HP's Reverb VR Pro Headset

I'm still confused about Imacken and his continued use of Win 10 (build 1903). We know there is a decidedly fuzzier image with that build and I don't think there is a hotfix available yet. I think those with build 1809 are getting a clearer image with Reverb. Yes?

What does this alleged bug in 1903 affect? Is it WMR settings only? Where is there a source of info about this'?

Glad to see you're getting such good results with version 1803. Its the same version I have, and my Reverb will be here on Wednesday. Just tried DCS 2.5.5 on the Rift, and even though the VR performance update got broken and didn't make it into the latest patch, the performance in VR is still tremendously improved over 2.5.4 and 2.5.3. :D

Yeah the performance with the reverb is absolutely fine. Low level flying over the caucasus map with dense forest in the f14 is really smooth. I have no complaints

What does this alleged bug in 1903 affect? Is it WMR settings only? Where is there a source of info about this'?

You've missed the references in this thread mate...

https://forums.eagle.ru/showpost.php?p=3940685&postcount=1290

https://forums.eagle.ru/showpost.php?p=3940844&postcount=1303

https://forums.eagle.ru/showpost.php?p=3940560&postcount=1277

There are also links within those posts if interested..

You've missed the references in this thread mate...

 

https://forums.eagle.ru/showpost.php?p=3940685&postcount=1290

https://forums.eagle.ru/showpost.php?p=3940844&postcount=1303

https://forums.eagle.ru/showpost.php?p=3940560&postcount=1277

 

 

There are also links within those posts if interested..

Thanks but I didn’t miss the references in this thread. I would like to see some other reference from a reliable source. Can’t find one anywhere.

Just wondered where all this comes from.

Also, I can’t really make sense of the ‘bug’ as altering SteamVR SS and PD values make a big difference. So, not sure where this ‘bug’ is affecting anything in DCS.

Well I think I am getting screwed by pre-ordering with Amazon. Friend ordered from the HP store (US) the "pro" model and just got his shipping notification. I think all of the Reverb's are going to the "pro" edition and those of us that ordered the consumer edition are getting shafted.

Anyone else order with Amazon and sticking with it? I may have to order from ShopBLT as a backup. Amazon has been getting worse and worse with high end electronics.

Well I think I am getting screwed by pre-ordering with Amazon. Friend ordered from the HP store (US) the "pro" model and just got his shipping notification. I think all of the Reverb's are going to the "pro" edition and those of us that ordered the consumer edition are getting shafted.

 

Anyone else order with Amazon and sticking with it? I may have to order from ShopBLT as a backup. Amazon has been getting worse and worse with high end electronics.

Might be best to stick to your guns and wait it out at this stage. Who knows what stock is going where right now. I ordered the consumer edition but they shipped the pro version, and that was ordering direct from hp

Well I think I am getting screwed by pre-ordering with Amazon. Friend ordered from the HP store (US) the "pro" model and just got his shipping notification. I think all of the Reverb's are going to the "pro" edition and those of us that ordered the consumer edition are getting shafted.

 

Anyone else order with Amazon and sticking with it? I may have to order from ShopBLT as a backup. Amazon has been getting worse and worse with high end electronics.

I dropped my Amazon pre-order as soon as I had order confirmation from HP. The pro version was what I really wanted to begin with. The expected ship date was 6-18-19 with delivery on 6-26-19, but they shipped much earlier, and I will have mine on Wednesday. As far as ShopBLT goes. Be careful there. They have already pushed their expected ship dates back with people in this thread who have already given them money. Missing deadlines, and apparently, increasing prices in the middle of the deal is something a lot of negative reviews have mentioned about ShopBLT. Could all be hysteria, but I'd stick with ordering from HP if I were in your shoes. They're a known quantity. Until HP has more stock though, I'd keep my Amazon pre-order, since there's no financial risk to doing that.

Thanks but I didn’t miss the references in this thread. I would like to see some other reference from a reliable source. Can’t find one anywhere.

Just wondered where all this comes from.

Also, I can’t really make sense of the ‘bug’ as altering SteamVR SS and PD values make a big difference. So, not sure where this ‘bug’ is affecting anything in DCS.

The WMR setting 'renderTargetScale' being stuck on 1.0 on 1903 is separate from SteamVR SS and PD, as it is something that happens before that stage. Think of it as the 'source' image that happens before any SS. The new default in the settings file is 2.0, but it's ignored if past Win10 1809.

I don't think you'll find someone from Microsoft to give you a reliable source that they caused a bug, other than the links that were provided to you. If you don't see a difference between 1809 and 1903 then don't worry about it and sim on.

A Microsoft rep on reddit chatted about what the value is used for, to the extent that it got changed by default to the new larger value, just because people could have clearer displays:

Again, there is no real way to prove this person works for Microsoft.

But its just WMR and doesn't impact Oculus, steamvr, etc?

The WMR setting 'renderTargetScale' being stuck on 1.0 on 1903 is separate from SteamVR SS and PD, as it is something that happens before that stage. Think of it as the 'source' image that happens before any SS. The new default in the settings file is 2.0, but it's ignored if past Win10 1809.

 

I don't think you'll find someone from Microsoft to give you a reliable source that they caused a bug, other than the links that were provided to you. If you don't see a difference between 1809 and 1903 then don't worry about it and sim on.

 

A Microsoft rep on reddit chatted about what the value is used for, to the extent that it got changed by default to the new larger value, just because people could have clearer displays:

 

Again, there is no real way to prove this person works for Microsoft.

Thanks again. Let’s see what happens with this, it’ll be interesting.

But its just WMR and doesn't impact Oculus, steamvr, etc?

Yep, just the Windows Mixed Reality for SteamVR part. If you have a HTC Vive it would use SteamVR direct, and if you use Oculus it would either use the Oculus SDK or SteamVR.

On the Samsung Odyssey Plus the bug is quite noticeable, because that uses the WMR for SteamVR app to run DCS via its SteamVR API.

I'm pretty sure it will be fixed and rolled out in 1903, but sometimes it's hard to find out what fixes are in each plain Win10 update. Hopefully not long though.

Btw, this is all of us just trying to figure it out as we go. I wrote some stuff here to try to gather what I've found out so far: https://forums.mudspike.com/t/vr-news-hp-reverb-second-gen-headsets-are-on-the-way/8165/266?u=fearlessfrog

I’m still confused by this. If renderTargetScale is defaulting to 1.0, then surely just hiking SteamVR SS or PD values on top will have the same effect as if it was 2.0.

Yep, just the Windows Mixed Reality for SteamVR part. If you have a HTC Vive it would use SteamVR direct, and if you use Oculus it would either use the Oculus SDK or SteamVR.

 

On the Samsung Odyssey Plus the bug is quite noticeable, because that uses the WMR for SteamVR app to run DCS via its SteamVR API.

 

I'm pretty sure it will be fixed and rolled out in 1903, but sometimes it's hard to find out what fixes are in each plain Win10 update. Hopefully not long though.

 

Btw, this is all of us just trying to figure it out as we go. I wrote some stuff here to try to gather what I've found out so far: https://forums.mudspike.com/t/vr-news-hp-reverb-second-gen-headsets-are-on-the-way/8165/266?u=fearlessfrog

You say in your link on mudspike that the Vive Pro reports correctly in SteamVR. Well that’s not true. See my post #1319 in this thread.

I’m still confused by this. If renderTargetScale is defaulting to 1.0, then surely just hiking SteamVR SS or PD values on top will have the same effect as if it was 2.0.

The renderTargetScale is used to make the scene bigger so it apply a distortion warp, so that it looks correct when viewed in the lens and panels. Think of it sort of like a fisheye image because we aren't viewing a flat 2D box.

Before:

After:

That buffer has a hard limit, so that renderTargetScale == 1: Limit is 1657x2065 per eye, while renderTargetScale == 2: Limit is 2072x2582 per eye. SteamVR (when not using WMR for SteamVR) uses a default 1.4x value.

What this means is that even if you use super sampling to increase the resolution sent to the headset, the 'source data' of what is up sampled is bigger or smaller (more detailed / less detailed), dependent on the renderTargetScale value.

Put another way, even if you request a more detailed image / higher resolution, it hard limits it at the that stage of the pipeline regardless. It still gives the headset more pixels, but it is upsampling something of a lower detail, making it look fuzzier. That's not a great explanation, but that's what I've got :)

You say in your link on mudspike that the Vive Pro reports correctly in SteamVR. Well that’s not true. See my post #1319 in this thread.

In that you wrote:

> my Vive Pro defaults to 100% which SteamVR reports as 2016x2240!

2016 x 2240 is 1.4x (the SteamVR render scale target for pre-warp, as mentioned above) of the native resolution of 1440 x 1600. So SteamVR 100% SS by default.

EDIT: I should expand on that a bit. It's like why the Odyssey Plus is clearer with a DCS PD of 1.0 and a SteamVR SS of 210%. The base resolution of the O+ at SteamVR 100% SS is reported as 1426x1779. If you increase it to 210% (2064x2576 per eye) then you can see it is taking the native resolution of 1440x1600 and getting close to that 1.4x warp distortion size. What that means is that in the center the pixels are the clearest because they are close to 1 pixel per 1 native pixel where you need it the most. For sims and clarity that helps a lot.

In that you wrote:

 

> my Vive Pro defaults to 100% which SteamVR reports as 2016x2240!

 

2016 x 2240 is 1.4x (the SteamVR render scale target for pre-warp, as mentioned above) of the native resolution of 1440 x 1600. So SteamVR 100% SS by default.

 

EDIT: I should expand on that a bit. It's like why the Odyssey Plus is clearer with a DCS PD of 1.0 and a SteamVR SS of 210%. The base resolution of the O+ at SteamVR 100% SS is reported as 1426x1779. If you increase it to 210% (2064x2576 per eye) then you can see it is taking the native resolution of 1440x1600 and getting close to that 1.4x warp distortion size. What that means is that in the center the pixels are the clearest because they are close to 1 pixel per 1 native pixel where you need it the most. For sims and clarity that helps a lot.

Thanks again. I'm trying to understand what you are saying, but still having difficulty in relating it to SteamVR SS values.

This is what I see:

Vive Pro - native 1440x1600, Steam SS 100% 2016x2240 which is the 1.4 factor you are talking about. OK, so far so good!

Reverb - native 2160x2160, Steam SS 100% is 1612x1572. Obviously way less than the native res. To achieve the 1.4 factor you are saying is best for clarity, the pixel count would have to be 3024x3024 (2160*1.4). That would mean a SteamVR SS of 380% according to the slider!

I'm clearly missing something obvious, as I don't understand why the VP 100% is 1.4 x native, and the Reverb 100% is 0.75 x native.

Anything to help my aching brain would be appreciated!

Someone tested Reverb in IL-2 BoX?

Can You share your experiance?

Anything to help my aching brain would be appreciated!

That 1.4x factor does NOT need to be the same for every headset. Why you essentially need that 1.4x SS in Vive is to achieve barrel distortion which is a counter to the inherent pincushion distortion issue in headsets.

I have talked about that before here (in a reply to him only apparently):

https://forums.eagle.ru/showpost.php?p=3940596&postcount=1279

All of this depends on the physical shape and type of the lens in the headset. Some headset need more correction, some less. And therefore this 1.4x SS need NOT be the same for every headset.

That 1.4x factor does NOT need to be the same for every headset. Why you essentially need that 1.4x SS in Vive is to achieve barrel distortion which is a counter to the inherent pincushion distortion issue in headsets.

 

I have talked about that before here (in a reply to him only apparently):

 

https://forums.eagle.ru/showpost.php?p=3940596&postcount=1279

 

All of this depends on the physical shape and type of the lens in the headset. Some headset need more correction, some less. And therefore this 1.4x SS need NOT be the same for every headset.

OK, so how is anyone supposed to know what the correct factor is for any given headset?

How do we explain the fact that the SteamVR 100% number is 1.4x native on VP and 0.75x native on Reverb. How can it be less?

This is totally confusing for the average user!

OK, so how is anyone supposed to know what the correct factor is for any given headset?

How do we explain the fact that the SteamVR 100% number is 1.4x native on VP and 0.75x native on Reverb. How can it be less?

This is totally confusing for the average user!

Only the manufacturers know what the ideal number should be. But the thing is they have to balance it with user hardware also. Let's assume the ideal number is 1.4x but that will kill frame rates. So they might make 1.2x the default which fits most users and then give you the slider to increase or decrease it based on your GPU.

There are also some bugs and keep in mind this is a brand new release and maybe the values will change with software patches.

Knowing all of this is good if you are curious but honestly at the end of the day,you should just try to find a balance between FPS and resolution that YOU are happy with. YOUR tolerance to barrel distortion or dropped frames might be different than mine. This is not an exact science because biological processes are involved which can be imprecise.

EDIT: I should expand on that a bit. It's like why the Odyssey Plus is clearer with a DCS PD of 1.0 and a SteamVR SS of 210%. The base resolution of the O+ at SteamVR 100% SS is reported as 1426x1779. If you increase it to 210% (2064x2576 per eye) then you can see it is taking the native resolution of 1440x1600 and getting close to that 1.4x warp distortion size. What that means is that in the center the pixels are the clearest because they are close to 1 pixel per 1 native pixel where you need it the most. For sims and clarity that helps a lot.

Only the manufacturers know what the ideal number should be. But the thing is they have to balance it with user hardware also. Let's assume the ideal number is 1.4x but that will kill frame rates. So they might make 1.2x the default which fits most users and then give you the slider to increase or decrease it based on your GPU.

 

There are also some bugs and keep in mind this is a brand new release and maybe the values will change with software patches.

 

Knowing all of this is good if you are curious but honestly at the end of the day,you should just try to find a balance between FPS and resolution that YOU are happy with. YOUR tolerance to barrel distortion or dropped frames might be different than mine. This is not an exact science because biological processes are involved which can be imprecise.

OK, that's all very well, and that is the approach I have always followed over the years. 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. I have not been aware of this, and have always assumed that SteamVR SS and PD do similar jobs, and getting the best 'for me' was a question of getting the balance right with SS, PD, MSAA, etc. etc.

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?

OK, that's all very well, and that is the approach I have always followed over the years. 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. I have not been aware of this, and have always assumed that SteamVR SS and PD do similar jobs, and getting the best 'for me' was a question of getting the balance right with SS, PD, MSAA, etc. etc.

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 am not sure about the DCS PD vs SteamVR SS thing. ED would have to tell us if there are any differences but I just assume PD just applies the factor to both the x and y axes meaning 1.4 PD is 196% more pixels. While SteamVR SS simply adjusts that final pixel number. There might be some other differences but I don't know about them.

Regarding the Reverb default values, like I said it may be to achieve balance. You have the headset, did you notice any barrel distortion at 0.75x? If not, then maybe the manufacturers decided that value is fine for most users. Like I said, even small changes to the lens can affect all of this.

Secondly, its also possible that 0.75x is not the correct final value. We know for a fact that some value are bugged in Windows latest version. Maybe it is to be fixed with the next patch? Maybe it is a limitation of SteamVR where Steam doesn't know the ideal value to apply for a Reverb?

These are just possibilities and guesses. I think that even Pimax has values less than 1x (based on reading other comments).

Another review on the Reverb form our friends over at IL2 forums:

https://forum.il2sturmovik.com/topic/43948-reverb-new-offering-from-hp/?do=findComment&comment=781289

I came back home late today from my vacation trip and had ONLY 30 minutes time to do a quick first test with the Reverb and IL-2 BoX using the Kuban map and a FW-190A8 and a Spitfire MK9 in some quick missions.

 

My System:

- i7-8700k

- 1080Ti

- 32Gb Ram

 

I owned the Vive Pro, Oculus Rift CV1 and last the Odyssey Plus. I never tried or experienced the old Vive or a Pimax 5k plus or 8k. So I can only compare with the 3 HMD´s I owned and used before.

 

VERY FIRST Impressions:

 

The installation was easy and quick, actually NOTHING to install, because I used the OS+ WMR before.

 

Reverb itself:

 

The Reverb is very small, much smaller than I expected, very light weight. It is smaller than a Rift.

I love the design, the baseball cap design in wearing. It is very solid, the comfort on my head is superb and fits easy.

I love the leather VR cover, very comfortable. My nose is happy now, because the nose is free and nothing squeezes, my OS+ hurt while wearing it, Reverb does not at all.

The cable is very thick and heavy, that was better with the Rift, but I like that it does not use HDMI of the GPU.

Display port works great.

The Headphones are comfortable and easy to adjust to your ear position.

 

I like that the Reverb is easy to lift up in my face, like lifting a sunglass. That makes it easy in IL-2 to watch my keypad. With the OS+ that did not work that great.

 

Visual Quality:

 

First thing I saw is my desktop in the WMR app and I thought "Wow is that crisp and clear, what a nice picture". A huge wow effect. I never saw such a great image in VR before, compared to the ones I used day and night difference and now REALLY VR 2.0 and not 1.5!!!

Same thing happened when I started IL-2, the menu....the hangar, just WOW! And then I started a quick mission on Kuban summer map in a FW-190A8.

I saw the cockpit and was hyped. Crystal clear picture every small details easy to see and read. Even small text inside the cockpit was easy to read.....what a difference. In the Spitfire I could even easily read the model number of the gunsight ( very tiny text ).

The colors are much more natural compared to the Rift and OS+. The screen is brighter, with OS+ I used Gamma 1.0, that is too bright with the Reverb, I changed Gamma to 0.9 now.

 

The black levels inside the Spitfire Mk9 cockpit looked really great and real black, not as I expected from an LCD, honestly I was not happy when I heard that it will not use OLED, but now after I saw the great picture quality I don't say it´s a disadvantage.

 

The image contrast is very good.

 

No screen door effect visible. I did not see any Mura or God Rays or anything else disturbing.

 

Overall the picture is just great and I saw so many details now of the plane which I never saw before in VR. Mind blowing!

 

My IPD is 63mm, so for me the prefect fit.

 

Sweet spot.........honestly I had no problem at all with that, sweet spot is bigger than in Vive Pro, Rift and OS+.

 

 

 

Sound Quality:

 

 

 

I love the built in Headphones. They are easy to adjust, comfortable and have a very good sound, much much much better than the OS+ and even better than the Rift, which I liked also.

 

Sound is powerful and very balanced. Just great for such small headphones.

 

 

 

Performance in IL-2:

 

 

 

As I said before, Kuban Summer Map, daytime afternoon with a few clouds in the sky. I did use Steam VR SS 150%. Did NOT change any graphic settings in the menu from which I used before with the OS+ and got FPS from 79-90. Most of the time stable 90, but sometimes around 83-85, the lowest I saw was 79. The picture was smooth with no stuttering or flickering. Performs even better for my opinion than the Rift or the OS+, maybe it´s the Display port?????? I am not a PC hardware Geek, I can only say my impressions, but honestly I was surprised how nice the Reverb performs, I expected big FPS drops and stuttering images, but nothing like that happened.

 

With the Rift using ASW I had worse FPS drops and the picture artifacts and the flickering image inside the german gunsight was horrible, with the Reverb nothing like that happens, everything smooth and crystal clear.

 

The tracking was also working very solid, no problems for sitting simulation guys like us. I will not use it for other VR games using room scale and moving around with the controllers, so for me the WMR concept is good and really my cup of tea.

 

 

 

That's it for now, I will do much more testing next days, will try some different graphic settings and SS in Steam VR and will try also some Sim Racing titles.....iRacing, Raceroom and Assetto Corsa.

 

 

 

I try to install the Steam APP for the FOV test next days aswell.

 

 

IMG_5463.jpg

I am not sure about the DCS PD vs SteamVR SS thing. ED would have to tell us if there are any differences but I just assume PD just applies the factor to both the x and y axes meaning 1.4 PD is 196% more pixels. While SteamVR SS simply adjusts that final pixel number. There might be some other differences but I don't know about them.

 

Regarding the Reverb default values, like I said it may be to achieve balance. You have the headset, did you notice any barrel distortion at 0.75x? If not, then maybe the manufacturers decided that value is fine for most users. Like I said, even small changes to the lens can affect all of this.

 

Secondly, its also possible that 0.75x is not the correct final value. We know for a fact that some value are bugged in Windows latest version. Maybe it is to be fixed with the next patch? Maybe it is a limitation of SteamVR where Steam doesn't know the ideal value to apply for a Reverb?

 

These are just possibilities and guesses. I think that even Pimax has values less than 1x (based on reading other comments).

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!

Another review on the Reverb form our friends over at IL2 forums:

 

https://forum.il2sturmovik.com/topic/43948-reverb-new-offering-from-hp/?do=findComment&comment=781289

I mean everyone is going to have their own take on things, but really, what he says about audio quality, blacks and colour range of LCD cf. OLED panels flies in the face of anything I have seen or read, or experienced myself. You only have to load up Elite Dangerous to see a grey universe to know about blacks on LCD panels!

I'm happy for him!

After 3 or 4 days of use, I'm still doing a lot of testing of the Reverb in various games, scenarios etc. not just DCS.