DaveRindner

Wait 2-3 mins for eye to adjust to dark in 2.5 Given what I knew then and what I think I know now, I don't think they should mess around much with night illumination. At first I also though that nighttime is too dark. Not anymore. Now I like it , sort of, the way it is. It appears to me that ED implemented human night/low light level eye adaptation. It takes few minutes to adjust. One notices that after 2-3 minutes, things are more visible. Avoid using NVG during this time. Plus moonlight is illumination light. As in real life, the darkest time is before dawn. Additionally light is simulated , it changes across time. So taking off at 4 am, and coming back after 5 am, you get to see early morning light. Physiologically humans are weakest, least alert , when awake, between 2:30 AM and 4: AM. Standard military Stand To (everyone is awake and manning post) is around 3 AM, depending on mission commander's guidelines, and tactical situation. Above applies to 2.5 with Deferred Shading ON. Legacy HDR is too bright to be believable. Mission file attached. A private little war scenario. American , Georgian, and Ukranian PMC vs Insurgents suppourted by plethora of various nationals, also PMC,. Concept is inspired by Dale Brown's Chains Of Command, and Harold Coyle's Strategic Solutions Inc. novel series. The real life precedent, is modern day Syrian Civil War, War In Donbass, Angolan Civil War, and various Bush Wars of 1960's and 1970's. All made use of proffesional combat aviators and maintenance/ordnance personell. Proxy_SheepDip_scn.miz

Quadro P4000 and P5000 shares same chip with 1070Ti and 1080 . But not 1080Ti. 1080Ti Quadro sibling is P6000. An extravagantly expensive GPU with more RAM then 1080Ti.

Oh come on! It is not unreasonable to expect a decent machine to have decent FPS in DCS 2.5. I need to wait until the my new rig arrives. That one is XEON W-2125, 64gb ram, Quadro P4000. It will run 2560X1440 panel and one 1920X1080 panel. Both are Q series gaming and graphics panels. I know my current is dated, that is why new one is incoming. Just for heck of it, I set settings as follows; Textures: High Terrain Textures: High Civ. Traffic: Med Visibility Range: High Heat Blur: low Shadows: Med Resolution: 2560X1440 Aspect: 1.778 Monitors: 1 Res. of Cockpit Displays: 512 MSAA: 2X DOF: OFF Lens Effect: Flare+ Dirt HDR: Off Deffered Shading: ON Anisotropic filter: 2X Terrain Shadows: Default Cockpit GI: OFF Clutter/Grass: 500 Trees Visibility: 100% Preload Radius: 100000 Smoke density: 1 Gamma: 1.4 Disable Aero : ON Vsync: off Full Screen : ON Scale GUI: OFF Observed FPS with current rig. I-3770K, 32gb ram, Quadro K4000. 5-7 fps in Georgia, 7-9 in NTTR. To get 48-60 FPS with the new rig, the performance of DCS 2.5 would have to be 10X (1000%) better overall. I have no such expectation of a Dell 5820T wkstation with XEON W-2125 4.3GHZ, 64GB DDR4 2666mhz ram, Quadro P4000 8gb. I expect 2X maybe 3X perfomance increase, and that is excellent. I do not expect more then 20-24 FPS with DCS 2.5 at the settings above . That is low performance with a way above average system. The single biggest FPS drag in those config is trees with Deffered Shade.

I want to agree with you. But we don't all have the machine we think we need. Unless I spring for two home systems, a hell NO for me. The system I need needs to run DCS great and my career apps, lest I find myself unneeded by the world. My system needs to run DCS at near its max settings at above 48fps, and it needs to run Maya, Realflow, Houdini, Arnold Render, Redshift, NUKE, Mudbox, and host of other utills. While 1080Ti does do that , and more or less great. Non Quadro boards have issues with Maya's Veiwport 2.0 technology, and unstable OpenCL GPU simulation. Common conspiracy theory is that Nvidia intentionally cripples GTX boards so as not to cannibalize Quadro sales. But game people, developers at Krytek (FarCry Engine) , tell me that Quadro and RadeonPro hobble some game specific shading performance features to achieve wkstation board reliability and lifetime. Makes one want to rip one's hair out. DCS is not the only flight sim I run. Other is Prepare3D from LockheedMartin. That application is more of a prepatory, not entertainment.

I don't use MSAA or Nvidia's FXAA. It sucks away FPS and on 2560X1440 is not really needed. Hell, anti-aliasing is expensive in normal non RT renders. Motion blur! Fogetaboutit!

More like impossible. But it will be three weeks for new station to arrive. I am also considering swapping GPU in my legacy system, an HP Z240 wkstation. However the issue is space and power. Space I think I have, as K4000 is itself almost double height card. However the PCI-X is at v 2.0 not 3.0 , and power supply is may be 400 watt, 450 if I am lucky. It has I7-3770K with 32GB ram , which is max. It runs my 3D apps really well. Maya, Realflow, Houdini, Arnold, Redshift (GPU). When DCS was put on it in 2013, it got above 48fps with A-10C, UH-1H, KA-50, and FC3. In 2014 I added Mig-21bis, and added F-5E in 2017. Once we hit 1.5.2 FPS started to drop. Sometime it would surge on a build, then drop down. My 1920X1200 panel died, and was replaced with 2560X1440 Q series gaming and graphics panel, that dropped FPS to about 36-32 in 1.5.8. With 2.0 FPS went down again, and 2.5 is barely , barely playable. I don't think upgrading CPU is going to change much. On Benchmark I7-3770K does about 9350 - 10,500 score with single thread score, vital for DCS and gaming, of 2100 something. A Xeon W-2125 3.6-4.3 GHZ has single thread score of 2570. And overall benchmark of 13700 something. Fastest commonly available Intel CPU with highest single thread score. I9-7900X hs massive 22,300 score becouse of its 10 cores, but lower 2010 single thread CPU score. I regret that I have never seen my DCS use more then single thread. When I cook a sim using Houdini, or Realflow, with OpenCL or not, as soon as I hit SIM, the CPU threads in task manager spike and stay spiked on full. I have never seen DCS even break 1/2 of any single thread usage. But DCS does use 100% of GPU, and I can see that in Task Manager on Win10Pro (latest update 10.0.16299.309). So DCS 2.5 is GPU centric. Performance gains from CPU are not marginal , but they are not order of magnitude higher either. Expect maybe 25-33% faster, probably less, between my current CPU I7-3770K and I9-7900X, slightly less gain with I7-7700. DCS is all about the GPU and that GPU needs to be good. No doubt about it. Obviously I can't just stick 1080Ti or Radeon Vega, TitanXP into my legacy box. It can't power it. So which board would do? I don't think a 1070Ti would work either. 1070 power specs are nearly same as 1080 and both are PCI-X 3.0. Sticking a PCI-X 3.0 into 2.0 socket, might function, assuming adequate power, but board will step down in performance. I think. It has to. That is why I posted in a different thread. What exactly is Deffered Shading? How do we benchmark it? What common 3D benchmark/score that we can see on a given board that gives an indicator of Deferred Shading performance?

This is not really a bug, and perhaps it is a WIP. But trees, as they are implemented , especially in Georgia, do not work intutively. If the trees growing out of the ground artifact , with Deffered Shading, cannot be gotten rid, the trees need not be drawn at all. Its so distracting, it wrecks the flying experience. Especially low to the ground in a fast jet. If the only way to not have the artifact, is to have trees at 100% and visibility to high and above, with significant FPS cost, then it is not really a system worth having. Better go back to 1.5.8 noncolliding but fast tree draw. Aside from Deffered Shading being really heavy on GPU, the trees in Georgia is the most visually distracting issue. Now we cannot turn them off. THe lowest allowable tree visibility is 30%. Because invisible trees are still collide-able, fast paced flying NOE is difficult. Basically avoid all dark patches on ground texture, becouse that where an invisible, undrawn tree is at. Not using Deferred Shading in Georgia, makes for a somewhat weird unattractive experience. Prior to 2.5, trees in Georgia were darker then ground, which makes for a correct expected experience. Without DS , in 2.5, trees are very bright. Neon green. Even at night. The scene looks wrong and distracting. So we are required to use Deffered Shading in Georgia. NTTR still gets away without DS, except at night, when every texture is very bright. So Defered Shading is in practice a required setting for DCS 2.5. But man, does it suck away FPS!

I get it that Deferred Shading runs at 60-100 fps with top'o'line Nvidia GTX boards . So it is not necessary to guess how DCS 2.5 with DS runs with 1080Ti, TitanXP, TitanXP Collector's Edition, TitanV (Volta GPU), Quadro P5000, Quadro P6000. Now how about it performance at 2560X1440 or 3840X2160 with mid-range workstation boards using single panel with a single board. Specifically Quadro P4000 Quadro P2000 Radeon Pro WX5100 Radeon Pro WX4100 Or midrange gamer boards GTX 1070 GTX 1070Ti GtX 1060 GTX 1060Ti Lower end GTX 1050/1050Ti GTX 1030 Quadro P1000 Quadro P600 WX 3100 Instant Action missions would do. For me any FPS above 36 is great and acceptable. 48 and above is great. 60 and above is fantastic. To me once sim or any game hits 48fps it is very playable. Assume 2560X1440 1 screen or 3840X2160 1screen Visibility range High Deffered Shade ON Shadows LOW or Flat Only Trees and Grass 100% Terrain Objects Default Cockpit Global Illumination OFF What is Deffered Shading engine? The look of DCS in 2.5 with DS is so different from 2.5 with legacy shading engine. There are new features such as how NVG work. Is it Direct X 11, or DX 12, OpenGL, OpenCL, CUDA. CUDA is only on Nvidia. Radeon support DX11/12, OpenGL, and OpenCL , but no CUDA (obviously). Which benchmark 3D closely approximates DCS 2.5 with DS performance? While my new home workstation will have Quadro P4000. It won't ship for three weeks. My existing one can be upgraded by replacing its older K4000 2GB board. Unfortunately, the 2012 vintage HP Z240 workstation cannot power 1080, or any board that draws more then 75W, and I have no desire to replace a power supply that is most likely proprietary to HP Z workstations. I beleive it is 400 or 465 watt. So my choices are limited in that regard. Otherwise the HP has been a really good trooper. If I am overlooking or unaware of something basic, please point me in that direction. Thank you.

I get it that Deferred Shading runs at 60-100 fps with top'o'line Nvidia GTX boards . So it is not necessary to guess how DCS 2.5 with DS runs with 1080Ti, TitanXP, TitanXP Collector's Edition, TitanV (Volta GPU), Quadro P5000, Quadro P6000. Now how about it performance at 2560X1440 or 3840X2160 with mid-range workstation boards using single panel with a single board. Specifically Quadro P4000 Quadro P2000 Radeon Pro WX5100 Radeon Pro WX4100 Or midrange gamer boards GTX 1070 GTX 1070Ti GtX 1060 GTX 1060Ti Lower end GTX 1050/1050Ti GTX 1030 Quadro P1000 Quadro P600 WX 3100 Instant Action missions would do. For me any FPS above 36 is great and acceptable. 48 and above is great. 60 and above is fantastic. To me once sim or any game hits 48fps it is very playable. Assume 2560X1440 1 screen or 3840X2160 1screen Visibility range High Deffered Shade ON Shadows LOW or Flat Only Trees and Grass 100% Terrain Objects Default Cockpit Global Illumination OFF What is Deffered Shading engine? The look of DCS in 2.5 with DS is so different from 2.5 with legacy shading engine. There are new features such as how NVG work. Is it Direct X 11, or DX 12, OpenGL, OpenCL, CUDA. CUDA is only on Nvidia. Radeon support DX11/12, OpenGL, and OpenCL , but no CUDA (obviously). Which benchmark 3D closely approximates DCS 2.5 with DS performance? While my new home workstation will have Quadro P4000. It won't ship for three weeks. My existing one can be upgraded by replacing its older K4000 2GB board. Unfortunately, the 2012 vintage HP Z240 workstation cannot power 1080, or any board that draws more then 75W, and I have no desire to replace a power supply that is most likely proprietary to HP Z workstations. I beleive it is 400 or 465 watt. So my choices are limited in that regard. Otherwise the HP has been a really good trooper. If I am overlooking or unaware of something basic, please point me in that direction. Thank you.

After Hot fix, Mig-21bis still crashes. Others appear to work. Also FPS performance has improved a bit. Graphics options also could use a saturation amount in addition to gamma. Some colors are really vibrant. I think the whole DCS World could use 15% saturation reduction. But that is my opinion.

Instant Action missions crash or don't work correctly.

https://www.bing.com/images/search?view=detailV2&ccid=2cYMmDXi&id=D78EBEB5521723AD5038E35D9F9F79AD4F215531&thid=OIP.2cYMmDXi54iBrW7d7qlaiAHaER&mediaurl=https%3a%2f%2f3.bp.blogspot.com%2f-tN5BgI4LYUg%2fWEwdJ_MJd5I%2fAAAAAAAAKK4%2f-0gYDCg0Kso1piESbfRaYrL4nawVHtvkACLcB%2fs1600%2fBlackburn%252BBuccaneer%252BS.1%252BButtler.jpg&exph=924&expw=1600&q=Blackburn+Buccaneer%3a&simid=607997501038202670&selectedIndex=100&ajaxhist=0 Needs Free Willy nose art.

This one is interesting. This was F-15SA , which is a F-15E modified for Saudi with export version of Sniper (ER?) pod, and export version of radar, without ASARS. Not the pure air-air F-15C. The weapon was a improvised SHORAD SAM IR/UV. Yemeni 'rebels' or other parties, took IRST system from MIG-29C export variant, and R-73 mounted on a railpylon from Mig-29C, which was mounted on a truck. No lasers, no active radar emissions. Total passive engagement. Probably assisted by consistent ingress and egress of F-15SA crews. Clever bastards. I suspect that Russians from factory helped them. RAND, Boeing, and Raytheon are still studying it. As F-15's MAWS (missile approach warning system) should have started pumping out countermeasures. Grudgingly we have accept that export versions of R-73 are capable of engaging F-15s. The Air-Air record of F-15C stands. No F-15 has ever been shot down by another aircraft. F-15E, on other hand, have been hit by SAMs, during Desert Storm, Allied Force, and OIF. But loss rates are low. Less then 10 airframes across 25 years.

So I did reinstall. So things work , many don't SU-25T appears to work. But if I fly 25T in Georgia, Free Flight. Then free flight again , in NTTR, DCS asks me to authorize NTTR? Go figure. KA-50 appears to work. FC3 does not work A-10C appears OK, for now. F-5E is completely dead. DCS freezes as soon FLY is pressed. In InstantAction/TakeOff in Georgia, sim starts, but a/c is dead with all warning lights lit. AKA Xmas tree. The barrometric altimeter reads 98570 feet. Mig-21Bis is a freeze, I can hear the engine, but DCS is otherwise frozen. I still love DCS, and remind myself that it is a beta. Like a cistertian monk. With a flaggelum whip. "Its a beta" SNAP!, "Its a beta" SNAPP!

Latest DCS Update 2.5.01508 broke F-5E. With exception of Instant Action/Caucass/Take Off , all F-5E missions crash DCS. The sole survivor an Instant Action/Take Off Caucass gives a broken F-5E siting on runway. The engines are dead, and cannot be started and bar.. altitude read 98055 feet with aircraft sitting on runway. SR-71 territory. All other F-5E missions freeze and/or crash DCS 2.5.0.15085

Hughes (contractor) promo vid for ARBS in AV-8B and A-4M. I haven't yet got AV-8B, I don't my rig can do it justice, maybe after my new one arrives. Anyhow, I found this vid that explains the USMC ARBS on Harrier II and A-4M. I also realized that a Russian version of ARBS, or something like it, is on our SU-25T, and SU-25 modules. Engagement seems similar. Locate target, lock it, lase it, hold down weapon release m fly the aircraft to release point, bomb is automatically released. Maybe not exactly the same, but similar enough. Personally, I prefer CCIP or CCRP in A-10C.

No. Stock clean install 2.5 . Was working then came this latest 2.5.015085 and it all went caput! I did repair. No change. This update threw everything over the cliff. No doubt ED is working on it. I did gt KA-50 to fly in Free Flight: Nevada. But everything else is frozen.

Yep. Worse for me. All modules crash upon entering flight. Including 25T. I tried DCS Repair, got no errors. Its just that 2.5.01.15085 update has killed the sim!

Now nothing works. When it comes to Fly after starting mission or instant action, pressing FLY freezes DCS. On all modules!

If I can help it, I prefer not use FC3 or 25T. Except SU-33 with Kuz. I want to use PFM ASM enabled platform only. So its either F-5E, A-10C, or Mig-21bis. Of three F-5E is most appropriate, untill we get F/A-18C or Goshawk variant of BAE Hawk trainer, for actual carrier quasi-quals. Damn I would love to have A-4M Skyhawk USMC variant with ARBS . I have not yet plunked for AV-8B, as my current rig is obsolescent for that module. But it is on my to-do list. Anyhow F-5E is right bow the best, subjectively, module for DCS for training and tactics work. I noticed that if SA threat is SA-6 (Kub) or SA-8, or their Naval variants. Beaming the radar and/or missile while low to the weeds, is effective. Don't even need to punch CM. TOR, Tonguska are nearly impossible for me to shake, if I am in their engagement range. S300 is doable in F-5E, by staying low. Hawk is almost ineffective, provided I stay low. When I mean low, I mean 100 AGL. Never below 50, even at 100 AGL it feel like I am 10 ft AGL. Above 200 feet AGL SA-6 and 8 are deadly. Also deadly if SAM radar and launcher are at higher AGL then F-5E. For MANPADS, with skill set to RANDOM, AI usually fails to take the shot. This is reverse of A-10C, where MANPADS almost always shoot. With F-5E, dusk or dawn attack, and MANPADS don't engage. What I find difficult is that if I am trying to bomb vehicle column with Snakeyes, I have very poor accuracy with snap runs. The sight is such that I have to be in 20 or 30 dive, and at 400knots, and release at 1500 meter slant range. So I try to bomb by eye, with DEPR at 80 mil(down) and eyeball the release at 200-250 feet. With MK-82 , I frag my self, so MK-82HD is required, or rockets/guns. Even with rockets and guns with DEPR at 15 (DOWN) and at 200 or below AGL with 400 knots, there is very little of downnose allowed for engagement. So in DCS with F-5E is like learning to fight with bare rifle with iron sights only. With A-10C with CCIP, CCRP, and Guns, is like an M4A1/M16A3 with combat optics. Much more precise and flexible. F-5E is old school air-ground work.No sensors, no lasers, no Mavs. Just iron bombs with PD fusing, rockets, and guns.

I think its a little tad too dark. The A-10C cockpit with defered shading. Canopy glass acts as a sunglasses at night, there is considerable darkening when viewing terrain through canopy glass then outside. Flood lighting inside A-10C does not work with defered shading ON. Some buttons, switches, and dials are tottally black, and difficult to find. NVG performance is completely different with defered shade. Is it really defocused everything inside cockpit, except of HUD and outside of canopy, and gives pilot circular non-stereoscopic porthole of a sight? If that how it is in RW then thats a dangerous system. I get it that ED programmed adaptation to night visibility of human naked eye, into the Defered Shadeing engine. Takes a few minutes flying to adapt. Excellent touch.

OK, I think I see the issue. Without deferred shading, tree opacity maps are not being used. The trees appear brighter then terrain and appear as growing out of the ground per LOD. Trees also grow out in Differed Shading engine,but in DS the opacity map masks the visual artifact present in non-DS graphics mode. THe visibility range needs to be on high or better, otherwise even with DS, viewer sees trees grow and scale up. Oi'vey. Now is a bad time to purchase new system , as GPU prices have gone way up. 1080Ti is more expensive then Quadro P4000, nearly $1300. TitanXP Collector's edition is 1600. Workstation card price point. Unfortunately, I need Quadro for my home station. Realflow, and Maya's VP2 don't play nice with GTX. All sorts of issues. My incoming box has P4000 speced out. Go on Amazon, Newegg, TigerDirect, even ebay. Older model GPU , Maxwell based quadros, and 980 series boards, are priced higher or almost as MSRP , when those boards were introduced. Lowly GTX 1070Ti is tad lower then P4000. These prices are insanely high.

Thank you.

Accidentally enabled MFCD on one monitor, and now I cannot it off or make changes as OK and CLOSE buttons are off screen in the black. I need to reset back to 1 Screen. Now do I undo the damage. MFCD layout sets cockpit and world into a center strip, and displays on each sides. Unfortunately it truncates home screen and settings, such that settings cannot be changed. I guess its one of those quirks of DCS that was neglected.

Unless one cheats with F10 or marker on kneeboard, of which I am guilty on both counts, ground is necessary for dead reckoning navigation. TAS, ground speed, and cockpit chronometer, are instruments that are used to navigate. A map also helps. A-10C, KA-50 have moving map display. A-10C hud has time to waypoint to keep you on speed. in RW , gnd spd is what is used, unless one has GPS. KIAS (knots indicated air speed) is necessary when flying. F-5E requires 400knots for ballistic computer to deliver AG weapons. My opinion is that navigation in DCS is very easy. Too easy in fact. But sophisticated navigation procedures does require sophisticated weather and meteorology implementation. For a consumer sim, it would just suck away performance, and give developers headaches. Their time is better put into improving flight models, optimizing speed of existing features