darkman222

As he says. I heared about a real effect thats modelled in DCS too which is, if you are in the line of sight behind an aircraft thats being locked up, he will get a lock up warning on his RWR but you will get a RWR warning too. Even if you are not the one being locked up and maybe 20 nm further away from where the actual aircarft thats being locked is. So neither RWR nor radar are magical devices. It all depends on how the pilot interprets it.

Yes you are right. I just talked about the non-linear fly by wire curve. I just did not realize that there is a deadzone also being applied to the F16 input. I am mostly doing dogfighting. So I might not have run into any issues by that. When would I realize it except when air to air refueling or doing a very very precise landing?

I think there are always compromises being made with PC input devices. So you might realize for example on the YAK52 that the stick travel is not linear compared to what your physical input does. It also has a curve applied to it. You can see that if you move your physical stick and watch what the virtual stick does. No big deal for me. This behavior is much more exaggerated in MSFS2020 for instance where the last 20 % of input just really kick in. I already had the idea for the linear F16 stick curves as option, and posted in the "input" section. But I think its really obvious that if you would follow that approach to recreate the original input curve, the F16 is the perfect aircraft for that. So what I did to counter that using an input curve of -20 in the DCS inputs menu, which kind of smoothes it out to a more linear experience. I feel very comfortable with it now, using a traditional stick. I am fine with it either way. But I also would not be disappointed on the other hand if we got that option you ask for at some day.

These are issues popping up now but they have been reported long time ago. Apparently they have not been adressed by ED (yet). I posted about the F16 locking the wrong target. Result: Investigating: I also tried the method @Frederfdescribed. It seems to help a little but might be just my personal feeling.

Fatigue was just an additional idea I had. I dont want to lie, so better look the numbers up somewhere. I think some air to ground configuration are 6,5 Gs maximum.

You were pulling 8,6 Gs with that loadout when the aircraft broke apart. Thats way too much. Airframe overstress has just been added recently. And maybe there is some kind of fatigue added to it too. So you might have been over G-ing the aircraft several times before, but nothing happened up to this final pull when it was finally too much.

I started a thread about it quite some time ago. Its two separate issues.

Seems to be a DCS thing. Yesterday my virtual pilot got shot when I was fighting on a MP server, while I had the guns button depressed. I was watching my aircraft from the outside drawing nice tracer circles. So I guess when you eject its the same like the pilot being dead. The last input is kept. Which will be of course most of the time pulling up when ejecting.

Forget about what I said above. I was on that server again. Bad CPU frame time. After a few minutes I watched the replay through the headset. The EXACT same stutters and CPU frame times ocurred during the corresponding phases of the replay. I was remembering them because I was just a few minutes online on the server. If I could remove the clouds temporarily in the replay I could check if they are causing the issue too and do further investigation.

So I have a theory. First we got the new smoke und flame effects. They look like some kind of volumetric, non sprite/ bitmap based technology. When we had this, but no clouds yet, I realized when there were lots of players shot down, and lots of those smoke trails were in my sight, my CPU was slowing down. Because the smoke effect has to be synced between clients. When the amount of smoke got less, the CPU frame time recovered. Then we got the new clouds and frame time jumped up in general. Of course GPU demands were rising too, but fpsVR shows that my system is still capable of the clouds GPU wise. Also there are 2 similar dogfight servers out there. One with and another without clouds. With a comparable amount of online players on the one without clouds the CPU performance is still okay.

You can get it partly working from isometric / external view. You can get some of the aiming symbology back as long as you have control by mouse. As soon as you touch your primary control and the mouse control is disabled the symbology stops working. I hope ED fixes that. CA is too much fun in VR.

Yes, welcome to DCS. I started with Strike Commander ( as my avatar shows ; ) , then got my hands on Falcon 3.0, which was way too complex for a teenager. Then went to Falcon 4.0 which was a lot more accessible in usability. Then I forgot about it. Realized that Falcon BMS was out, went back to flight sims and the F16. Again forgot about it because it was already an old sim. When I heard about DCS, was when the F18 came out. I knew systems were comparable to the F16, and VR available, I started right away with the F18. I learnt a lot of, in deed, comparable systems and technical approaches. When the F16 came out I immidiately went back to it. Systems and flight model are still under development. I am dogfighting a lot in it and wait until it is finished. But sometimes I just fly around in it like cruising in a car. And this is amazing as well. I enjoy the unrestricted view in VR and the good feeling that I know that systems and aircraft from so many sims for such a long time now. Next challenge for me: Private pilot license. No kidding. I am taking the first lessons on the ground already EDIT: If youre new to DCS try the aerobatic servers. Friendly places to "just" fly around with real people. Much more fun than in MSFS2020... or if you wanna get shot down immidiately try a "just dogfight" server

Short answer: On a dogfight multiplayer server: Out of fuel or out of ammo. Not very realistic either I know

I get the point in doing a G warm up. And its good that ED is working on the pilots G tolerance. The best guess would be to look into the requirements about the required amount of time for real F16 pilots to withstand 9G. One year ago I did a side by side comparison with a Thunderbirds airshow 9G turn against the DCS F16 we had back then. Result, the DCS pilot was asleep while the Tunderbird pilot was still happily turning his jet. Its way better now. I would not want to waste my time doing that comparison again, because I feel the F16 and whats related with it is developed in its own pace and re-worked if needed.

Its sad that this fight Mover and Gonky did was before ED claimed the F16 FM tuning to be finished. And maybe the F18 FM is still undergoing revision. Dont know about the current state here, I have not been in the F18 subforum for a long time since the F16 came out. I was hoping the F16 FM would have been adressed as a whole and not just sustained TR, and then again 2 updates without any of the other announced FM changes. This seems to produce again a pointless discussion about an F16 FM work in progress... Counting the days until this thread is getting locked like the others. I hope Mover can arrange a rematch after ED states that the F16 FM finetuning is finally done.

Right, the canopy goes too. That adds to the pilot and seat weight loss. Now we have about 300 kg of weight loss in the front of the jet. But, wouldnt a functional FLCS still be able to compensate that?

Thats my question. Exactly. What would engineers design for that case? I would assume that the best case would be if the aircraft just should continue to fly the last direction. So the pilot can point the aircraft where he wants it to crash, after he ejected. A shift of the center of gravity would ocurr. But I googled that the seat weighs about 70 kg plus the pilot 80 kg. Thats a weight shift of 150 kg in front of the COG. The F16 compensates quick and well if you'd jettison all stores simultaneously. Not really sure about that explanation, although the pilot and seat are really far away in front of the COG. And if the controls surfaces remained in the last position, although the pilot would have let go of the stick, if the FLCS still remains functional, but kept the last input, it would prevent the jet from snapping the wings anyway. I guess there is some game design going on here. Imagine on multiplayer if you followed an F16, you did not see the pilots ejection of and you keep following him for 15 minutes in afterburner just realizing there was no actual player in it

I suspect the amount of objects (vehicles, clouds, stationary objects) that need to be synced between clients affects performance. A dogfight server that ran smoothly before 2.7 has a significant higher CPU load now, since the server started to use clouds, which also need to be synced between clients. Unfortunately that additional amount of CPU load the 2.7 update created, made CPU frame times so high that its not a smooth ride for me any more. But the same server has a BVR area whith less players in it and players being not as close together, which is still kinda okay performance wise. So I guess less players in the closer area around my own aircraft causes less syncing between client machines. Obviously the clouds were an additional thing that has to be synced for every player. Another dogfight server did not start to use the new clouds and the performance drop 2.7 gave was not as significant. Just occasional stutters. Almost like before the 2.7 update. I guess thats how it is. DCS evolves hardware comsumpsion rises with it. Fingers crossed that Vulkan and multicore optimization will optimize the performance in the future.

But the main question, beside that VR causes higher CPU load than playing in 2D is why does multiplayer add so much CPU load too?

That may sound like a stupid question. But why does the DCS F16 depart from controlled flight after pilot ejection? Looks like the FLCS gets disconnected and the aircraft gets into an unstable flight condition as it would if the FLCS failed. Is it realistic? Or is it just game design that other players know that this F16 is dead?! Seems to behave differently in multiplayer and single player. In MP the F16 likes to pull up and snap its wings. In SP it just slowly banks down into the ground after ejection.

No. I am using W10. I just checked it anyway. This option seems not to exist in W10. After all it does seem that its not only me having issues, regarding that 17 page long thread about performance loss by other users. I'll try to see if some of the tips in the thread DavePastry pointed out could help in my case. EDIT: Tried it, I already had everything as suggested in the thread.

Multiple things going on. The F16 FM is not completed. ED started to work on it, but no more updates in the last 2 ones. Then its just a game after all. Cheating in the F16 is difficult due to the restrictive FLCS. On the other hand, when playing in the F15 the "over over over G" sound is your best friend. Same for the cracking sound of the wings in the F14. Luckily the new G damage model for the F18 makes paddle pullers snap their wings. That was some kind of relief to see some paddle pulling F18 just explode after merge on the dogfight server, when the G damage improvement was introduced What I am saying is that you have to fight against human players on a regular basis to get a feeling how your chances are in aircraft X against enemy aircraft Y. My chances to win a fight in a F15 is higher than in a F16 currently. The reasons for the outcome are quite many. A big factor is that its a game and people fly unrealistic. But what I would say is, two fly by wire aircraft fighting against each other should bring up results that are expected by public knowledge. F16 should be outrating aircraft it should be able to outrate. I am not talking about dogfighting in general which is a fluid sequence of manovers. I am talking about when all options are taken away from two human players. Every card is played. Just sitting on the deck and and doing a two circle fight. That happens quite some times on the dogfight server. What counts now is if the pilot knows the best turn rate for his aircraft with the given fuel state. I'd say fuel state because on the dogfight server you mostly dont have missiles as extra weight. If the result is not turning out as expected something must be wrong in the FM of one or both aircraft.

Game plan in a guns only fight agains a good F18 driver is: Hope he makes a mistake, if not: run, and reset the fight. And try again. Gameplan in the F16 against the other fighters (with a good driver): Dont get shot until you are bingo fuel. Then fight a rate fight while your unfit pilot struggles to stay awake making some 9G turns. Thats the way it is at the moment.

I have to agree. You cant measure any aircraft performance against AI. If you want to put an aircraft to a test, go on one of the dogfight servers. Try to rate fight against a human. It takes a lot of time and patience. Against an F15 its often the 8th or more full circles when the opponent gets impatient and bleeds his energy. Then you know its your chance. Against a MIG29 AI, you also try to outrate him. But what he will do is go up and you often cant follow him. AI can bend the laws of physics for computing performance sake I guess. On the other hand it does not know how to jink. You can point your nose in lead on AI forever and it just does not care.

Thanks for pointing out that thread. 14 pages about it already. Gotta go through that. I think its a combination of graphics improvement and MP. Also almost all servers start to use the new clouds now. Which are nice of course, but they seem to add to the FPS loss with around 15 to 20 % too. It seems that all adds up. I also have the feeling that the clouds syncing between client eats up MP performance through higher CPU loads. Servers that performed very well before 2.7 became now unplayable.