EbonySeraphim

Clarify? I just ran the multithreaded binary (bin-mt/DCS.exe), logged onto my usual ThroughTheInferno MP servers (tried all three Syria, Persia, and Caucus) and every one of them saw the massive performance increase. From what I can see with some of the follow on discussion it seems there may be some interesting tools or "optimizations" some people might be forgetting applies to their system. I don't use particular tools myself and would recommend letting Windows and your UEFI/BIOS setting do their thing: Verify all of your cores are active through your UEFI/BIOS Keep Hyper-Threading / SMT setting enabled (unless you have 33 or more "threads"). Ensure Windows doesn't have any "CPU hardware management" software that tries to impact or configure the above settings through Windows. I can't think of a worse idea than user installed software impacting how the OS schedules work to be done on the CPU -- if it is even possible. Unless you really need it, I would uninstall such software yesterday. This is ambiguous: but is all core performance better than single core highest performance? Only ED can answer this effectively. For AMD overclockers this is PBO curve settings; I don't know what it is called on Intel or if all of the P-cores are all basically the same. At the end of the day, there could still be a CPU heavy "main thread" that wants the fastest core and the question is if the engine can find it and use it, or if Windows detects this and schedules it there appropriately. If neither the engine can find it, and Windows doesn't schedule the most demanding thread on your fastest core, then you're probably better off with an overall higher all core OC rather than a single fastest core. The reason being is thermals and power. Otherwise, the reverse is probably better if you see "CPU bound" enough popping up when the performance meter is on (Right CTRL + Pause). Disable any gaming software that claims to boost your performance. It's most certainly doing the exact opposite. For quick reference my setup: Specs 5950x, 4090 (Gigabyte Gaming OC); Gigabyte Xtreme mobo; 64GB 3200 RAM (4x sticks, makes a difference how many channels you use) Running on a 4k @ 60hz monitor; I have 2xCubesim monitors, and 2nd 1440p monitor which renders the AMPCD of the F/A-18 Running on up to date Windows 11; GeForce GameReady driver 531.26 Important: if for some reason you have the "Studio" drivers installed for an nVidia card, replace it yesterday because that driver runs game substantially slower than "Game Ready" drivers. Unique: In the nVidia control panel I changed DCS specifically to Vsync "fast" option; and in-game DCS I turn Vsync on. What this does is prevent frame rate dips that are just below 60hz from being penalized all the way down to 45fps or 30fps because it misses the monitor sync signal and waits for the next. The GPU driver will always let the (old school Direct3D::Present()) call return immediately even if the application asked to wait; and present the rendered frame when it does see the sync signal. UEFI: Hyper-threading on UEFI: PBO settings are automatically optimized based on my motherboard's opinion; I don't have curves set and don't know which one is the fastest UEFI: XMP profile enabled. Double check that this setting didn't turn off as it sometimes can own it's own; otherwise, your RAM is running at half it's potential speed. UEFI: Virtualization on (I say this because some people think it improves performance to turn it off, it doesn't and I use WSL2 often) I deleted my Saved Games / DCS.openbeta / fxo and metashaders2 folders. My DCS settings are essentially maxed except the settings that are about volume rather than quality like civ traffic Depth of field off Motion blur off Fullscreen is unchecked Vsync is off I actually source control my DCS config files here This patch made me appreciate that I went with a 5950x even more (I do other stuff). I'm curious if a 5800X3D still stacks up favorably in terms of performance gains with this patch. EDIT: clarified the labeling for Game Ready nVidia drivers preference over Studio driver.

Amazing job with this. It's not perfect (I'm getting UI issues), but they are minor and I'll try to have detailed reports if I don't see that they already exist. I'm a software engineer, ex hobby gamedev, and would LOVE LOVE LOVE to see a detailed blog or interview with the development team about the challenges, surprises, and overall development process of bringing multithreading to DCS. Players would love it; I would love as it as an engineer; and the rest of the game dev community will want to listen in on how a "game" improved performance by this much for just about all PC hardware. Though DCS isn't super well known to gamers at large, I get the feeling that news of this level of improvement will cause ripples in the general gaming community as well. Companies like EA/DICE will be wondering why suddenly players are expecting Battlefield to better exploit their 16+ core rigs for better performance. Most games do not have demanding simulation/computation needs and the devs know it, but there will be curiosity if they do something better and achieve even a fraction of the performance boost.

Inaccurate. It says more than 32 threads, and those CPUs (I have a 5950x) is exactly 32. Outside of serious workstation CPUs like Threadripper, server CPUs, or non-consumer and non-mainstream (Intel/AMD) x86_64 chips, or some other weird UEFI/BIOS hack that might report more, no one has over 32 threads. Being a software engineer myself, I'm confident about their precise language here -- translation, "if you have 33 threads or more." It would be a non-sensible decision to expect very popular high end CPUs to disable a very valuable general performance feature for your entire system, needing to reboot everytime to get it back. Though maybe a reboot isn't necessary for those who are aware: in Windows, you can limit the cores a process is allowed to run on by going to the Task Manager -> Details -> Right Click on the DCS process (yes, you will have had to launch DCS already)-> Set Affinity; and in that dialog deselect every other core. I'm fairly certain that typically the hyper-threading pairs are enumerated next to each other, which means CPU0 and CPU1 are on the same physical core; CPU2 and CPU3 are together. So basically you want CPUs 0, 2, 4, 6 etc to be selected. I haven't tested this myself and it could well cause DCS-MT to crash if it counts cores at process start (likely), tries to use them later, and then Windows tells DCS to f-off lol.

Also, you can check the Windows event viewer. Just use the Windows search feature and type in "Event Viewer" it should be the first result. On the left side bar, open up "Windows Logs" -> "Application". You can do this for crashes that have already occurred. For convenience, you probably want to use "Filter Current Log..." on the right panel, and select only Critical, Error, and Warnings to see bad stuff that happened. You can see what exe or dll crashed, and what the error code was if it was a pure software issue. Based on what DLL or exe that crashed (if it is a system file or DCS file) you'll have an idea of what might be a problem. You can also check out the "System Events" section for anything fishy about your other hardware or your overall system health looking for any errors/critical. Application crashes will definitely show up, and if the application crash was associated with a hardware failure, you should have two events at the same time. If you only see application crashes, then I would stop worry about any hardware and think about what might need to be reinstalled / updated, or simply changed just to have a different combination/result that lets you know you're messing with the right thing.

I use the mode dial as a DCS modifier. In VPC software / device firmware, they are just buttons and the only thing I have them do is change the color of all of the LEDs so I know one of the mode dial buttons is being held down easily. In DCS, I assign each mode (aside from the first) as a modifier. That way I can have a lot of buttons and each mode can have a consistent them acros s modules. Mode 2 might be radio+comms, Mode 3 would be lighting etc. For the most part, any mode other than the default is for startup or non-combat / non-common functionality. The LED colors are critical because in DCS, if you're holding down a modifier button on any controller, no other controls will work and it leads to confusion when you press a button on another device and it just doesn't do what you expect it to do. 99% of the time, this is because you have a modifier held down on another controller. Aside from "having a physical replica" of a throttle, the CM3 is pretty much perfection for me. The only way they could make this thing better is if they included configurable magnetic positional switches, and could be set based on game state. That is, the way the JFS switch auto turns off on the F/A-18 would be seen through the module export, and would trigger the controller to return the switch on the throttle. I would want that optional as different modules have wildly different natural behavior so through firmware/software you could say which switches you wanted to behave as momentary or not and in which positions. I don't even know if it's physically possible to do such a thing, but it would be hella expensive for one; and add serious power consumption. Surely it would need it's own power source beyond what USB delivers over type A ports. Maybe type C 35W+?

There's really nothing to worry about until the next generation of video cards as MSI Afterburner already seems to be working with 40xx GPUs. If the next generation has some architectural change that has something else to configure/control, and changes how some current/existing settings are signaled to the hardware, at that point there is a problem -- not before.

Can you be more specific about what to search for (though I will try). I'm actually seeing this issue only in the F10 map. Everywhere else seems to detects where I'm clicking just fine.

I think people are missing a much much simpler picture for why the Super Hornet is not a good module for ED to develop: more modern airframes have far more complex avionics systems that eases pilot tasking. Essentially, ED would be spending a lot more time to develop complex avionics for a module we spent less time learning because it almost entirely flies itself. I may be overextending in that the Super Hornet / Rhino / whatever isn't that automated, but if the E variant is not able to fly missions that the C variant can't touch, then the extra time simply isn't worth it. Yes, players will feel a bit cooler "oh wow, I get the latest and greatest Hornet!" but actual gameplay will actually be more dull and less skill based. Last thing I can think of, if some of the extra capabilities are aircraft to aircraft related, the support from other aircraft may simply not be in DCS world to really allow, or it may make the redfor vs blufor aspect too imbalanced.

Appreciate your help. Quick one: That was either a very funny/unfortunate typo, or did you mean "assisting"? I'm sorry if I'm being difficult, but I want understand things at a deeper level to know what the limits of my possible results might be and how (if any) impact this will have on my exiting audio setup which I'm very happy with. If that typo was meant to read "assisting" then it seems like SimShaker Sound Module (by name) and Simshaker for Aviators is working by looking at the telemetry from DCS and then embedding data into audio signal that the Buttkicker Game Plus responds to? I'm not an audio engineer so I don't need to know exactly how it works but I am I'm curious how it's possible to adapt the SimShaker settings without modifying the audible sound signal assuming decent speakers and/or headset+DAC/amp are already being used. I hope I don't sound like a broken record repeating, but the only way around this seems to be a direct interface the buttkicker. I like that it can operate as a sound card and work for any game, movie/video, or music you are playing. But I would hope it has a purely force/haptic feedback device mode as well.

@Immune and @raven06aDid pre-orders made for the original Orion joystick base + F-16EX grip + Orion2 F-16 Throttle turn into the same thing with a Orion2 joystick base instead?

Apologize for the percieved jab. Let's distill the offerings and I'll be a bit more careful to not add characterization. Only a matter of choices and preferences: Your approach is designed around having a singular configuration for all aircraft and helis. The convenience is zero effort on your device while switching. Players can add/remove afterburner detents if they wish, or leave them on between fixed wing aircraft. The throttle/collective travel for all aircraft is 0% before bottom detent start (while pulling back), and 100% after the afterburner (pushing forward). My approach is for those who are wanting and willing to change, add, or remove detents when they switch aircraft and they want to maximize stick travel (particularly for helis) reclaiming off-idle-start space at low end of the throttle. As with the other set up, removal of afterburner detents is optional between fixed wing aircraft. Is that better?

Can you clarify what you mean? From what I understand the Buttkicker can operate out of the box based off of the sound frequencies. And I think that would work well in DCS, but I'm looking for something that works through exports / DCS-BIOS / dcs-interface so I can both potentially tune or turn off individual effects, and isolate how Buttkicker works from my sound output configuration and wiring. Can I do this with the Buttkicker Plus and any software that is currently available?

Anyone know what is the definitive software set up for the newer Buttkicker Gamer Plus/Pro? The maker seems to be wanting to streamline and take control of the software experience Buttkicker HaptiConnect + official plugins. Their primary market is car sims so DCS isn't available, but it might be in the future. I'm curious with the new product is there a direct way to use the game data (instead of the audio) that has worked before or if I have to use the audio until an official plugin comes out.

I think I like the way I set up the CM3 throttle better. The weakness of imacken's approach is you are signing up to only use the travel between the detents for the mdoules that don't need it. Virpil have so nicely provided an easily swappable detent system that can be placed or removed probably faster than it takes to load into a mission, so I think it should be used. The only problem? Even if you physically remove the detents, you haven't configured the axes values any different to not include idle/off positions. The solution I went for adds two virtual axes that are only active in a shift state (2nd mode) on the joystick. I won't be as detailed unelss there's a lot of interest but here's some general details how I operate: When I fly helis or anything with no detents installed, I'm in mode A. The throttle axes values are on the full range. When I fly a jet with engine idle/cutoff detent in, and possibly my preferered afterburner detent (or not for A-10) in, I'm in mode B. This approach won't work if you need different jets tied to different shift modes. You can have more than one shift mode for jets because the full travel is always available on all shift modes. If you're using shift modes to control different button presses in DCS, I'd suggest moving off of shift modes and using modifiers in game for a latched switch or mode dial (your preference). Though I have little high fidelity jet experience other than drilling their cold start procedures, this set up worked with F/A-18, F-16, and F-15C without adjustment in terms of activating the afterburner at the right point when I used the Classic Detent+ pushed as far up as it goes. Afterburner is never triggered if I don't use the finger lift, and it is never shutoff if I come down to the soft detent position in-afterburner. I don't see myself needing to be picky about exact activation point in or out as long as it's clearly between the two. Before afterburner, I do see a discrepancy that someone else has reported where typically one throttle stick will see 66% and the other 68%. Again, i haven't tested significantly but I don't foresee this causing any problems. General steps: Create two new axes to mirror each of the throttle axes (neet to use the same sensor type; main port, sub port, and Calibrate the entire joystick again after axes creation Go to each axes' configuration and set begin an end input values based on the min position right before idle-cutoff, and the max position pushed through to max (through afterburner if that detent is in). Set the shift mode to something. Assign button to activate the shift mode. On the same shift mode, use existing tutorials to set up engine off/idle/start buttons (2 buttons per axes) so the same jet mode will work for all jet engine start procedures -> even FC3. Note: the off-idle/start buttons you assign to low throttle need to be applied to the original axes not the new ones. The new ones won't cover values on the low detent travel range at all. This is also why the original full travel axes has to remain on and should NOT be put on any shift mode. Only one shift mode is active at a time In DCS you can no longer use the Axis Assign auto-detect functionality and have to assign manually because when you physically move your stick in jet mode, DCS will see both the original axis and the shifted one move and won't be able to tell which you want. After writing this guide, I thought through the two shift mode limit I have with my set up realize that I shouldn't be using the 5-way mode dial for just two positions. I I can use T1 or T2 and free up that mode dial for a dial entirely! Happy flying every jet or heli with only moments to swap between!

How is the extra button a problem for you? Couldn't you just not bind it to anything? It will always physically click but it makes no difference if DCS ignores it, or the hardware doesn't send it.

Going from 20220407 to 202206xx (forget the exact date/number) didn't reset my already bound DCS controls. That being said, the update certainly resets any profile info set up and is certainly capable of resetting hardware IDs.

It looks like Winwing had a new Orion2 joystick base in development and they just released it 2 days ago. They also redid their storefront and it looks more well done now: https://wwsimstore.com/view/shopping.html. I am kind of annoyed that their website didn't make it clear what was happening because I definitely would have waited for the Orion2 and my eBay purchase of the Orion1 base certainly cannot be refunded. There is Orion2 joystick base stock to be had now, and the following is the only impressions video I've casually found. It seems like the Orion2 is a solid improvement, and most exciting is the cam system appears to support FSSB as a kit at a far cheaper price than RealSimulator. Unlike RedKite, I don't entirely crave super soft joystick centers. I agree that it's easier to control, and in some cases a big machine is easy to control but I don't want my joystick to almost fall out of center.

I've flown a lot of BlackShark in my day and I'm finding that the Deployment campaign is super punishing probably because the core game engine has updated in a way to make things incidentally more difficult for a number of reasons -- I'm referring to DCS World 2.5, and maybe even 1.5. Many of these missions I've played and beaten more than once so I know where to look for targets and where threats are. Three key examples: For every missions the first big issue is the AI is terribly bad. It was never that great to begin with, but I used to at least be able to have my wingmen fire Vihkrs at obvious targets; and do so safely if I've already taken out AAA and air defenses. Now, even if the only thing left are defenseless targets not shooting back, they fly off to nowhere or do nothing when I tell them to engage. Sometimes they just say negative. The only command they sometimes execute halfway is recon. This basically means only my payload counts for the mission. I have to engage every necessary air defense target, and every mission target. If I waste rounds or missles on something unnecessary, I might simply not clear the mission. Deployment campaign, mission 8 or 9 "Assault on Weapons Cache" -- Though I may not have the number right, this is the mission where you have 3 wingment (flight of 4); with a 50+ minute flight time, and you need to clear a very high elevation mountain pass before getting to the missiong area. This mission has a group of BMPs with AT missiles that will blow you out of the sky when you get within 3km; and they are very hidden in trees preventing you from killing them. I know they're there but they simply are not visible from Vihkr standoff range, and it would be terribly inefficient to attempt to fire cannnon rounds. I did eventually figure out that my mission targets were light armor and I could gun them all down out of BMP missle range and pass the mission. Deployment campaign, mission 9 or 10 -- this mission has you taking out some ground targets so that a ground convoy can move in. The mission is mostly fine -- a bit of a mundane scan+search and shoot, buth that's what Ka-50 is. However, the mission script messed up as I unloaded on some friendly tanks that had moved past living enemy infantry and neither side engaged each other. It wasn't until I landed on the FARP that the infantry at a certain angle shot at me and I kill them instead. Only after I got mission success did I see the fratrice issue, and I knew it was the tanks I killed prior but it made no sense as I saw them drive right by the FARP. Deployment campaign, mission 11 -- is a coordinated stike mission against armored targets. First, this mission unreliably has the allied AI flights engage air defenses. The mission has you standing by while allied flights take out the SA-8s. Well, first the allied flights seem to frequently fail in that mission so you should not entirely wait if you observed that they have all failed. And given SA-8's have 12km range, this can get tricky. But the real problem with this mission is that there aren't enough Vihkrs to take out the serious AA threats AND the mission targets. Even if every Vihkr goes to a mission main battle tank target, there are still more targets left which you'll have to get within gun or rocket range, and to do that you will be in range of AA that hasn't been addressed. This mission wasted the most time from me and I flat out skipped it. I've played this mission to 95% efficiency taking out everything that could even shoot at me and still not able to pass the mission due to running out of weapons. This mission has a more fatal AI flaw: the AI winment doesn't even seem to recognize mission targets are right there. The mission also takes place right at sundown and a flight all the way back home turns it over to night time and that is not a flight the Ka-50 is capable of. I ended up skipping this mission. Deployment campaign, mission 12 -- I just got here tonight and is the first mission I don't member playing more than several years ago. I spent a lot of time flying slow and scanning around once I was in the target area. This mission has a low cloud deck and the ground is 85% or more covered with trees. Nothing was in sight until I saw a FARP. Nothing in the FARP fired at me so I flew over it initially, to get to the secondary target and again saw nothing, but looking back I noticed my wingman getting shot at and engaged air defenese (it died of course). I flew back into the area and the rounds came from the trees. I fired back and somehow it seemed to work because I wasn't fired on more. Returned to the FARP destroying three trucks that couldn't shoot back at me, and then a facility before meeting my end from something. When I looked at Tacview, I was surprised to see that the initial rounds from the trees that fired at me was a massive force of tanks. There was nothing shy of two additional groups I either did, or nearly flew over but I didn't see them at all. I'm pretty sure every target and threat was under the trees. How is a BlackShark supposed to handle that? I haven't even tried this mission a 2nd time but I'm certain I'll skip it after only 1 more attempt most likely. I think this mission is broken in terms of trees being where they weren't before. Anywas, the TLDR is that missions in Black Shark are very broken. I may need to quit this entire campaign in favor of one that has the right and intended difficulty for the state of the engine now.

Got the F16-EX Stick yesterday and experienced the wrath of the ww export lua on my FPS. I didn't even bother checking what my frame rate dropped to as it was clearly in unplayable territory compared to what I'm used to (just under 50-60, monitor is at 60hz). I knew there was an interval value to check, but didn't get into that last night so I will check that today and see if I can improve. It's a bad design for SimappPro to believe it should modify Explort.lua on startup without manual user action or confirmation. I commented it out, and was disappointed to see they inserted more lines putting it back rather than understand I didn't want it there. Essentially any time I might use the app to configure or examine my joystick, I have to remember to also fix Export.lua or else the next mission I load will run terribly and I'll end up reloading again. I have it in mind to misdirect SimappPro about where DCS is installed just to break this functionaliy. Overall, I just kind of want to mess with rumble, but I don't know how much I'll care for it. Most of the conditions it is useful for I have no issues detecting via audio and visual cues (shaking, actual aircraft audio warnings, or listening to the air flow). The stick was a clear choice for me based on buttons alone. If I had the Takeoff and Combat panels and needed SimappPro running for LED control, I'd be disappointed. I use ViLA for Virpil panel LED control and couldn't be happier how functional and lightweight it is.

I'm posting this because I don't know if people have noticed what I saw while I had been waiting for an F-16EX joystick + Orion2 F-16 throtttle set for over a month starting sometime in mid to late May. If you look at the Winwing store inventory carefully, they have had everything in stock for a whle except for this: https://wwsimstore.com/p/333.html -- the Orion joystick base. If your order does not involve this item in a combo, Winwing will start processing your order and ship it in days. I did this myself getting only the F-16EX grip and it is already on the way in a quarter of the time I spent waiting in the pre-order queue for the full set. I ordered the Orion joystick base from eBay which is more expensive, but I really wanted that joystick combo over anything else on the market. I used the waiting time to do further research and reverse my decision on the F-16EX throttle (my own reasons). It seems well in-stock should I want to come back for it later. I realize this approach misses the huge savings you get with the combo. If you do the math, the combo is essentially the cost of the Orion2 throttle base + an Orion2 throttle grip (of choice, take the most expensive one, F-16), + Orion joystick base which brings you to $544 and you're still shy a joystick grip of a full HOTAS. It's $548 for the F-16EX combo which is the full F-16 HOTAS. The F-16EX grip alone is $186 currently so if you drop the combo, you're speding $182 more, and still technically cannot get the Orion joystick base. Hope this helps someone decide where to go.

Had to turn on Easy communication. I kept the prior setting to Force options for all missions/campaigns off. I'll just use different settings KA-50 campaign and multiplayer when I get back into it.

Edit, whoops, not a solution: I didn't do a deep enough test to see if it was working on one of the campaign missions. I opened up the menu successfully while waiting on the runway, but checked in a couple minutes later and the action stopped working. Truly bizzare and frustrating but at least I didn't fly for 30+ minutes to the target area before realizing it was broken which meant I wasn't going to clear the mission. I'm going to test further options. This is what I previously wrote:

Ouch. This made “The Other Side” a very very difficult mission. I cleared it years ago before Black Shark 2 came out after maybe 3 attempts only failing on the mountain ascent before it. Now I just can’t get my wingmen to take stand-off shots rather than “running in” and dying to the AAA. Even if I make every Vikhr count and lots of stand-off (3.km) 4km) cannon picked off with both cannon round types, I still basically end up on empty with more targets to kill. I must have failed like 5 times yesterday.

It does. I wonder if there's a way I can mark this question as "answered."

Does Winwing Orion2 throttle base save settings/profiles on the device itself (like Virpil) or does it require SimAppPro to be running in the background? I'm really hoping that if I change settings like adding buttons for axis position ranges, that the device itself retains the information for easier DCS integration. If done through running software virtualization, then at some point a reinstall or PC update would cause DCS to see an entirely new device and require setting it to be set up again. Really what I'm asking if Winwing's input configurations are virtual joysticks done through software or is done in their hardware? Which of their devices can do this or not? If it's virtualized, it seems Joystick Gremlin or vJoy would be a better option than SimApp pro.