RuecanOnRails

http://forums.bistudio.com/showthread.php?126795-flight-model-editing/page2&p=2054734#post2054734 An old thread for TOH which brought up possible cause and solution to "input lag". Could something similar to "damper" be the issue in Arma 3, a tweak that got forgotten, never added, or maybe obsolete? I have no idea how relevant that thread is today, but it seemed interesting when I stumbled upon it.

Last couple updates (VRS) the xH-9 seems to have no weight at all. The slightest increase in collective and the helicopter jumps off the ground. Flying upside down is not only possible, it's easy and stable with little altitude loss. Lowering altitude for a landing and bleeding off speed is difficult due to being too "floaty" and able to climb with zero collective with low angles. Helicopter flight feels closer to fixed wing than rotorcraft with current amount of lift generated from speed. I'm not a pilot and have no real world experience, for all I know everything is behaving exactly as it should. I will also mention that I too suffer from a lack of engine on/off control mapping option. I have controls for all other newly added functions, from wheel breaks to slight loading but nothing for engine. It's not an issue for myself, as the only need for it I have is auto-rotations, pressing space twice works well enough. During flight, with max wind, rain/storms, the flight is still rather stable. Would it be possible to see a sort of random air temperature/density map layer generated at the start of map loading to cause random, but somewhat persistent pockets of turbulence? Something to make long cross map trips require more than trim and forget. Using some sort of data layer would allow a seamless multiplayer experience where flying in formation has the same effect on all players, not semi random wind gusts seen only by individuals. If there is already something like this in place, perhaps make its effect on helicopter flight much more prominent?

There really needs to be a higher multiplier to damage per second when excessive stress is continuously applied. For example each consecutive second of applying over 90% torque increases the multiplier by 1.65x damage per second. If the current is 1 damage per second up to 300, applying stress for 5 seconds continuously would result in 1.65^5 = 12.23 damage done, increasing exponentially, 10 consecutive seconds would do 149.56 damage and 12 would cause failure. This would allow many short high torque maneuvers, but very few long high torque maneuvers. Another option would be increase multiplier higher but place an upper cap on total damage it can do per second. My opinion would be to define how many excessively high torque maneuvers the helicopter should be able to perform before damage occurs. Then find the average time for how long the torque is above 90% during those maneuvers and calculate the multiplier around that. There's also the option to set different multipliers based on the percentage over 90% torque. 95% torque could do 5x damage per second, and 100% torque 10x damage per second. This is just a suggestion, and my math could easily be wrong. But something about the current damage model still doesn't feel right. Loading up the editor and flying as recklessly and hard as possible pushing torque to max, destroying the littlebird takes too long. It Feels closer to being on a timer and less like it's influenced by high torque maneuvers. Side note; it's still impossible to land inside the hangar at the central Altis airstrip. Doing so causes instant damage and death.

I really like the quick break down on what's simulated, knowing the thresholds will make flying within them easier. Now that torque related damage is being added, will we see working torque (and other) gauges like we had in TKOH? Flying without a hud gives no indication of torque, and with the hud enabled a static white/yellow/red light doesn't give enough information to make adjustments. I'd assumed these features were waiting on stress damage to be implemented. I would really like to be able to fly, maintain a heading, speed, and altitude using cockpit instruments alone in the advanced FDM, a near hud-less experience. Little bit away from the FDM discussion, but an option to changing the HUD altitude from above ground level to above sea level would be nice for formation flying. Things become complicated with an ever changing ground level while maintaining a level flight path.

After reading the announcement, the lack of a single player campaign/story driven mission set has me a little disappointed. I was really hoping to see the return and rise of Larkin Aviation on Altis, I found the civilian flying missions relaxing and fun. Time trials and VR "training" are fine, but they play more like minigames than full missions. Hopefully it will be possible to port TOH content including semi-functional campaigns/missions into Arma 3 to at least replay those with the updated RotorLib. To incentivize purchasing of the DLC, there needs to be more paid specific content than 4 time trials and some achievements.

Forget free, I'd be more than happy to pay a premium for that extra content. Adding more authentic and simulator based content costs time and money which needs to be recouped with additional sales and support. Content like that is something I would pre-order in a sort of crowd funding campaign. Given the option I'm sure others in the community would be eager to show support as well. I'd think the likely success the Helicopter DLC sales and community support will be weighed in on future decisions to add other simulator like content. Even if the chances are slim, I can always hope. I am patient when it comes to these things, I'll ask this again in January 2016 to see where we're at :)

Seeing as RTDynamics, the company behind RotorLib also has a product called FixedWingLib is there any chance of seeing a possible future DLC or expansion making use of this?

Slingloading does seem to be a little glitchy this time around. @KeyCat I don't mind the wind noise, at first it bothered me but I quickly got used to it. Remember the MH-9 doesn't have doors, the wind should be pretty loud. I really enjoy the creaking sounds even though as you said, it's overplayed even when sitting with engines off. It makes it almost a little nerve racking as it sounds like something is about to fall apart, I'm anxious to see how the new creaking sound will play into the damage and stress models. Flying with no HUD and judging damage based solely on sound and responsiveness of controls will definitely make it more interesting. The creaks and rough sounds should occur during engine start up, warm up, and anytime you put stress on the helicopter during flight.

With a focus on flight, will we see improvements to night flying? Currently flying at night is really only possible with night vision (or gamma/brightness increase), many large structures lack aircraft warning lights (some models have them but they don't work). This isn't too friendly towards civilian pilot missions at night/dusk. The large Altis airport at night looks absolutely beautiful while viewing from a distance or landing. It would be nice to get similar lighting effects around the other airstrips and some small residential lighting around buildings and structures. Three quick suggestions from things I recently flew into at night, powerline transmission towers, large military structures, and wind farm turbines. It would be a great to see with warning lights. Obviously I was flying fairly low which I'm sure is part of the problem ;) With the heavy shaking from overspeed, will we eventually see this cause stress damage from pushing the helicopter into speeds beyond "safe limits"? It still needs to be tweaked as the vibration shaking begins at too low of a speed around 95-100 knots, below the rated cruising speed of 125 knots? But it's nice to see some negative effect for trying to push 150+ knots.

I just loaded up TKOH to do some quick testing... What I found is the civilian light heli's are nearly indestructible in terms of engine damage. 100 rounds from a machine gunner into the back/engine location did next to nothing (200 rounds and it blew up). It wasn't until I used another littlebird with mounted weapons before I was able to get a result... The heli bellowed smoke after a dozen or two rounds but still flew without much issue, and the instrument panel showed the engine was fine. There doesn't appear to be any bullet penetration into the engine block, instead all damage goes towards the overall hull. I don't think it's a direct port. A balance needs to be found between TKOH's and Arma 3's littlebird damage system. In Arma 3 the hummingbirds are currently far to fragile to be considered for transport remotely near a firefight. Even taking pot shots at one from a distance can send it falling to the ground. Given the option, more times than not, I'd increase the weight by reinforcing the hull around the engine because a 9mm should not be a one hit kill. Off the top of my head, my solution could be something along the lines of: engine and internal instruments can not be damaged until the hull is yellow, or red. This thread is for the FDM so I don't think we should continue discussion of this here, it was more of an after thought to my previous post.

The latest update with 4.1.1 made the helicopters (IMO) much more responsive and generally felt much smoother and more enjoyable to fly. The helicopters seem to react much closer to my inputs leading to less over-correcting. Or this could all be a placebo based on seeing an update ;) While playing as zeus, when using the remote control on AI the advanced FDM does not work regardless of game settings. Another small issue not directly related to the FDM, now that I'm spending more time with the helicopters (littlebird in particular) it's armor rating is far too low. A single round from a 9mm can effectively, and reliably disable the aircraft. To reproduce, shoot the back passenger seat towards the engine, or the hull near the engine. This will instantly turn the engine to yellow or red, both of which render the littlebird unable to take off. This leads quite often to falling out of the sky from a single stray round from any weapon.

You're only half right, but mostly wrong. Arma 3 directly can only use 2-3gb of ram due to 32bit limitations, however it makes great use of external memory allocators to boost usable ram cache to a max of 32768mb. This is a great workaround for 32bit limitations, native 64bit would be better, but for the task at hand this works just as well. The OP's system only has 16gb of ram, my suggestion of 14336 will allow a lot of headroom but prevent low virtual memory errors from the OS. I know how crossfire and sli work, the total available memory is mirrored and not shared. By setting maxvram to 8192 Arma 3 will use the full 4096mb available by crossfire and any overrun gets stored in ram for quick swaps, faster lods, and generally less stutter. One of the main killers in terms of performance is I/O and how data is streamed instead of loaded directly into memory. Utilizing the high transfer speeds of ram is essential in getting smooth playing experience.

I'm not sure how it's calculated, or why it's giving different results each run. It's likely calculated during launch with the quick benchmark/hardware test. Try running -nobenchmark and see if it still changes. Would making the arma3.cfg read-only be a temporary solution? Are you able to make a repeatable mission or set up that others can confirm and test?

Try fred41's registry change and custom experimental memory allocator. https://github.com/fred41 When running the game, keep as little running in the background as possible. Try to do a reboot before playing if necessary. As for your starting parameters, try to use the new launcher and set cpucount 8, exthreads 7, check hyperthreading, malloc tbbmalloc (if installed), maxmem 14336, maxvram 8192, nologs. Ensure you have adequate pagefile as these settings will make use of pagefile swaps for quicker lod switching when it exceeds available hardware memory. If you have an SSD, you will notice a performance boost when Arma 3 is installed on it. If you have limited room on the SSD, you can monitor various missions and using symbolic links move over some of the files with the highest i/o.

The "input lag" while a little extreme I believe is part of the way they simulate control system effect and phase lag. Of course I could be completely wrong too. Try the demo of x-plane, dcs huey, fsx, or other flight sims, they all simulate phase lag differently, but they generally results in delayed angular movement from control input. I would like to see some G-Force induced head/camera movements when flying, a static model feels weird and out of place. A slight head movement during rolls could add a lot of immersion.