopter "appears" under the ground, engine calculates force necessary to push it out (as a inverse force to force necessary to put it that deep). You may possibly imagine that results may be extreme with low fps.

This reason made me both sad and at the same time the mental image of being violently rejected by Armaverse earth made me laugh  :) 

I don't quite understand how most land vehicles are able to flip and roll at will while aerial vehicles need to explode though.

[CaptainAzimuth 714]

Then the weight of the helicopters should be increased so they don't bounce and roll so often. I feel that they are far too lightweight.

Erm... Helicopters are generally light weight to begin with. When you start getting into medium and heavies, of course their weight won't be an issue. But making light vehicles more heavy just for a work around is by no means the solution.

[breech99 47]

Just to be sure, this is a simplified reason, it depends on many other aspects. But generally speaking, this is the most common issue - as engine calculates states vehicles each frame, it's kind of a quantum physics - the helicopter may be above ground in one frame and under it in the next, that depends on time between frames. And as the helicopter "appears" under the ground, engine calculates force necessary to push it out (as a inverse force to force necessary to put it that deep). You may possibly imagine that results may be extreme with low fps.

The way the system is now is absolute crap.  Helicopters in Arma3 need to have some sort of damage system that reflect using bladder tanks and not just a simple hit point, health bar equals zero then explode kind of crap every time a simple tip over or when red. There must be something they can focus on with having explosions when rockets impact to simply forced landings, crashes that don't involve explosions every single time. This is a huge negative in the game that is turning me and a lot of my friends off from this game. Years ago for example, screamin demons over europe used a break away system and games like mechwarrior that have a damage model that transfers damage inwards to a core with break away points like you already have with tail rotor failure and main rotor.  Keep working on this issue. Don't simply say oh well nothing to be done. You made it.  Try to make this game 64 bit and improve the damage models or implement some better damage system. It sucks.                        

 

 

No more half measures and no more bullshit excuses.

[twisted 128]

Pettka, is this something that is on the radar to fix or will heli's always explode when they roll, and the rolling will continue?

 

this tendency to roll and explode has caused me to stop playing with advanced flight model.

[MadocComadrin 12]

You made it.  Try to make this game 64 bit and improve the damage models or implement some better damage system. It sucks.                        

 

 

No more half measures and no more bullshit excuses.

 

"You made it." The game split a long time ago from VBS. Parts of the engine are very much arcanely mysterious for BIS staff. You ask for no more "half-measures" or "bullshit excuses," but the things you ask for require not only major research into the engine itself (to determine plausibility), but major refactoring of the engine itself. It's not possible for ArmA 3 because the effort that many of the "major refactoring" requests that people have would be the same or more than making ArmA 4, if not ArmA 5 (especially the 64 bit part, which doesn't automatically guarantee increased performance or fidelity, especially when naively implemented).

 

I'm not saying they shouldn't try to fix these things. I would love to see some heavy engine loving between ArmA 3 and ArmA 4, even if it means waiting longer than normal for it.

[oukej 2911]

Still, maybe scroll wheel + key would be a nicer way to adjust collective and the speed of adjustment could be controlled by how fast the user scrolls it. I know it's still digital so the same could be done by allowing the user to control the speed of adjustment by tapping a key fast or slow but scroll wheel feels more comfortable and then the keys could be used for smaller adjustments when needed.

As a reaction to this older post and http://feedback.arma3.com/view.php?id=21189I've created a simple test scenario with slightly different behavior of the collective for digital input (linear with bigger increments). Try it out here http://steamcommunity.com/sharedfiles/filedetails/?id=532066985

[St. Jimmy 272]

As a reaction to this older post and http://feedback.arma3.com/view.php?id=21189I've created a simple test scenario with slightly different behavior of the collective for digital input (linear with bigger increments). Try it out here http://steamcommunity.com/sharedfiles/filedetails/?id=532066985

Cool, I haven't heard about this. Need to test it later today.

[CaptainAzimuth 714]

Big problem with Advance Flight Model and PhysX Lod.

http://feedback.arma3.com/view.php?id=26186

 

EDIT: Also, i noticed that Vanilla helicopters using AFM have a terrible Taxing capability. Slight turns flip the heli, and send you into a fireball. You need to tweak that, because Taxing in heli's is a big part of the Advanced Flight Model.

[Gruman 123]

Wow, since BIS stopped any work on the AFM (IMHO still needs a lot of work), this topic got real silent.

Are there still other pilots relying only on AFM?

 

While running AFM, I always notice a heavy drop in frames. Is this also observed by other pilots?

 

Any recommendations for settings / optimisations?

 

Can we hope for any inprovements from BIS in that matter?

 

Thank you

[SSG-Levi 7]

not only still using the AFM but almost exclusively (SFM only used with non supported choppers from mods).

at least with the small-medium choppers (xH9,xH6 MELB and UH80) the AFM is awesome (the AAF helos and the CUP MH53E also works good). 
the huron is a little too easily going into VRS so it requires a more stable flying style.

 

the AFM makes you "feel" the weight of your cargo/passengers, weather and RPM, and overall makes flying much more interesting from the SFM which allows for insane maneuvers and perfect hovers. 

 

as for taxing and rolling over-
taxi your chopper slower. you can't expect a narrow chopper with a tail wheel to handle a 180deg turn in 30kmh - especially with ARMA lower gravity value (at least with a stick with a rudder capability or pedals - haven't used keyboard since OFP or early ARMA1).

rolling over is not an AFM issue (SFM choppers also explode and bounce on impact) and AFM at least lets you make a more realistic autorotation.

there's no need to disregard people who worked/s a ton of hours on those features, there's no need to talk to them like they leave those issues out of laziness.

 

[Lobotomy Lobster 6]

As a reaction to this older post and http://feedback.arma3.com/view.php?id=21189I've created a simple test scenario with slightly different behavior of the collective for digital input (linear with bigger increments). Try it out here http://steamcommunity.com/sharedfiles/filedetails/?id=532066985

 

Tried this for an hour or so and its a real improvement over the way the AFM collective currently works. You can do precise but meaningful adjustments by tapping the key and you can make major adjustments much more quickly. Its amazing how much more responsive and fun flying the choppers is with this. It almost seems easier flying a Mohawk with it on than flying a Littlebird with it off.

 

If at all possible it would be really good if this could be the normal behaviour for the collective from now on.

[ruPal 143]

I was thinking about helicopter explosion when helicopter rolls on ground. What if you add more ground contact points not only for skids or wheels but for body part too. Will this solve that problem?

[Kryptongame 14]

I've been doing a lot of crash landings lately with the AFM, and my biggest issue is that a succesfull auto rotation is often ruined by the helicopter flipping over and exploding. Helicopters should not bounce when they hit the ground, or flip over so often. Even if they do flip over they shouldn't just explode.

 

Please make it so the helicopters do not explode when they roll. The pilot should be able to get out with minor injuries.

Please consider this! Still very annoying to explode when flipped?

[Gruman 123]

It would be amazing if they fix it. But as they stated in the past, an advanced damage model is not planned. Which I basicly translate into: The AFM Project is finished and stays as it is. Probably not suitable for the casual gamer approach BIS has currently on their game plan.