RealDCSpilot

The Brunner base is targeted for the custom sim professional market. It still doesn't support directinput FFB, only gets it's data from DCS's telemetry output (for motion seats and forcefeel mats). Trim force has to be emulated by it's own software. However, DCS is telling a lot of haptics and other information to us, but very few of us have all the hardware to listen. Those haptic forcefeel mats and simshakers for aviators (software) for instance improve immersion a lot by giving info about airframe vibration directly to your body. You'll instantly know if a stall is imminent, extremely helpful in a helicopter.

That's what i did to get the most out of a Logitech G940 base. It has pretty strong motors (but they need a little help from oil dampers to tackle all the weight). Best helicopter flying experience so far, but i really hope that someone comes with a new FFB base to the market sooner or later.

@PickleMonster Hello! Test, check... :D

It's so simple to test the FM, hold your stick a couple of cm left or right when in air and watch what happens. This kind of behaviour is absolutely impossible with a helicopter, there is no mechanical or aerodynamic or SAS driven way to do a continuous roll with just a bit of stick deflection. It's also against everything how an engineer would design a helicopter. A real pilot said "NO, IT DOES NOT FLIP ON THE BACK WHEN TILTING THE CYCLIC TO GET ABOUT 20 DEGREES OF ROLL AND THAT WE HOLD THE CYCLIC." (translated from french, link is below) In real life you hold and move the stick all the time during maneuvering, there is no self centering and automatic straight leveling. The over-sensivity is also completely wrong, look how much stick movement is needed for just keeping her steady: Here is what really was going on with french ALAT Gazelle pilots: TLDR: Their first impression was good, they saw the cockpit and felt at home. Polychop took this for marketing the "tested and approved by real pilots" claim. After testing the FM more in detail the same pilots even recommend not to buy the DCS module. "A colleague from ALAT advised me against taking the Gazelle (rightly so). Qd I tested it, my first impression was a very positive shock, the cockpit and the sounds are so realistic. I had this "Wow" effect that all Gazelle IRL pilots have when they discover this module. And it is completely understandable that ALAT pilots gave a positive opinion on this Polychop gazelle when it was released, for 2 reasons: -they had a "Wow" effect, and as long as it is not simmers, testing this gazelle with a Hotas, on a 2d screen etc .... Well, in the end it is not easy to do this direct analogy with reality. Personally, it took me 2 good hours of solo testing (knowing DCS and my Hotas very well, as well as the axis curve settings) to manage to write precisely the FM problems." "Today I recommend instead the Huey and the MI8 which, apart from the Vortex, are realistic." Sources: http://www.checksix-forums.com/viewtopic.php?f=465&t=206209&p=1742308&hilit=inertie#p1742308 http://www.checksix-forums.com/viewtopic.php?f=465&t=206209&p=1742308&hilit=inertie#p1742326 "Some guys tell that ALAT pilots were happy with the FM when testing the Gazelle on DCS. It's normal because when you test a Sim for the first time when you are also a real pilot, you miss the real sensation of flight that you are used to. And you pilot the Sim in response to the instant behavior of the module. I made the same mistake during the first 2 hours of flight with the Gazelle. I was like a child when I saw the cockpit for the first time, it was like 15 years ago when I learnt to fly in ALAT school at Dax. But passing the first 2 hours trying to apply curves on axes to have a realistic feeling, I never succeeded to reach a realistic FM. Then I searched for the mistakes and I have discovered some. After that I performed a test flight with Polychop guys and now we are waiting for the new FM. Perhaps one day." Source: https://forums.eagle.ru/showpost.php?p=3929939&postcount=44

@DflippinK a bit late to the party, but dropping GPU load on purpose is not a good thing. You want your GPU usage always at the highest level as possible. It says "i'm making the most out of it.." in a CPU bound scenario like DCS. In your case, 60% could mean that there is enough headroom to not downgrade visuals with the shader mod and even maybe choosing higher options for better visuals in graphics settings OR your CPU is too weak to deliver data fast enough to the GPU.

When i get my RTX 3090 i will be able to get to the core of this. If DCS has a problem with it's garbage collector routine or doesn't have a routine to clean up unused assets, then it will show up as ridiculous VRAM usage with a 24GB card. Right now it's easy to fill up 11GB of a 2080ti, just happened to me again yesterday an a MP server, when i switched from a A-10 to the Ka-50. In my case i was lucky and after 5-6 minutes the normal FPS came back. I think ED has to test these scenarios more, because the problem is present for years.

Brunner CLS-E's problem is the lacking Direct Input FFB support. They have to use telemetry data from DCS which is missing several effects. However a good old MS Sidewinder or a Logitech G940 is perfectly working in DCS (except the Gazelle). You have FFB for trim forces (trimming moves the stick) and extras like stalling or runway bumps. The absolute best are helicopters, here you have fantastic force trim support - makes DCS the best Heli sim out there. Playing DCS without an FFB stick is like playing a racing sim without FFB on the wheel. You are missing a lot, sadly... It's about time for a new gen of FFB sticks...

The thread starter described exactly a problem that seems only to occur in VR in multiplayer sessions. After a while something unknown triggers a performance hit that stays forever and needs a restart or goes away sometimes after a while (5 or more minutes). There is no external process involved, but it might be provoked through loading several different DCS modules during a session.

I also suspect a VRAM problem more than anything else. My 2080ti's 11GB get filled up in MP very quickly after switching to different modules. There seems to be no efficient cleanup process for unused textures and geometry. When i get a 3090 i will see what's really going on there, will DCS eat up all the 24 GB? :joystick:

Atm the mainrotor doesn't produce simulated air pressure for lift. The collective is just an anti-gravity lever, so the skids on ground behave magnetized as long as their collision detection says "true" for contact with ground object. As soon as you use the anti-gravity lever, it shoots the gazelle-thingy straight up and the skids say "false" for any ground contact and completely ignore that there ever was a ground object.

Despite the bad description of the problem, it is a fact that especially in MP something like a memory leak appears and causes a huge drop in fps. It may be triggered by choosing a new plane or different type of plane or by another player entering the server and choosing a plane that wasn't choosed before in the session. But that is only me guessing, DCS is still not handling memory very well (video memory and RAM). The thing is, sometimes the problem goes away after a few minutes, but it sucks however. The only thing that seems to mitigate this, is to lower the visibility range (lowering memory usage). But that's all guessing... What i don't get is why the "extreme" option exists, because there is no hardware that can tackle it (unless you set everything else to extremely low and ugly).

A much more efficient method would be support for Variable Rate Shading for supersampling in DCS. This MSAA mask feature is kinda neat, but would have much more potential in combination with VRSS.

I have a Valve Index and i use Steam VR for custom SS for DCS. I pushed it up to 2880x3200 pixels per eye at around 90% GPU usage. I don't know enough about oculus, i think you have to use the tray tool. Disable MSAA, SSLR, SSAA and motion blur. Set terrain shadows to flat and eyeball how your fps change.

In VR, DCS is running better than ever on my system at very high settings. From what i see in your settings, a PD value of 1.5 is quite high (DCS internal SS is very costly!) - it's better to use external SS settings and leave PD on 1.0. Use Oculus settings to raise SS for DCS, this shifts the SS workload over to your 2080ti which is hungry for work while running DCS's CPU bound tasks. The most actual driver optimization comes from this guide: Make sure your CPU runs DCS at high clock speeds 4.6+ GHz without any energy saving options. Don't use VR shader mod, ignore older optimization guides - DCS constantly changes, we have to keep up with it.

Hmmm, so the real Gazelle's stick does have a huge dead zone and as soon as you are moving a couple millimeters out of it and hold there it goes into a continuous roll?? Why does it fly like by-wire instead of direct stick to main rotor? And the real collective can be moved like crazy fast without friction for instant jump starts? Why does everyone have to tinker extremely with control settings to make it even a bit flyable, while other DCS modules just work out of the box?

Hi, just got a realtus forcefeel and tested SSA with DCS. Feels great! Adds another level to VR flying. I couldn't find a way to enable full support for all 8 motors of the forcefeel, did i miss anything? I can't use the Forcefeel Control Center app when SSA is running, it has this little checkbox (6 to 8 conversion).

The Rift OLED panels are clamped in bandwidth. Because switching times of OLED cells from OFF (full black) to ON where not fast enough, they decided to let them not switch completely OFF. Instead they just go into a very low energy state that looks like nearly black. However, VR HMD with OLED's aren't per se better regarding Blacks or color space, they also suffer from "black smear" "black crush" which shows as less detailed display of detail near black or dark scenes in general. I've seen that in several VR titles, where my old VivePro just showed pitch-black-nothing and my Index now still shows details and objects that i couldn't see before. The 100% perfect panel tech still doesn't exist.

Great post, thanks!

I upgraded my frankenstick with smooth low friction oil dampers. They compensate the weight of the heavier grip and counteract the mechanical deadzone of the G940 base. https://www.thingiverse.com/thing:3849592

The Gazelle is DCS's worst helicopter module, i would recommend to start with the Huey or Ka-50. Polychop did nothing to fix this since release in 2016, all we get are excuses like: https://forums.eagle.ru/showpost.php?p=4373660&postcount=51

I had to do this, FFB is so essential :) https://www.thingiverse.com/thing:3849592 The base delivers the FFB and X+Y axis, the grip delivers buttons buttons buttons... @fat creason Afte the latest OB update i have a bug with FFB switching off in the F14 in MP. If i change from another aircraft to the F14, FFB gets disabled :(

I checked the cockpit in modelviewer, it's not the shader maps... Must be something with the lighting or the 3d model that makes the gi lighting go nuts.

I'm on 445.98, don't have the issue...

+1

Thanks for looking into it. I had the problem in MP, don't know if it also occurs in SP. The tail wheel always tries to turn itself facing backwards, like on a rubber band.