HP's Reverb VR Pro Headset

Ah got it backwards sorry. Weird that its seemingly better for you than him.

Yeah it's puzzling. If anything I've been pretty thrilled how much better the performance is with the Reverb than it was with the Rift, although the 2.5.5 update could be a significant factor here as well.

Or the fact he has a 4790k and IIRC you have 9x proc.

An i9 would make little to no difference over an oc`ed 4790k in DCS. My 5yr old 4790k is still nipping on the heels of intel`s latest cpu`s. Intel have been running on the spot for the last few gens, absolutely no value in upgrading from Haswell for gaming. AMD with their new Ryzen 3000 may be the best out there in the next few months while delivering a great price and worth consideration.

Certainly worth consideration , but the reported high price of even lower-end 570 motherboards is a concern .

..However, fearlessfrog seems to be suggesting that there is a sweet value for SteamVR SS, one at which the image looks better with no PD in DCS. ..

fearlessfrog's posts seem to suggest differently and I'm fascinated to know more.

If, as you say, only the manufacturers know what this value is, it seems like a nonsense! Why would this not be made clear for each headset?

Also, back to my question about the 1.4 and 0.75 for the VP and Reverb, I almost understand 'the value' being 1.4 or 1.6 or 1.8 etc., but why should the Reverb be LESS than native?

I really only meant to mean that because DCS PD and SteamSS effectively do the same thing and combine, then like Panthera_Tigris replied, it was just easier to change just one 'end' of the A + B = C equation, especially as the DCS PD is quadratic and the SteamSS is its only funny scale. At the end of the day PD 1.6 and SS 0 or PD 1.0 and SS 190% is all good if it means the center of the warp is clearer for people reading gauges etc.

I'm sure that's all true. The whole concept of an custom factor for each headset to counteract barrel distortion is a nonsense if no one actually knows what the correct value is!

I think it's because in a lot of common games the ability to read cockpits or text clearly isn't so much of a big deal. We're a bit of a niche, so the WMR people picked low defaults out of the box that are more suited for something like Beat Sabre than DCS.

Edit: one thing I did notice in IL2 was the mouse not working in VR, I had to prop the Reverb on my forehead and navigate with the mouse on the mirrored image on my monitor, so a bit of a glitch there, although that might be due to inexperience with the WMR/IL2 combination.

WMR is a pain (and like imacken said, roll on the native implementation in DCS) but one tip I got used to do is (1) put on the headset (2) in the WMR 'house' use the Desktop widget to click the DCS icon. If you do it the otherway around then WMR gets confused on what 'owns' the mouse pointer (due to the sensor in the headset detecting if worn or not). The WMR bug here seems to be it gets out of 'sync' with that, so doing the above steps every time makes it work reliably.

interesting points but just wondering what your suggested settings are for the reverb and the DCS in terms of custom resolution (in steam VR) and also PD in game? (have been running 188% custom res (on the basis this aligns the res with the native) and 1.0 PD in game?

I personally would start with DCS PD at 1.0 and try 188% and see how that goes. Because we're talking personal clarity then we're going to see different people seeing different results, due to IPD, DCS settings like MSAA used plus even different versions of windows, WMR, WMR for SteamVR and SteamVR itself! :)

 

 

WMR is a pain (and like imacken said, roll on the native implementation in DCS) but one tip I got used to do is (1) put on the headset (2) in the WMR 'house' use the Desktop widget to click the DCS icon. If you do it the otherway around then WMR gets confused on what 'owns' the mouse pointer (due to the sensor in the headset detecting if worn or not). The WMR bug here seems to be it gets out of 'sync' with that, so doing the above steps every time makes it work reliably.

Yeah, the mouse thing is a bit annoying and worrisome since I use the mouse to run the pit. And It makes me wonder how point control is gonna do with it.

As for the A+B+C thing for res, is it really like that or are there multiple processing steps?

I.e.

is it A+B+C = final output

or A then B then C = final output

Certainly worth consideration , but the reported high price of even lower-end 570 motherboards is a concern .

The 570 mb`s will come down in price after a while, but they bring some big improvements. PCIe 4.0 and AMD Ryzen 3000 series CPUs would be able to hit a maximum of 5000 MHz DDR4 memory speeds with X570. That will do wonders for latency and VR.

If the reviews back up the hype, I`ll likely go AMD 3000/X570 for a new build, we`ll see how the Black Friday sales go.

An i9 would make little to no difference over an oc`ed 4790k in DCS. My 5yr old 4790k is still nipping on the heels of intel`s latest cpu`s. Intel have been running on the spot for the last few gens, absolutely no value in upgrading from Haswell for gaming. AMD with their new Ryzen 3000 may be the best out there in the next few months while delivering a great price and worth consideration.

Its not all about the GHZ, things like memory interface, cache size etc. matter quite a bit... I saw a pretty large performance improvements going from a 4x to 9x.

Yeah, the mouse thing is a bit annoying and worrisome since I use the mouse to run the pit. And It makes me wonder how point control is gonna do with it.

 

As for the A+B+C thing for res, is it really like that or are there multiple processing steps?

 

I.e.

is it A+B+C = final output

or A then B then C = final output

That's a fair point, only ED could answer that for sure. In things like Unreal and Unity the target resolution is requested from the driver (think of it like how a monitor would enumerate what resolution is was supporting) and then the game usually uses that to make up the final scene size.

DCS was quite early to VR in some ways, so they offered a in-engine setting of their own as a multiplier. SteamVR used to use a different scale similar to DCS's, but changed it to the percentage when they realized that different headsets had different aspect ratios, so the Oculus way of '1.2' was too large a set of steps I guess.

I've never seen a visual difference between A then B etc, but would love to see it if it can be worked out.

The 570 mb`s will come down in price after a while, but they bring some big improvements. PCIe 4.0 and AMD Ryzen 3000 series CPUs would be able to hit a maximum of 5000 MHz DDR4 memory speeds with X570. That will do wonders for latency and VR.

 

If the reviews back up the hype, I`ll likely go AMD 3000/X570 for a new build, we`ll see how the Black Friday sales go.

Thanks for the info . I'm planning an upgrade and considering AMD for the first time in decades .

Its not all about the GHZ, things like memory interface, cache size etc. matter quite a bit... I saw a pretty large performance improvements going from a 4x to 9x.

It`s very much about ghz and single core speed atm. Hopefully Vulkan will rectify that when we eventually get it. Add to the fact that intel chips are getting worse over time with an estimated up to 20% loss or more via security vulnerabilities - Zombieload etc. I have always had intel but now I`m seriously thinking of dumping them for AMD who have matched or surpassed them with their IPC jump, more cores and better security. Add Vulkan to the mix an it`s even more suited to AMD and the extra cores.

Intel had a very good run, but got sloppy and complacent. It`s good for them to get a kick in the arse now and then. Competition is good for us consumers.

Yea I'm curious what the newer cpu's will bring. I've had Intel as long as I can remember and I've thought about going to amd but would need to see how much that changes for me with the current cooling solution that I have. I'd probably still use an Nvidia card in the end.

Yea I'm curious what the newer cpu's will bring. I've had Intel as long as I can remember and I've thought about going to amd but would need to see how much that changes for me with the current cooling solution that I have. I'd probably still use an Nvidia card in the end.

AMD Ryzen 3000 is 7nm so will deliver more for less, Intel i9 is a 14nm processor, they haven`t even hit 10nm`s yet. Your current cooling will work fine, i 9`s are known to run quiet hot, I`d say the new AMD chips will be cooler.

Yeah it's puzzling. If anything I've been pretty thrilled how much better the performance is with the Reverb than it was with the Rift, although the 2.5.5 update could be a significant factor here as well.

I am on stable branch......

That link is the same as what I said to you in an earlier post. It's called motionReprojection in WMR, not motion smoothing. That is a SteamVR setting.

yes got it - as far as I am aware don't think you can enable motion smoothing for the hp reverb? (but might be wrong?)

I did enable the motion Reprojection in the config file and it has definitely improved things (it seems to be by default disabled) - strangely before I enabled reprojection was showing in steamVR has on but from your other posts guessing this must have referred to the steam version of it......

I received shipping notice from Fed Ex on Saturday and my Reverb arrives on Tuesday 6/11! Ordered on 6/1 direct from HP. I am going from a Lenovo Explorer 1440 x 1440 to the Reverb.

I got received shipping notice from Fed Ex on Saturday and my Reverb arrives on Tuesday 6/11! Ordered on 6/1 direct from HP. I am going from a Lenovo Explorer 1440 x 1440 to the Reverb.

With all of the testing that had been done already with 1080 ti's and plain 2080's and the Reverb/DCS I am not too worried about performance at this point.

Will wait for you feedback. Since you have a 2080 it would be nice to get feedback on how it runs.

With all of the testing that had been done already with 1080 ti's and plain 2080's and the Reverb/DCS I am not too worried about performance at this point.

Not sure if its just my luck but started playing around with DCS PD vs SteamVr SS. With the o+, setting the super sampling with DCS PD, I get a lot of ghosting and video defects on the headset and more micro-stuttering. I get a much more stable video feed to the HMD with SteamVR SS turned up and DCS PD @ 1.0. For example BFM with an f14 with PD turned up makes the f14 constantly shimmer and ghost during the fight. I don't have that issue with SteamVR SS, and the visual clarity seems on par.

Current settings - PD 1.0, SteamVR 180% and MFAA and MSAAx2 on. I set textures other than water (@medium) to low, no grass, half trees, shadows flat with both object distance and preload set to extreme and max respectively. MotionVector is on in Steam for re-projection.

Anybody else seeing similar results? This might help on the Reverb since the render pipeline is the same.

I hadn't noticed a big improvement in the ghosting and jittering when I switched from the 1080ti to the 2080ti, so I did a lot of troubleshooting and found that most of the issues I was having were in one aircraft only (the A-10C) and only when I had images displayed on the MFDs from the TGP and MAV. As of the 2.5.5 update, that issue has all but been eliminated. Its still there, but once things get loaded into the VRAM on the card, it gradually goes away and stays away. Must be that VRAM memory leak that they fixed with the new version. If I knew back then what I know now, I probably wouldn't have upgraded from the 1080ti. Granted, the quality of the image and sense of distance seem to be improved with the 2080ti, but I'm not sure its worth $1500.00. Oh well. Hopefully I see that difference in other games.

Tomorrow AMD is representing new Navi GPU. The top model will be RX 5900 for 600$. This will be matched to RTX 2080 Super by performance but will have 16Gb GDDR6 and will be PCIe 4.0 so this could have some benefits for VR.

So combining with Ryzen7 3700X 8/16 could beat I9 9900 and RTX 2080ti on Vulcan engine.

Point is the price. it is actualy100$ less than Nvidia pair. So I believe Nvidia will lower the price as well and AMD didn't go with the lowest price just to leave headroom for further lowering if Nvidia answers on this with cutting prices.

Definitely, it is not the time to buy new GPU or CPU yet, but after summer ... :pilotfly:

I ended up with nearly the same thing with the same results but adjusted the display quality in the WMR driver instead of SteamVR which is somewhat coded to vendor specific headsets (the May update affected the Lenovo Explorer among others). I have the display quality now set to Very High (beta) and my other settings are very similar. ED said they will eliminate Steam VR eventually for the Reverb. It is smooth as glass.

I will be trying to get the Reverb going on Tuesday.

Not sure if its just my luck but started playing around with DCS PD vs SteamVr SS. With the o+, setting the super sampling with DCS PD, I get a lot of ghosting and video defects on the headset and more micro-stuttering. I get a much more stable video feed to the HMD with SteamVR SS turned up and DCS PD @ 1.0. For example BFM with an f14 with PD turned up makes the f14 constantly shimmer and ghost during the fight. I don't have that issue with SteamVR SS, and the visual clarity seems on par.

 

Current settings - PD 1.0, SteamVR 180% and MFAA and MSAAx2 on. I set textures other than water (@medium) to low, no grass, half trees, shadows flat with both object distance and preload set to extreme and max respectively. MotionVector is on in Steam for re-projection.

 

Anybody else seeing similar results? This might help on the Reverb since the render pipeline is the same.

HP's Reverb VR Pro Headset

So combining with Ryzen7 3700X 8/16 could beat I9 9900 and RTX 2080ti on Vulcan engine

what is this comparison based on? number of processing cores?

Looks like all my hopes are now on the Reverb. After finishing my new TOTL computer for exclusive VR use, I was finally able to give the Pimax 5K + a try.

 

 

It was horrid. I honestly thought the rez was just as bad as my CV1? I CAN'T believe I spent 3 Grand on a new computer for this? I need to try some other "games" I guess, But first up was Xplane, and it put me in a big hanger with aircraft, that from a distance, looked like big white blobs.

 

 

If Imacken is right and The Reverb isn't dramatically better than Gen 1.5 such as the 5K+, then I'll go triple monitors. because from what I can tell Gen 1.5 sucks hard. Extremely low rez with Moire effect everywhere.

Oh, something went horribly wrong there. The difference between CV1 and P5K+ is pretty big, IMO. And don't forget, you have to adjust the physical IPD to make sure it's clear for you eyes.