VR Clouds any improvements form the few of us using VR?

[Jonnti46]

I'm really interested to know if the original problem of VR clouds has been fixed that I raised in June Last year v4 (clouds making you feel cross eyed as you get near them and clouds / coming into cockpit) - I can't really get any clear information that it has been looked at. It a shame as I have x3 different VR headsets and could have helped to trouble shoot inc PIMAX 4K. I'm told VR is a fraction of the user base, so probably not very important - but I'd like to know as I'm really not able to use v4 due to this. If anyone has information on the new version 5 working with clouds in VR I'd be really interested to hear feedback? The feedback from developer is (it would seem) they are not sure.... 

[Gildahl]

Same question here.  I purchased V4 for VR and was never able to put to any serious use.  Between the rotating clouds and (most significantly) the eye convergence in foggy conditions I ultimately put it on the shelf to wait for a fix.  The effect was extremely obvious, even just sitting on the runway in foggy conditions. For me, I was told that the developers didn't have an Index to test this in.  So I'd love if a user who had been experiencing these two issues in VR before could report on what things look like now.

[Edit] Just read the post here, which probably answers my question.  So will be holding off for now as this was precisely one of my most major issues...

 

Edited January 23, 2021 by Gildahl

[Cameron]

31 minutes ago, Gildahl said:

Same question here.  I purchased V4 for VR and was never able to put to any serious use.  Between the rotating clouds and (most significantly) the eye convergence in foggy conditions I ultimately put it on the shelf to wait for a fix.  The effect was extremely obvious, even just sitting on the runway in foggy conditions. For me, I was told that the developers didn't have an Index to test this in.  So I'd love if a user who had been experiencing these two issues in VR before could report on what things look like now.

[Edit] Just read the post here, which probably answers my question.  So will be holding off for now as this was precisely one of my most major issues...

 

That doesn't appear to be the same issues. He seems happy about the clouds. His issue is the shadows (which are optional).

[Gildahl]

Ok, I'll keep my eyes out for further reports with people expressly indicating that they have shadows turned off.

[benjaprud]

Well... As a VR user, I can assure you that the rotating clouds/crosseyed effect problem is thankfully fixed, which is all it would take to make me a happy panda but !...

Volumetric clouds in SMP v5 do come with their own set of issues at the moment, which are to my eyes more disturbing than the "rotating clouds" effect I've gotten used to with v4 over time. See here and there if you want to get a rough idea and make an informed decision. Also the "clouds to the horizon" is not really a thing in VR yet. I have to keep my distance slider to the default 10000 I used before to keep similar performance, your mileage may vary depending on your GPU (I have a 1080Ti). To be fair, setting the slider to max does not impact performance nearly as much as it did before, but it does so enough that it's not suitable for me yet. Shadows do look awkward in VR. They "move" with head movements, which feels kinda similar to the rotating effect of v4 clouds so you'd want to have these disabled as well, as recommended by SMP's manual itself.

Currently you'd be trading rolling clouds for the occasional sea of milk or sharp-edged mountaintop showing through, until things hopefully get improved in the future, although it was hinted that a fix may come with a performance hit. Clouds do look and perform fine when they don't break though. They don't look great, but they're believable enough and have defined shapes and edges which is all I'd wish for. For now I'll test it a bit more to see what can be done to tune the performance and allow for longer viewing distances and maybe help the team with testing and feedback, but I may be rolling back to billboard clouds soon.

Edited January 24, 2021 by benjaprud

[Gildahl]

I actually had two issues in v4 that seemed separate from one another.  One was the rotating clouds issue, which was clear at altitudes when flying near or in cumulous clouds.  I'm glad to see that that appears to be fixed.  The second issue, however, and possibly the greater of the two for me was related to low visibility conditions on or near the ground in which there was fog.  The problem was that within the fog there would be banding effects, like posterization, associated with distance, but which led to serious eye convergence issues.  For example (and I recall posting some screenshots of this), one might see a building or part of the scenery more heavily obscured by fog in the left eye than the right eye, and when rotating the head the line that separated these two levels of obscuration would shift (though not appearing rotational).  These convergence issues were very pronounced and made fog effects unusable in my Index (I think there was a recommended mitigation that I tried, but it only helped slightly).  So while happy that the rotating clouds issue was addressed, I'm particularly interested in hearing from anyone who had the low visibility/fog banding issue in v4, commenting on whether that is fixed in v5. 

[benjaprud]

I've never experienced issues with fog myself, which may or may not be because I'm using ASXP alonside RWC/SMP or have different graphics settings. I'd be happy to try to reproduce the issue for you and see if v5 makes a difference there, but I'm not entirely sure of what I should look for. The only times I felt crosseyed is when I'm flying close to a cloud's edge and it rotates so close from my position that it will enter the cockpit, and obstruct each of my eye's viewpoint in a different manner.

Edited January 24, 2021 by benjaprud

[Gildahl]

On my Index, it is immediately obvious by just turning on low visibility and sitting on the runway.  Stock fog, on the other hand is perfectly fine. 

[Daikan]

9 hours ago, Gildahl said:

The second issue (...) was related to low visibility conditions on or near the ground in which there was fog.  The problem was that within the fog there would be banding effects, like posterization, associated with distance, but which led to serious eye convergence issues.  For example (and I recall posting some screenshots of this), one might see a building or part of the scenery more heavily obscured by fog in the left eye than the right eye, and when rotating the head the line that separated these two levels of obscuration would shift (though not appearing rotational).  These convergence issues were very pronounced and made fog effects unusable in my Index (I think there was a recommended mitigation that I tried, but it only helped slightly).  So while happy that the rotating clouds issue was addressed, I'm particularly interested in hearing from anyone who had the low visibility/fog banding issue in v4, commenting on whether that is fixed in v5. 

I don't have v5 (yet) but using my Oculus CV1 and SMP v4 I cannot reproduce any of the issues with ground fog you are describing.

But maybe you are using the word "fog" wrongly. FYI: "fog" is what X-Plane only applies BELOW the bottom cloud layer and is controlled by X-Plane's "visibility" setting i.e. without at last one cloud layer there will be no fog at all. Additionally, X-Plane uses a "haze" model which is still somewhat influenced by the "visibility" value but has no significant impact on ground visibility under ~10km (e.g. if you set visibility to the absolute minimum possible and no cloud layers you will still be able to see the ground from miles away). AFAIK the way "fog" and "haze" are represented is not changed by SMP at all (neither in v4 nor in v5).

However, I can somewhat relate to the described issues in v4 when inside or above the lowest cloud layer, so IMO what you are describing is actually due to differences in representation of the cloud billboards for the left/right eye. I assume these artifacts will be gone when using volumetric clouds in v5 - although seemingly at the cost of a whole new set of issues (and the reason why I'm still holding off and watching this topic with interest...).

Edited January 24, 2021 by Daikan

[benjaprud]

The clouds looks have started to grow on me, they're really cool. As far as VR is concerned there's probably some room for a visual downgrade. I'll be testing the 5.0.4 fix shortly which is supposed to resolve the terrain blending issues. However during my last flight with v5.0.1 I had more concerning issues in regards to VR performance in some overcast situations. My last flight started at 90 fps (clear skies over water and/or easy scenery), then was 45 most of the way (various cloud coverage above water along the way) but ended at 16 fps (low overcast at night with two broken layers underneath, easy scenery).

My setup is an Oculus CV1 with a 1080ti. I usually keep supersampling at 1.3 and can lock 45 fps most of the time with my current settings (HDR, textures max, no AA, anisotropic 8x, world objects high, reflections low). I'll try lowering some of that but I'm not sure if that'll be workable at the moment. I may be able to approach 30 fps by ditching supersampling and going from HDR to medium, which is about how low I'd be willing to compromise the rest.

Pic 1 looking down (GPU time is 15 ms without the capture software), pic 2 looking up (GPU time is 60 ms without capture software), two successive layers of overcast cumulus and low visibility, moderately dense overlay and mesh. This is with the latest v5.0.4.

Edited January 25, 2021 by benjaprud

[Gildahl]

2 hours ago, Daikan said:

I don't have v5 (yet) but using my Oculus CV1 and SMP v4 I cannot reproduce any of the issues with ground fog you are describing.

But maybe you are using the word "fog" wrongly. FYI: "fog" is what X-Plane only applies BELOW the bottom cloud layer and is controlled by X-Plane's "visibility" setting i.e. without at last one cloud layer there will be no fog at all. Additionally, X-Plane uses a "haze" model which is still somewhat influenced by the "visibility" value but has no significant impact on ground visibility under ~10km (e.g. if you set visibility to the absolute minimum possible and no cloud layers you will still be able to see the ground from miles away). AFAIK the way "fog" and "haze" are represented is not changed by SMP at all (neither in v4 nor in v5).

However, I can somewhat relate to the described issues in v4 when inside or above the lowest cloud layer, so IMO what you are describing is actually due to differences in representation of the cloud billboards for the left/right eye. I assume these artifacts will be gone when using volumetric clouds in v5 - although seemingly at the cost of a whole new set of issues (and the reason why I'm still holding off and watching this topic with interest...).

In the original threads on the issue, I think we concluded (or suspected) that the issue was most noticeable in higher res headsets.

[Jonnti46]

I'm afraid to report that actually the original problem is not resolved. If you use normal textures (Cumulus FAST), then you still will get double projection (although less bad) but it is certainly there. If you use the BETA Volumetric then you get no double projection issues, *but* the frame rate drops to 5 to 7fps in VR, as apposed to 90fps without VR. And as apposed to 60fps if using Cumulus FAST. So, the new beta will resolve projection issue, but frames drop dramatically even with draw area as low as possible, shadows off, Crepusular Rays Off, Cloud Terrain blending 0. 

There is something wrong in VR if you use the BETA Volumetric, (it is unusable) if you use the Cumulus FAST then same issue is basically there reported in June 2020.. never fixed. I new this was gamble paying for this upgrade - I figured I was going to be disappointed as I kind of new this would not have been fixed by the vague reply I got re the issue. As I waited *so* long - and then paid for a update that might have fixed it....  I think I'm fairly justified to be unhappy? 

[Cameron]

@Jonnti46

I think this comes down to your own misunderstanding. No one here thought to talk about any cloud settings other than volumetric because they are what's new to speak of. All other options are simply SkyMaxx 4 billboards, and you shouldn't expect a change.

Performance will vary by your GPU and rendering settings.

The vague response you received was because I truly didn't know. Actual users ended up answering you in detail above.

[Jonnti46]

But you answered when I checked with you on the other thread - on the announcement thread. Surely you must know the issue - it is a long running thread, you've responed many times to it, not just to me but to others with the same issue. If there was a doubt, you could have outlined this - i,e, said don't pay for upgrade as issue was actually never fixed, but you said it was (not in this thread) but the other one. I'm finding it quite hard to accept, I bought something that whilst I understand was not fully VR supported, but was to some degree (v3 and then v4) then we had various issues trying to sort it out, you said you'd go back to xplane developers - then there was a "silence" for several months (I think from Sept last year) then next thing I hear is new version, I question before upgrade if it fixes the original problem, you did not say only with Volumatic new Beta clouds.... you actually were really vague. Then we have a kind of fix, but a frame rate issue. Never had this with V4.... When you say you truly did not know, why did you not beta test it? I have x3 different VR headsets, low end, to high end - could have done this for you *and* still paid of course for update... I do not understand this concept.... now we have situation where beta is unusable and original clouds no fix afterall, so I paid again for something I can't use? Can you understand why I might be upset at this stage .... after months of patience... 

[Cameron]

I'm not going to argue with you.

All I said was essentially, "we think so but don't know because we were never able to reproduce it." 

[Jonnti46]

But you've been saying that since actually May 2020 - "we can't reproduce it" yet it is there. I've offered to help but you don't want it, you need users to trouble shoot it who can reproduce it, last I heard was you were discussing with Xplane developers - I'm not sure what happened as I never heard any follow up. I understand VR users are not high up on the agenda... but the issue is not just for me. And it certainly sounds like there won't ever be a fix for this. Is it possible I can have the upgrade $19.99 refunded? I mean I can't use it, I flew with it once just now as an expensive test, I've been waiting for the fix in the v4 threads for months. I just won't use v4 or 5 and will continue to use Xplane default clouds. The whole thing is very disappointing as I had some faith it would eventually be fixed. The new beta clouds / I don't get the point aside from the poor VR performance (I've got all settings right back) and I'm running a *really* powerful system. 

[Cameron]

You may find our refund policy here: https://www.x-aviation.com/catalog/refund_policy.php

I refuse to engage in a "he said, she said" argument with you. Therefore, I'm not responding to the rest (trust me, it's better that way).

[benjaprud]

@Cameron You may want to be aware that SMP5 will be looked upon by a big chunk of the VR simmers crowd. The main reason why I use SMP/RWC is that currently it's the only VR compatible solution that allows us to get rid of instant changes in weather, and I bet that's true for a lot of us. The improved graphics while welcome is not necessarily what I'm looking for. The last bit that was an issue with v4 was the rotating effect with the billboard clouds, which can feel nauseating to some, and highly distracting to all of us.

V5 is viewed as a potential solution to that last issue, meaning we may finally fly low VFR around the weather without any sort of distracting or immersion breaking behavior. While we need more experience and testing with v5 to see if it is the ideal weather solution for VR users, my experience with it so far is that the performance is hardly workable. It'll still take me lots more tuning attempts to see if I can make it work for me though.

In my opinion it highlights the fact that the expectations from flatscreen and VR users may diverge quite a bit, and there's no "one size fits all" answer to it. While the former will value a high fidelity graphics depiction, the latter strives for believable weather that doesn't hurt performance too much.

I'd understand that catering for the VR flightsimmers crowd may hardly be valuable as it represents a niche inside a niche. From where I stand, the SMP team is the closest to crack the issues we're having with weather in X-Plane, thus why we're looking at it with hope and passion.

[Cameron]

@benjaprud,

I can absolutely get on board with your constructive criticism, and I'm even happy to hear it. Being honest and providing the feedback is key and totally acceptable, so please do outline your issues as they come. We listen best we can!

What I can't get on board with is the desire to argue, such as shown above by someone else. It's just not going to happen with me.

[Gildahl]

Thanks folks.  I have my answer.

[benjaprud]

You guys may want to remove this for now, it's a bit misleading.

[Cameron]

2 minutes ago, benjaprud said:

You guys may want to remove this for now, it's a bit misleading.

Sure! Done.

Thanks for the feedback.

[benjaprud]

Thanks, I've toned down my settings a bit to some point where overall performance is less of an issue but still very much on the edge. It seems like from time to time I'll hit an edge case scenario that makes the performance go to trash, yet I can't point my finger to what's happening exactly. I've also found a way to capture my flights with the frametime info overlayed so hopefully I can catch some of these and send them to you for more constructive feedback.

Edited January 26, 2021 by benjaprud

[Cameron]

1 minute ago, benjaprud said:

Thanks, I've toned down my settings a bit to some point where performance is less of an issue but still very much on the edge. It seems like from time to time I'll hit an edge case scenario that makes the performance go to trash, yet I can't point my finger to what's happening exactly. I've also found a way to capture my flights with the frametime info overlayed so hopefully I can catch some of these and send them to you for more constructive feedback.

Any and all is appreciated and helpful!

[sundog]

VR performance issues are probably driven by the same things causing performance issues on people using 4K monitors that we're seeing. Just too many pixels to render; at some point something pushes things over the edge. I'm going to spend some time today trying to understand what's going on under the hood better.