Gunnars Driver

Good work Thud!

OK, it looked free for at least iphone. didnt DL as I bith know my ipd since before and already have the euemeasure app thats seems giving proper results. My friend had measured hes ipd to about 75 mm and because of this had excluded all VR Headsets except pimax (8KX). When we did measure him with the app it was 70.7 and thats much lower and probably a lot more vr headsets would be possible to use for him. Just saying that a improper measurement can lure one to make the wrong decision. Of course 72 mm is possible, your measurement could be correct. That would take you to the 99 percentile for ipd.

It of course depends on what CPU and at which clock speed together with what GPU at which clock speed. And what you play, a highly crowded server or single player. And which VR Headset you run. And of course its not only one setting but the combination of settings.

I did a few tests with my 2080ti together with the 8KX, but as my 3090 arrived only a few days after the 8KX, I didnt tune anything. The 3090 is fine even with some high settings in normal FOV but the 2080ti do struggle. It would probably had been possible to get it to run ok with lower settings. I guess a 1080ti would have a hard job but with low settings and not widest FOV it might work ok.

Ieyeque PDcheck. I didnt have that good luck with the mirror ipd measurement, still I knew my real ipd so I think I would have it easier than not knowing. the app I used say ipd within 0.5 mm accuracy and my result was exactly the same as my real ipd.

Did a online flight this evening, on [Aerobatics Online Kaukasus]. There was 40 or more aircrafts(of 70) and I only flew formation flight with others for 13 minutes exept for engine start and take off and landing. What we can see is that the frame times and CPU/GPU loads are not that different and that my computer isnt CPU or GPU bond but instead being the limit shifts between the CPU and GPU. Im quite sure the highest numbers on the CPU comes from the numerous screenshots I took. I do not think the CPU would have been above 22ms(=45fps) at anytime except for the screenshots. I guess that 40 aircrafts and flying around with 10 other aircrafts rather close should be somewhere like the typical multiplayer mission when it comes to CPU and GPU load.

I use the keyboard for all F-commands: -Reload, refuel at FARP. - In Air refueling -Cat launch on the carrier -Change view,( weapons released to see it hit) or look at the map. Screen shots. Recenter the VR Its not that un-immersive if you ask me. In real flight, specially night flights with NVG you cannot even see al switches because of the NVG so you learn to find them and navigate the hand by different references(also true for some daylight profiles when you dont have the time/resuorces to look when needing to föip a switch) This can also be done on the keyboard as well and thats how I do it.

Yes. I have 65.5 and the app I used also gave me 65.5, repeatedly. I did look far away during the app measerement. It was possible to get the value a bit off if I looked close.

Yes, these home tests seems quite unreliable. For starters, anything you do when tou look at something close( lets say less than 8-10m) will not give you the far ipd but it gives something between far ipd( which is the ipd we normally mean when using this term. The closer you look at something, the more inwards is the eye angled, giving a smaller ipd number than the actual distance between your eyes. I’d say, in most cases the IPD you set in VR is the far IPD. Together with this above we have the unreliable measurement. Of course, if you have a VR Headset and dont know the IPD, it might be better to try to measure than to only do a wild guess. I had my IPD measured since long time with quality equipment, I tried doing it as some of the recommendations, and didnt get that close. I have 65.5mm, its 33.0+32.5mm from the middle/nose. Results 62-63 when doing mirror measure. If you can get a app, i found at least one for android.

IPD Measure: I have my values from the work, measured with state of the art systems. When I try the recommended mirror etc I get it quite wrong. I tested an iPhone app [Eye Measure] and it actually gave me the correct IPD repeadetly. I recommend using that app. my friend had gotten values about 75 mm for pwn tests but the app said 70.7 mm. Quite big difference. If you are well above 70 I dont know if you have any choises except Pimax. Spec say 75 mm I think, my 8KX goes to around 73mm and there is a electronic offset that probably can do the rest. Also, many seems to get best results with pimax if setting a ipd lower than actual= up to 75 should be fine. I dont know but I guess the electronic ipd adjustment will be still there for Reverb G2 as it is a feature of WMR. Think there was a limit that could be overrided if tuning some .ini files for WMR.

[Heat blur] is one of those, isnt really made for VR I think, ahould be set to off. I think [terrain textures] is almost not detectable from VR, so it can be set to low. theres a lot about aettings but we can cover this when you have got the vr set. theres some very useful links with info also.

The Reverb ”G1”/I have the pro version of it) and the upcomming reverg G2 have the exact same resolution and because of this the hardware performance needed will be more or less exactly the same. You can use current data for the G1 to compare for G2. As DCS is a bit different when it comes to pc hardware specs than other games we can not use the official pc hardware requirements only(posted below). Also I would say ”half resolution” is a bad term. You wont run it in half resolution but you might end up with supersampling lower than 100%, thus using a resolution thats lower than native....you wont need to go even close to half. You probably can use 100% SS and thereby use the native 2160*2160. Some of the settings in DCS are used to decrease the jagged lines, for example MSAA but practical SS also function about the same way. Using the right settings to achieve the same graphic level will reduce the load on the CPU and GPU. There is also settings that actually take a big hit on the CPU&GPU but that doesnt work in VR...these shouldnt be used. With the right settings you can get good results with the 2070 super together with HP Reverb, we already can draw this conclusion from the G1. The most limiting part will probab ly be the CPU clock speed, and keeping settings in pair witht the cpu resource will make it playable. Playing on very crowded servers with a lot of activity etc will tend to max the CPU, we see this on multiple threads in this forum. Seems to happen to many people and it is not limited to slow CPUs appearently. https://www8.hp.com/h20195/v2/GetPDF.aspx/c06630565.pdf

This thread cover Reverb + RYX2070/2070 super. New Reverb. Which GPU? - ED Forums

1) as G1 and G2 have the same resolution its possible to test how it will look in the G2 when it comes to ’undersampling’ (havent done it). you probably wont need to undersample with a 2070 super, but you might have to stick with supersampling at 100% ( = the reverb resolution of 2160*2160). Also you might need to lower the settings for threes and grass. Supersampling increases the sharpnes when going above the panels physical resolution, higher supersampling/resulotion gives a sharper picture. I do 150% with the reverb, 200% was even sharper but it made a too big impact in FPS. The same function should be for undersampling, the sharpness would progressively decrease when increasing undersampling. A qualified guess is that you can stay att 100% with the reverb G2 for your system. some settings affect the CPU more and some the GPU more. For your 4.5 Ghz max clock, its possible to use and you may need to lower some settings further or have some stutter you otherwise hadnt had. Just make sure the CPU stays at 4.5ghz during the whole gaming without clocking down. Probably already covered as you have OC:d already. As from what I did read about not possible to both upgrade and get VR, I would go for VR with the current system tuned as good as it can get and then take the 2) FS2020 we dont know about VR yet but it also currently more or less use one cpu thread for the main core= easy to get cpu bound. theres lot of settings so you should be able to loser these to get a smooth ride. I expect Microsoft to fix the single core thread issue thus take advantage of cpus with multiple cores, but when this will happen is not clear. Both FS2020 and DCS will need quite low settings. Still, the immersion in VR is good. DCS developer works on a multiple cpu solution but it seems to be one year or more away. DCS has good levels in the settings and the main problem isnt to find lower settings but that the users usually set to high settings.

I have the Reverb since one year and did run it with a 2080ti until yesterday(now 3090 :-) ) The hardware performance needed shouldnt differ between G1 and G2(same resolution) and my 2080ti could easily handle the Reverb G1, I use 150% supersampling. I think the 2070 super is ranked as about 25-30% lower performance than my 2080ti, so it isnt a very weak gpu. Just keep the settings appropriate, not to much supersampling, shadows flat etc and the known fps killer settings low or off and it will be fine. Its also possible to set the Reverb frequency to 60hz, which gives the possibility to use 30 fps as a target with motionreprojection to 60fps. That would allow for even lower hardware performance if needed. The main thing for OP is taking advantage of the CPU Possibility to overclock and make it shine to the level that it mostly isnt the limit for performance. If we play with the thought that he reaches 5ghz or close to it, it will perform DCS-wise about as good as any cpu. Then there wouldnt be any big win even if buying a new CPU(+ mobo, ram etc). I know overclocking isnt so easy to get a good grip asap, but youtube should be a very good help, 4770K was commonly used in OC and the actual motherboard settings might be covered in many videos

High cpu clock speed is much more important then high RAM speed in DCS. High ram speed does make a difference but marginal compared to the cpu clock. Also, you would like to have the high clock freq on all cores. And you might be helped by disabling hyperthreading, thus making each physical cpu core a bit more powerfull. Now you now what to hunt during settings :-) Reverb G1 and G2 will have the same resolution, 2160’2160 for each eye( if they didnt change the specs for G2 very revently). Main difference should be better sharpnes in the lenses outside middle focus on the G2. G1 is very sharp in the middle but a bit blurry on other spaces. Also more cameras for the inside out tracking(which is a problem/issue with G1 for me at least). I would recommend tuning of current hardware and getting the reverb G2. I would say the 2070 super can do it, but probably nbot with the highest settings, and also the cpu might be the limit anyhow if you not tune it. My neighbour have a low end rig just fullfilling the specs for VR and DCS has to be set at very low graphic settings and also its more or less a slide show, at least when we do multiplayer with lot of assets. This with a oculus rift.

Speed-of-heat is sort of right. DCS is harder on hardware than most other VR games. The recommended hardware, if you are on the margin you will have either very low settings or a slide show. You can use your hardware, you should make an effort to OC the CPU. I had my old CPU(3770K) at 4.5ghz for a couple of years and was happy with it before getting VR, and all other games run fine but DCS needed more CPU speed. I had the oculus rift since the release. I did increase the overclock and found it working progressively better the higher the OC was and ended up at 4.8ghz as the normal setting. Even higher clock was better, but to have it run safe for the wife when I was away at work I stayed at that value. You also should make sure the settings keep the cpu at the set clock during the complete gaming period. Default setting for turbo mode is turbo clock for a while, then clock down and it then clock up for a while etc. DCS need that clock speed all the time, if it clocks down to 4ghz or so you will have stutter. You can test this at the same time as you stress test the overclock of your computer. HWinfo64 of something similar can show you the clock speed of all cpu cores dureing a stress test, and if you use a temperature software you can keep an eye on the temps as well. Your watercooler should be fine, maybe new good paste on the CPU and make sure the coller is free from dust etc.

Try set the visib range to at least high or ultra. Might take a hit on fps, but I think it'll work. The eye seight value 20/20 is at longer range, are you sure you se very sharp at let say 1,5 to 2 m? I have >20/20 at range(airforce pilot) but due to age the close visual is degraded so I use reading glasses to read. With Oculus Rift and HP Reverb I see very sharp but recently I got the Pimax 8KX which has the focal length somewhere below 1m. I need reading glasses in these to see sharp. Without it is quite blurry. The focal distance in the Index is probably more then in the Pimax, and I would guess about 2m, thats the reason to check how sharp you se at 2m.

Do you have the visibility range in DCS set to ultra ? Otherwise, it could depend on either your Supersampling setting, the personal ability to detect/spot or the difference between the eye sight sharpness ?

Thats my experience as well. If not using ASW you have stutter which is much worse in my eyes than the ghosting ASW can produce. This is a question about personal reference, so OP have to find hes own solution. For the HP Reverb(and probably all WMR headsets), it do not use ASW but a similar technology, motionreprojection. The difference is that its much better not to produce ghosting. In many cases I dont notice the difference between motionreprojection on or of when it comes to the bad things, but it takes the stutter away. The OP GPU is probably to weak to be able to reach 90fps in any case so that isnt really a possible way. ASW/motionreprojection is needed to get out of the stutter when the frames per second is lower than the headset framrate, or accept the stutter if OP finds ASW/motionreprojection unacceptable.

Yes, I agree. The 4770K should overclock quite well. With the right CPU-cooling/chassies cooling it should be able to reach about 5ghz, and if it does it should be more or less as good as any new CPU. I went from a 4.8ghz overclocked 3770K to a 9900KS one year ago and there wasnt that much win (for DCS spefifically). As I see it, with intel CPU the stessing spikes in DCS tend to reach 100% on the core doing the DCS thread if the CPU clock is ”only” 4.5-4,6ghz, and when going above this freq, say 4.8 or more moste of the CPU spikes will be within the CPU calculation power. (Not talking about the worst online servers). The other thing (but actually first thing to check) is to make sure the computer is properly setup for gaming. This means, the BIOS settings allow the CPU to stay at the overclocked/turbo frequency during the whole gaming session and not to clock down after a minute or two. Also, it should run the turbo/overclocked freq on all cores. Having a little weaker CPU I think a good thing could be to disable hyperthreading. This will couse lower performance for some applications, due to going from 8 threads to 4, but for DCS this probably increases the performance of the CPU core driving the dvs main thread. Also, checking that the RAM memory actually runs at the clock speeds it is marked(commom that they run on the default DDR speed(could be 1666 instead of 2333 etc). The Reverb (G1) has a very sharp picture and using a lower resolution isnt as bad idea. The G2 probably will be at least as sharp as G1 so...

Before chaning these settings, make sure the cooling of your computer/CPU is up to it. Install som temperature software to see the cPU temps(core temp for example, as I use). There should be safety built in in the motherboard bios that protects the cpu, but to be sure, use a cpu temperature software and keep a look on the temps when tuning. Package Power Limit TDP (Watts) / Package Power Limit Time hould be set to "diabled" of something like that (the manual only states auto or enabled). Core Current Limit (Amps) could be changed, but I dont know this motherboard and the settings. Either we get a tip from someone else here, or you check youtube for youer motherboard(preferebly with the same CPU) and find appropriate settings. https://download.gigabyte.com/FileList/Manual/mb_manual_b460m-aorus-pro_1001_201014_e.pdf (the manual I looked in)

The "Intel turbo" is designed as to only give the full clockspeed at a limited time, after this the CPU should throttle down to keep within the limits of TDP. This means that hte CPU cooling requirement is set lower than it actual is for running the CPU at full speed over time. If you have sufficient cooling you can use hte higher clock speed(more heat) a longer time. A nice setup can keep it all the time. Some motherboards follow this Intel turbo guidance and doesnt keep the turbo speed over time. Some may, Ive only done Asus mobos and never even tried stock settings so...I'm not the best to tell for that part. Read the linked article for more info. The settings differ a bit between manufacturers(even between chips by same manufacturer and even between motherboards using the same chip), different names for same thing etc. Asus use: turbo mode parameters, long duration power package limit etc. Intel TDP Investigation: Boards Violating Turbo Duration (Z390) | GamersNexus - Gaming PC Builds & Hardware Benchmarks

What motherboard do you have? (one thing noticed after the update of the forum, no attatched info....which is very good when supporting in a forum. :-( )

Yes, I think so. Have been looking around because I was curious but now I know :-) Time to see how it performs with my 8KX.