twistking

Sorry, but this is nonsense. Not only can you simulate it in a game, you can also easily make it enjoyable and immersive. Same as battling with a broken aircraft can be fun and exciting in sim, while being less fun in real life. Driving a car through a field of rocks: Driving sims can easily do that. The important aspect is, that you cannot only convey it through the flight dynamics (or driving mechanics in this case), but you'll have to utilize visual art and sound-design. Without sound-design especially it might just feel as if your controller is jittering, but with proper sound it will immediately communicate the bumbs and vibration. You can add camera effcts and if you want to go the extra mile, you can add cockpit elements visually engaging with the vibration. Can be done, has been done since ages. Could be done in DCS, requires effort though!

It should at least slow down the initial startup. I don't think it's good practice to delete those everytime you run DCS...

You are not CPU bound! If you have your GPU at 99% usage, your CPU is very obviously fast enough to feed the neccessary data to the GPU.

correct. it's not possible unfortunately. ED would need to build a more complex system for DRM in MP environments to allow this. such a system would also allow paid mp/coop campaigns (which would be nice). i don't know how liberations works, but if it's not paid then it does not have DRM and therefore should have no restrictions. if you build your own multiplayer mission in the editor for example, you can still play it alone in singlepalyer on your pc locally, or you can host it on a server and play from a client pc, or you can host and play it in multiplayer.

1. missions that run on a dedicated server are technically multiplayer missions, even if they are designed to be played by a single player. it's set up differently during mission creations. 2. paid campaigns have some sort of DRM and only run on an account that has purchased the license. this authorisation does not work when connecting to a server. the server would need to check for the client's license, which it cannot do.

your computer is using whatever version required by the application.

I feel you!

don't mess with scheduling settings unless you have a very good reason to do so. if your "good reason" is someone on the forum saying so, the burden of proof is on him. so i'd strongly suggest to leave it on whatever the default setting was. there is that nice german word "verschlimmbessern": making something worse by trying to improve it. i wish there was an english equivalent.

fair enough! if the technical staff is too busy with doing more great MT magic, it would also help to have a slightly more high level overview of what the midterm goals for MT and general performance are. maybe this is something that a team-lead/producer could say sth. about without knowing all the nitty gritty details. i (we?) don't expect any guarantees how DCS will perform on a specific system, but - again - some technical overview over current MT implementation and goals would already help a lot to form a new understanding of DCS performance, now that the old dogmas are shattered (hooray!).

@BIGNEWY do you think you could - in the near future - make a newsletter happen that talks a bit more about technical implementations of MT - especially looking into the future a bit? Background would be that i (and i think many others around here) are thinking about upgrading our PCs with DCS MT in mind. Unfortunately user benchmarks are all over the place (regarding reliability of sources due to different setups and quite frankly different levels of technical knowledge) and only a fraction of users run newest generation of CPUs anyway. For example your MT FAQ mentioned that DCS would only use intel e-cores for data streaming. Is this just a WIP stopgap solution (E-cores should be capable of doing much more: A 13th gen cluster of 6 or 8 e-cores alone should be faster than my current CPU (i7-860@3,4ghz), which runs ST DCS just fine in pancake.)? I don't expect a CPU recommendation, but it would be nice to get some more technical background on current implementation of MT as well as goals for future improvments to get a better understanding of how different architectures might handle DCS MT. Examples: Focus on ST vs MT performance, Cache, Memory speed and as mentioned before: Thoughts on classical multicore design vs big/little design (intel 12th gen and newer). Maybe an engineer can write a little piece about their design goals and expectations... Thanks.

thanks @Taz1004 you're doing great work. here's hoping that this thread wakes up the responsible devs: 5/5 stars! texture sizes are ridiculously oversized. i bet that you could squeeze them even more before quality would degrade to a point where it was really notable.

i think waiting for official clarification is a very valid approach. if you decide to tinker for yourself, you can just make a shortcut from the exe (in the bin_mt folder) and put that newly created shortcut on your desktop. all the command lines would go in the shortcut anyways, so if it does not work, you can simply delete the shortcut. i could imagine that someone at ED just did not bother to put all combinations of different executables and different launch parameters in...

my guess would be, that you have to start dcs.exe "manually" from bin_mt folder and add all your vr commands as arguments to the exe with windows. this is just an educated guess though.

my guess would be, that you have to start steam dcs with the MT exe from the MT folder (" .../bin-mt\DCS.exe" ?) and force openXR with a windows command line on the exe itself (making a shortcut f.e. and add your argument to the shortcut). hope that helps! (not tested myself)

yes, i fear that's a compromise you'll have to take. MSAA only works on aliasing from geometry and the rough lines on the F-16 are a texture problem i think. SSAA would smooth these out, but as you've experienced it comes with quite a performance hit. DLSS might indeed help in the future, but i think there is also a real issue with the textures on the 16 (i'm guessing missing mipmaps): I think it is already reported since many years, but it might be worth reporting it again. *edit* while @trevoC has good setting recommendations in his answer, i'll advise to at least test with SSAO. It will tax the GPU, but in certain situations it makes the cockpit look much more natural. I think your GPU should be able to handle it if you don't use SSAA.

it appears to me that you are running at the wrong resolution for your monitor. too high and wrong aspect. also you're running ssaa, which quadrubles the load from the high resolution. choose the correct resolution for your monitor, disable SSAA and enable MSAA to compensate.

this is true. i feel that ed's modern jets wouldn't even qualify for "basic" DM at the moment. they do need work in that regard!

Well, a spherical map would be more realistic of course. Some things are easy to "fake" like radar horizon, radio comms etc. but i'd guess especially visual aspects are more complex. For example the horizon on a cloudy day (especially with a low sun), or where aircraft visually appear on the horizon. It's just the proper thing to do in a realistic simulator game. I don't think a casual flyers will notice it without knowing what to look for, but it will properly also make life easier for the devs, because they won't have to constantly "correct" the simulation to make it appear as if the game world was spherical. I think the biggest motivator for ED to go spherical though, is to offer a flying experience where you can access the whole globe eventually. You could not do this on a flat model (well, you could, but this would not be scientifically correct ). Honestly i feel that for a combat focused simulator the seperate maps are "good enough", but it's also clear that a whole globe offers many new possibilities...

thanks for the answers. i think you all made equally plausible arguments! i have skipped through an interview with him now, where he talks more casually and there he sounds definitely british. so, it was probably just his deliberate intonation for the promo video that made him sound less british...

The F-14 and F-4 rank quite high in my favorite planes list, however i did not pull the trigger on the tomcat, because i did not enjoy the dependency on the RIO, or AI RIO (Jester) to be precise. I think HB did a good job with Jester, in the end i just found the singleseaters more immersive for me, despite them being significantly less cool. I now wonder how WSO dependency will be in the Phantom. I understand that the Airforce has a bit of a different concept concerning twinseaters. Do you think that in the phantom there will be less AI interaction required, or more compared to the tomcat? Thanks

After watching the 2023 and beyond video, there was a little discussion in our group (all non native speakers) about the "accent" (if you can call it that) of Nick "basso profondo" Grey. I always assumed that Mr. Grey is British and his way of pronounciation in the official ED videos is just him trying to sound a bit more dramatic, theatrical or maybe also more "American" (because international audiences are more used to americanized English...), but others did not hear any britishness whatsoever in his voice, but even suggested a slight german accent (which i did not hear at all). Can some native english speaker enlighten us about the "accent"? Thanks.

I must agree. I am extremely dissappointed by stable moving to 2.81. It feels like mockery to me. Stable version has always been a bit of a meme, but at least it avoided the biggest issues. Now with stable on 2.81 it has become a complete farce.

it would definitely make a lot of sense to allow semi-manual/semi-dynamic missions. i think this would be the only reasonable solution to get the (huge) benefits of a dynamic and unpredictable situation together with the "level of detail" of hand-placed assets. this is by the way the one big concern that i have with the EDDCE: obviously a dynamic cmpaign is hard enough by itself (but absolutely doable as other titles have shown), but what i see as a bigger difficulty is how believable assets will move across the map and be placed on the map. That other flight sim has simpler terrrain graphics which makes it easy to put down ground assets without them looking out of place. In DCS i often spend a lot of time searching for a good spot to set up an AI SAM somewhat realistically: Taking into account the heightmap, obstacles likes trees and powerlines etc but also tactical considerations. Since DCS has such high visual fidelity, it will be harder for the AI engine to place assets in a believable manner i think. Hopefully ED can solve this. Alternatively it would be good if mission designers could "help" the DCE by placing some assets by hand, or assigning points on the map to help the AI with placement. Maybe the terrains could also come with handplaced "logical" positions that help the AI to "understand" the terrain better and make the DCE's assets placement more believable.

Thanks! I could imagine however that the radar not showing up at all might be realistic and intended. Unfortuantely i don't know for sure. It showing up 360° for everyone was definitely a bug/oversight though.

water has it's own setting and for this discussion i think one could claim that "water is not texture". With this i mean that the effect is mostly shader based and therefore has a totally different performance foodprint compared to what we normally refer to as textures (albedo maps and material maps...). Maybe it's technically correct to refer to the water as texture, but i think in general it's a good idea to think of it as somewhat seperate... ps: without testing, i'd guess that water has more of a processing foodprint and less of a vram-size/bandwith foodprint.