I think we should provide a precise scenario for them to look at, with all variables (weather, lighting etc) the same, as then they should definitely see it (and if they don’t, I am keen to understand the config of their testing system!)
One that we have discussed before is the Rio A320 landing challenge. For me the shimmering is immediately apparent on the buildings to the left immediately when the challenge starts. This is not curable for me with any settings changes. I have Quest 2 but others have the same with different headsets.
I would say that the shimmer is actually much worse on my system. If someone provides me with simple instructions on how to record a video of the VR source, I can try and post it.
EDIT: Looks like it’s pretty simple directly via WMR, I will give it a try.
EDIT2: Tried it with the regular WMR capture function and the shimmer simply isn’t visible in the video. Surely this means that the problem is occuring to the signal somewhere along the way of it getting processed for, and put into, the HMD?
Good idea. I’ve never done this challenge but will take a look at it when I have time. Bascially, I get really bad shimmer in basically any autogen building area that is not in my immediate vicinity. It looks soo ugly and yes, it’s immersion killer #1.
It’s not just that the distant buildings are shimmering. Even the cockpit, text boxes, the wing edges etc. It is evident that anti-aliasing must not be working properly as every straight edge is jagged when tilted.
Also, at the exact same time, the issue appeared in menus and overlay windows, so this must be somehow connected. Asobo have acknowledged this issue and say it will be fixed, so we can only hope that this also solves the problem in general.
I honestly think it may only be apparent with certain hardware / driver configures if some people are really not seeing it.
There should be no need to set a specific environment for the simmering / non-existent AA to be seen. I see it on the runway as soon as my flight starts, and in VR the second I put on my VR headset. Runway markings, wing edges, cockpit edges, lines on the PFD and ND etc. It’s very prominent.
According to the roadmap, only the antialiasing in the menu will be repaired. The problem with the shimmer in the game is supposedly not reproducible for Asobo.
I can’t explain why Asobo doesn’t see this shimmer problem in game, especially in VR…
Here’s where I see new shimmering/aliasing since SU5:
-Overlay windows
-Buildings, particularly in bright sun (As described in this thread, visible in Rio landing challenge and similar to video shown by Captain Lucky above)
-Thin and high contrast text/lines on glass cockpits such as the A320 displays.
Since release I have also had constant shimmering on thin objects such as radio masts, fences etc.
10850k, 3090, Quest 2 via both Steam VR and Oculus VR. I fly at native resolution and 100 render scale, but like others I have experimented with many different settings, drivers etc but have been unable to resolve this.
@CptLucky8 : What recording method did you use to capture your shimmer video? I’d like to have a go too and tried with the standard WMR method, which produced a very choppy video where the shimmering was not really apparent.
I see it in almost all the time as well. But in certain conditions it is not as prominent. I just tried the new Dusseldorf Windy landing challenge and it is super prominent on the houses there as well.
This is a very representative video of the issue, and to set the record straight, there has been no change between SU4 and SU5 on the system used to recording it (drivers, hardware, etc…), or even if there has been one that I don’t remember, it is so widespread an issue that it makes it unlikely this is just a system related problem only (but it could be).
I’ve already posted all the details in my posts:
Of course it is also a possibility and a reason why this issue won’t be addressed as-is, but in this case it would have been clear for everyone if clearly announced. However, it is possible this can’t be announced as being solved with DX12 only because DX12 on PC is not fully implemented yet and they can’t therefore commit on any future uncertainty, which is expected and normal.
I’m the founder of Reality XP, we’ve been contributing to building this industry which was nascent when we started 20 years ago, we’ve kept raising the bar with a proven track record of renown, awarded products and innovative R&D ever since, on 3 different major simulation platforms. Yet MSFT stopped talking with us dead cold since March 2020. Being successful developer or experienced enginer in the field of video games, flight simulation and professional simulation doesn’t seem to be any valuable argument to help making this franchise any better.
I agree with both of these statements, and I will add that the issue may materialize differently for different hardware configurations. In my case, I see some shimmering, mainly in the distance, but my overall image both outside and inside the cockpit is blurrier than pre-SU5. But as someone above mentioned, whether or not this is really even due to an AA issue or something else that has changed is an unknown. The sim is a black box, and I can just go off of my perceived difference in image quality.
I am not quite sure what to offer up for Asobo to repro this. I can’t really offer visual proof, since I don’t have any screenshots from pre-SU5.
@Jummivana : I would be glad to detail my hardware specs, GPU & SteamVR configurations and MSFS configuration if this is something the devs would like to see, but I suppose in the end, whether or not they see anything wrong with the image is somewhat subjective.
Yet Lionel’s interview hinted they made some changes to the TAAU (accumulation buffer related IIRC), and when I look at the videos (and it shows better on the longer ones I have but couldn’t post due to size) it is clearly showing a repeating jittering pattern. Either it is jittering done usually for example to reduce posterization visually, or it is jittering due for example to cutting the precision of some maths in the shader code (loosing the less significant digits), or even an overflowing computation of some sort (like with base2 integer maths). But this is clearly repeating in a cycle.
From the look of it it seems to affect only certain materials and only at certain distances. You can see in the video some larger nearby buildings are not affected. This all speaks LOD related problem here, probably the BShark AI buildings at this LOD level using a wrong material, causing clipping, and overflowing. Just speculating of course, and in any case, what can I know.
Just a question about video length - why not post a longer video on onedrive, youtube or somewhere to better show the full extent of how ugly the shimmer can be?
I tried making a recording of the screen both digitally and with my phone filming the screen, but neither provided satisfactory results =(
Because I believe I’ve given enough technical details about this issue for any core rendering engine engineer to find out what this could be in their implementation, or otherwise to call me…
I too am exhausted with this, I just want to get this fixed and move on. Is it too much to ask for?
I believe I’ve also given enough technical details, it’s the devs that are playing games now by saying they can’t reproduce it and make a blind eye to the fact it’s really there.
If Asobo and Microsoft can’t fix this, I don’t have much hope for this sim to prosper long term.
I’ve just posted this in another topic but it is in fact in my opinion much more relevant in this one:
What I really find peculiar, and every one should find this peculiar, even those not seeing any problem, is why it is all of a sudden since SU5 that there are a lot of people complaining about this… Is this our tolerance threshold to the FS2020 flaws which all of a sudden got lower because of any SU5 let down, or is there any truth in the technical problems people are collectively reporting… food for thought.
very good description, i also have seen this! For example in Tokyo with the autogen Buildings in the distance (better to see than with the Photogrammetry).