in review Spotting dot bugs in VR

[falcon_120]

I agree it needs some tweaking, but its not as bad as it was before at least thats the way i see it.

Tune down a bit how far you see the squares and work a bit om the transition to LoD and you're good. Its never gonna be the same as RL.

Enviado desde mi ELE-L29 mediante Tapatalk

[YoYo]

Yes, its a little better than before but still too big and too mutch visible.

[TAIPAN_]

20 hours ago, YoYo said:

Yes, its a little better than before but still too big and too mutch visible.

Dots are very small and hard to see on Pimax Crystal, though I have very high resolution and quadviews.

I have heard if you use DLAA or DLSS it makes the dots smaller, could be worth a try.

For myself if they were any smaller they would be invisible, I guess everyone has a different setup/settings.

[YoYo]

This is example direct from my VR.

Full picture (one eve):

Part from full view:

Flying box.

[KoN]

On 5/11/2024 at 5:21 PM, YoYo said:

What's going on in this subject at all? The problem is not solved, you can still see a large square that breaks the immersion, then when the LOD of the model is loaded, the object becomes smaller. Unfortunately it looks very bad still. Any news about improvements?

We are still waiting for improvement, still waiting for MT improvement.

I've gone back to old style spotting, it's not ideal in VR , but neither is the new style. I tested on 4K monitor and the spotting looked good. 

But in VR it's way too big . 

[TAIPAN_]

15 hours ago, YoYo said:

This is example direct from my VR.

Full picture (one eve):

Part from full view:

Flying box.

It doesn't look too bad to me for the non zoomed view (first picture), and I wouldn't even know there was another plane there unless you zoomed/circled it.

But I don't doubt you because headsets always display things differently to a 2D monitor, and can be very specific to the individual lenses, DCS settings, resolution, antialiasing etc.

For that reason I still think it's a good idea for there to be a slider in the settings where you can (optionally) reduce the size, or the range of the dot. Being optional means it won't affect those who don't see large dots currently (eg pimax crystal with quadviews and 5000x4000 per eye with 200% AA)

 

[YoYo]

4 minutes ago, TAIPAN_ said:

It doesn't look too bad to me for the non zoomed view (first picture), and I wouldn't even know there was another plane there unless you zoomed/circled it.

But I don't doubt you because headsets always display things differently to a 2D monitor, and can be very specific to the individual lenses, DCS settings, resolution, antialiasing etc.

For that reason I still think it's a good idea for there to be a slider in the settings where you can (optionally) reduce the size, or the range of the dot. Being optional means it won't affect those who don't see large dots currently (eg pimax crystal with quadviews and 5000x4000 per eye with 200% AA)

 

BUT the main problem is that its NOT ZOOMED. This is what we see in VR.

The cutout is on a scale of 1:1 and this is what we see in VR. I agree, this first screenshot look good on 4K or even biggen monitor. The resolution of whole picture for me is 5536 x  2832 (with my FOV). Looks really nice on 2D, I agree, but the cutout is in scale 1:1, so you see something like this, quite big box from long distances.

I think the solution should looks like this:

1/ one system of "dot" for 2D picture (and according me on 2D is ok what we have now),

2/ second system for VR mode, but importantly, separate from 2D. If you are in VR mode, it will load automatically. This requires an in-game code change, but it is possible, am I right @Flappie ? This would allow the developer to set other parameters, so I think that's where we should start.

The current system spoils the immersion of the dogfight rules. Phantom incoming soon and I would like to go back to short-distance air fighting and BVR, but with the system we have it looks terrible, unfortunately. I don't know how people can fly wiith this in multiplayer, it's a kind of cheat, what we have now in VR vs 2D. Someone doesn't see me, but when I see a big box. 

This case has been going on for quite a long time and there is no end in sight  .

[YoYo]

18 minutes ago, TAIPAN_ said:

For that reason I still think it's a good idea for there to be a slider in the settings where you can (optionally) reduce the size, or the range of the dot. Being optional means it won't affect those who don't see large dots currently (eg pimax crystal with quadviews and 5000x4000 per eye with 200% AA)

I also thought about this, but I'm afraid that the slider is not a good option, because everyone can have something different set (remember about multiplayer), which may again cause some cheating by other players. By the looks of it, the current dot for VR should be 1/4 of what it is now. It will be better than what it was before, smaller but also noticeable. They should give this option to test for 4 - 6 players with different headsets (Quest, Reverb, Crystal 8K, 5K, etc.) and a compromise should be reached. I think 1/4 would suit everyone, that's my guess. It will never be perfect due to different resolutions in goggles and the possibility of upscaling and downscaling, but one compromise option should be enough for everyone.

[Tippis]

4 hours ago, YoYo said:

I think the solution should looks like this:

1/ one system of "dot" for 2D picture (and according me on 2D is ok what we have now),

2/ second system for VR mode, but importantly, separate from 2D. If you are in VR mode, it will load automatically. This requires an in-game code change, but it is possible, am I right @Flappie ? This would allow the developer to set other parameters, so I think that's where we should start.

It's pure speculation but by all accounts, this seems to be the biggest problem and snag at the moment. It feels like the game naively just looks at the rendering resolution — possibly just the display height — and scales the dot based on that without any awareness of or compensation for what kind of display it is targeting. So a VR display with a 2k vertical resolution for each eye gets fed the same dot size as a 4k display since, hey, it also has a 2160 pixel height, so it makes sense to let it be the same size. Or something.

…except that, of course, the 4k display will be viewed from a meter or so away, whereas the VR display distance is better measured in millimetres. Ultimately, there probably needs to be a method of differentiating the two, like you suggest, and to try to compensate for the up- and downscaling. The latter could conceivably be handled by just having a different scale factor for “near-eye displays”, the logic being that the up- or downscaling of the dot size is inherently counteracted by the down- and upscaling of the rendering resolution so it all comes out in the wash.

As for the slider idea, it's probably a reasonable one irrespective of the exploitation potentials even though that's a legit concern. You can largely do that anyway by adjusting the resolution and, more practically, by just leaning forward. There's no way to get around that so why not offer it as an option for the much larger audience where you want to dial in the size for your particular physical setup, and if some numpties want to cheat in MP, there's always the screenshot, kick and ban functions. Numpties will numpty and will undoubtedly figure out a way to tell the game to feed them the VR scale factor in pancake mode or some such so restrictions won't help much anyway. The benefits would probably outweigh the disadvantages. Some of it could be mitigated by having some fairly tight restrictions on how much you can adjust it. If it's just a ±1 pixel tweak, and you're mostly restricted by range as far as how early the dots show up, the worst excesses can probably be handled.

[YoYo]

We focused here mostly about air contacts, but worth to admint, that this issue concerns also ground units. Here is good example, about 40 kilometers from units, quite far. Default units from the map are not visible, but added units, even the same models, has very visible black box too:

Again on full resolution on 2D looks quite ok:

but on 3D, 1:1 looks like this (not zoomed picture):

Filling in goggles is the same.

Generally, in VR you don't need to use any radar or anything to locate targets from a long distance. Comparing it to 2D, people have their panties on their knees. 

Edited May 13 by YoYo

[SharpeXB]

5 hours ago, YoYo said:

I don't know how people can fly wiith this in multiplayer, it's a kind of cheat

Yes it is. Right now it’s possible to totally exploit this by selecting either the 2.8 or 2.9 version (Off just switches to the old 2.8 dot). What the game needs is a true Off setting that’s mission/server enforced. 

[Tippis]

12 minutes ago, SharpeXB said:

Yes it is. Right now it’s possible to totally exploit this by selecting either the 2.8 or 2.9 version (Off just switches to the old 2.8 dot). What the game needs is a true Off setting that’s mission/server enforced. 

That would reintroduce the cheat that this change has gotten rid of where targets can be seen at utterly ridiculous distances. That cheat is gone for good reason and should never be allowed back in.

[YoYo]

But the worst thing in current system is probably this. Target about 10 km from me (large box), target about 7.5 km from me (different LOD loaded) - the dot got much smaller. What's the point?

10km:

Target about 7km:

It should be the other way around if anything ...

[zildac]

But the worst thing in current system is probably this. Target about 10 km from me (large box), target about 7.5 km from me (different LOD loaded) - the dot got much smaller. What's the point?
10km:

Target about 7km:

It should be the other way around if anything ...

Agreed

[Tippis]

2 hours ago, YoYo said:

But the worst thing in current system is probably this. Target about 10 km from me (large box), target about 7.5 km from me (different LOD loaded) - the dot got much smaller. What's the point?

The point is that the second one isn't a dot and isn't subject to resolution normalisation, whereas the first one is and thus isn't subject to LOD — just range attenuation.

This is why there is no single solution to the spotting problem, and a multitude of methods are needed to cover the different range bands, each with its own normalisation and compensation schemes. But more than that, the point is that, bit by bit, we're moving in the right direction.

[TAIPAN_]

On 5/13/2024 at 7:36 PM, YoYo said:

I also thought about this, but I'm afraid that the slider is not a good option, because everyone can have something different set (remember about multiplayer), which may again cause some cheating by other players.

 

Without the slider, it would be yourself and older headsets that would be cheating. People WILL literally see something different (thinking about multiplayer as you said). ie currently the dots are very small and hard to notice for me, but for you they are very large.

Imagine we reduce the dot size (without a slider setting) to a size that makes you and lower res headsets happy, then it will be invisible for me and not future proof as everyone will eventually be on newer high resolution headsets. We WILL see something different (ie some people will see nothing).

In any case I didn't suggest a slider to increase the size of the dot, only to decrease. Just for people who do not like the experience like yourself, you would reduce it to be the same size as everyone else. For everyone else who finds it ok can leave it correct as is. To me the current dot system is perfect, and changing that for lower resolution headsets would ruin that.

  

On 5/13/2024 at 7:36 PM, YoYo said:

By the looks of it, the current dot for VR should be 1/4 of what it is now. It will be better than what it was before, smaller but also noticeable.

1/4 would be literally invisible (except for low res), ie you would have advantage in multiplayer which you mentioned one of your concerns.

Turning on DLSS/DLAA reduces the dot to near half, and I cannot see it at all after that. Without DLSS/DLAA it's already minimum for my headset & DFR/Quadviews.

  

On 5/13/2024 at 7:36 PM, YoYo said:

 It will never be perfect due to different resolutions in goggles and the possibility of upscaling and downscaling, but one compromise option should be enough for everyone.

To me it is perfect already, and the current system is already a compromise. Eventually the older headsets will phase out and everyone will be on newer headsets with high res in the long run. 

A user setting could be a solution for people who find it too big as a personal choice. There's no way it's a "cheat" since it would only reduce the size.

Edited May 16 by TAIPAN_

[TAIPAN_]

On 5/14/2024 at 12:08 AM, Tippis said:

It's pure speculation but by all accounts, this seems to be the biggest problem and snag at the moment. It feels like the game naively just looks at the rendering resolution — possibly just the display height — and scales the dot based on that without any awareness of or compensation for what kind of display it is targeting. So a VR display with a 2k vertical resolution for each eye gets fed the same dot size as a 4k display since, hey, it also has a 2160 pixel height, so it makes sense to let it be the same size. Or something.

…except that, of course, the 4k display will be viewed from a meter or so away, whereas the VR display distance is better measured in millimetres. Ultimately, there probably needs to be a method of differentiating the two, like you suggest, and to try to compensate for the up- and downscaling. The latter could conceivably be handled by just having a different scale factor for “near-eye displays”, the logic being that the up- or downscaling of the dot size is inherently counteracted by the down- and upscaling of the rendering resolution so it all comes out in the wash.

As for the slider idea, it's probably a reasonable one irrespective of the exploitation potentials even though that's a legit concern. You can largely do that anyway by adjusting the resolution and, more practically, by just leaning forward. There's no way to get around that so why not offer it as an option for the much larger audience where you want to dial in the size for your particular physical setup, and if some numpties want to cheat in MP, there's always the screenshot, kick and ban functions. Numpties will numpty and will undoubtedly figure out a way to tell the game to feed them the VR scale factor in pancake mode or some such so restrictions won't help much anyway. The benefits would probably outweigh the disadvantages. Some of it could be mitigated by having some fairly tight restrictions on how much you can adjust it. If it's just a ±1 pixel tweak, and you're mostly restricted by range as far as how early the dots show up, the worst excesses can probably be handled.

Agreed, it IS a problem it takes the raw resolution.

When using DFR, eye tracking quadviews, the resolution is massively scaled up in the focus area ie 200%.

So we have a view that's got a dot made for 100% resolution but we are actually at 200% which is 4x the pixels.

[YoYo]

30 minutes ago, TAIPAN_ said:

Turning on DLSS/DLAA reduces the dot to near half, and I cannot see it at all after that. Without DLSS/DLAA it's already minimum for my headset & DFR/Quadviews.

What? Turning on DLSS/DLAA makes no sens, the picture for VR is blurry, no sharp, so don't expect to see anything from the long distance in any goggles. DLSS+DLAA is good but for 2D.

BUT The most important thing is this DLSS is not a benchmark here, check AA=off or MSAA. If it works well on these settings, then compare how well it performs with DLSS and DLAA - this is a compromise, not the other way around. Currently, it must be adjusted to the no anti-aliasing, and only then should it be taken into account how it compares to other settings. The current system is not even a compromise in this way.

Edited May 16 by YoYo

[draconus]

The whole idea of fake dots is a fail. It should be a job of a 3D graphic engine to either render the object with proper angular size on the screen using 1 or more pixels, or render 1 pixel fading into background until disappearing when reaching range far enough. How to tell if it's 1 pixel or more? Basic math taking resolution vs fov vs object angular size - gfx engine's job. This way purposely lowering the resolution would be disadvatageous because you'll worsen your spotting - the 4K would still show a faint pixel while FullHD will lose the sight entirely at the same fov & range. It can't correlate well with RL spotting since we also have different sizes of monitors and can change fov on the fly.

It is absolutely wrong to fill more pixels for the object that is farther than the one that is closer at the same aspect and light.

Current dot also fails to take into account important features from the model like skin colors, light reflections/glints. We can see white parachutes from far away depicted by 4 black dots (a box), same as the aircraft this pilot just ejected from.

Edited May 16 by draconus

[Raven (Elysian Angel)]

26 minutes ago, draconus said:

We can see white parachutes from far away depicted by 4 black dots (a box), same as the aircraft this pilot just ejected.

Yes that's the silly thing: at the moment we can see parachutes at 5nm+ easier than aircraft at 1-2nm 

[TAIPAN_]

3 hours ago, YoYo said:

What? Turning on DLSS/DLAA makes no sens, the picture for VR is blurry, no sharp, so don't expect to see anything from the long distance in any goggles. DLSS+DLAA is good but for 2D.

 

 

I agree, I don't use it personally. I don't even have a choice because I play cold war I wouldn't be able to see any enemy aircraft if I used it, but it's not needed for me because Quadviews gives enough performance (and a 4090 :)). 

I only mentioned it to highlight what happened when the pixels halved.

  

3 hours ago, YoYo said:

BUT The most important thing is this DLSS is not a benchmark here, check AA=off or MSAA. If it works well on these settings, then compare how well it performs with DLSS and DLAA - this is a compromise, not the other way around. Currently, it must be adjusted to the no anti-aliasing, and only then should it be taken into account how it compares to other settings. The current system is not even a compromise in this way.

 

Don't add too much importance to the DLSS comment, all it meant was that I already tested reducing the size of the dot and found that when the pixels were about halved it was mostly invisible (for crystal users on high settings).

I think for a lot of people the current implementation is working well, there have been some tweaks by the devs since the first implementation I have noticed some improvements. A user option to allow people like yourself to tailor it to suit you (reduction) would be acceptable as it would prevent damaging the good system that is currently in place.

But I don't think it's fair to impose a reduction in size across the board, stating multiplayer fairness when in fact it's the opposite. The last thing we want is to go back to the old days where people used to reduce their resolution to 1080p to get an advantage in PvP and have an expensive graphics card sitting there wasted.

  

1 hour ago, draconus said:

The whole idea of fake dots is a fail. It should be a job of a 3D graphic engine to either render the object with proper angular size on the screen using 1 or more pixels, or render 1 pixel fading into background until disappearing when reaching range far enough. How to tell if it's 1 pixel or more? Basic math taking resolution vs fov vs object angular size - gfx engine's job. This way purposely lowering the resolution would be disadvatageous because you'll worsen your spotting - the 4K would still show a faint pixel while FullHD will lose the sight entirely at the same fov & range. It can't correlate well with RL spotting since we also have different sizes of monitors and can change fov on the fly.

It is absolutely wrong to fill more pixels for the object that is farther than the one that is closer at the same aspect and light.

Current dot also fails to take into account important features from the model like skin colors, light reflections/glints. We can see white parachutes from far away depicted by 4 black dots (a box), same as the aircraft this pilot just ejected.

 

Agreed on this, all the while we rely on a dot is more time we end up arguing about how to size and transition it.

Scaling of the actual aircraft appears to be happening in the another sim (a WW2 sim) since they did the spotting fix (maybe 2 years ago). I see faint blurry small aircraft that gradually become clearer. All at high resolutions too, so there must be a way to make it work. And it's not the oversized smart scaling that you see in that other F16 sim.

Maybe DCS has a difficult engine to work with, I'm sure there are complexities we haven't imagined. But it would be nice if it was model scaling, though those at low resolutions would still find the scaling doesn't fit in 1-2 pixels so they would end up with something like a dot. Long term those resolutions would phase out.

Edited May 16 by TAIPAN_

[YoYo]

4 minutes ago, TAIPAN_ said:

The last thing we want is to go back to the old days where people used to reduce their resolution to 1080p to get an advantage in PvP and have an expensive graphics card sitting there wasted.

There will always be cheaters, if someone wants to spoil the fun just for some virtual point and to cure their complexes, well, there's nothing we can do about it . There will probably be no perfect solution here, it would be best if a fix was made and given to owners of different goggles to test before release.

[draconus]

11 minutes ago, TAIPAN_ said:

But it would be nice if it was model scaling, though those at low resolutions would still find the scaling doesn't fit in 1-2 pixels so they would end up with something like a dot.

DCS does the proper scaling already but at some range the model disappears and the "dot" is put in place instead. No matter the resolution there's always some range that the aircraft becomes 1 pixel or smaller. The point is to handle it well by gfx engine and depict it realistically - this should be priority, not some impossible PvP balance where everybody have different display size, resolution and fov.

[SharpeXB]

This has all been discussed to death with not much new to add. But again the idea of having artificial dots is a fail. For the following reasons:

- It’s not realistic to easily see distant aircraft. 
- Dots cannot portray target aspect so a nose-on MiG-21 looks the same as a top view F-14. 
- The system is exploitable by players online. 
- The Spotting Dots largely duplicate the Dot Labels which are already a feature. 

I get it that ED can’t solve this because they’re up against other video games which enhance the distant aircraft in order to make the game “fun” as opposed to realistic. But the very least that could be done here is to add a true Off option that’s mission/server enforced. The current Off setting doesn’t turn these off it just switches between v 2.8 to 2.9 making the setting more exploitable. 

[YoYo]

43 minutes ago, SharpeXB said:

This has all been discussed to death