Visibility: 1920x1080 vs 2560x1440

nighthawk2174 · November 19, 2019

This post is the result of me looking into how does resolution and the various AA settings inside of DCS impact visibility. Since then i've been paying a lot more attention to aircraft and spotting irl and how it compares to in games such as DCS. One particular observation will form the basis of this post:

I had a 4 ship of A10's pass right (and I mean RIGHT) over my house and I did a little test. I took about 40-50 sec for them to go to a dot where I couldn't even tell what they were. 80-90 sec to the point where if I hadn't known where to look I would not have seen them and around 100-110 sec to completely disappear. Let's just say it was 350mph for example. That's .1 miles per second so for a rear aspect A10, dot at around 5 miles, need to be cued somewhere around 8 miles and over 10 is when they disappeared. Now, 350mph is a bit slow and after talking to habu a speed of around 390-400mph would be more likely.

2560x1440, 24” monitor, MSAA off, SSAA off, NVIDIA sharpening and AA off

~38-36nmi - first visual in game with max zoom; one pixel only which is flickering in and out of existence rapidly.

https://i.imgur.com/1SKYYAJ.png

~11.2nmi - approximate maximum possible detection range irl in DCS target is invisible to 90° fov and still incredibly hard to see zoomed all the way in.

~11nmi - In DCS first visual with 90° FOV single pixel only that is flickering in and out of existence.

~8nmi - I would say at this distance he is about as visible as the real A10 about 10sec before it disappeared IRL although the A10 irl I would say had more contrast with the background and more of a nebulous dot. Low opacity on the outside and increasing towards the center.

5.5-4.5nmi - I would say at this distance the A10 in game is at the range i’d call max possible range for ‘easy’ tracking. Although the IRL A10 I’d say was a bit easier largely due to it having more contrast with the background and being a bit darker + bigger. In game the A10 is still only a single line of pixels with most of the pixels at the edges of the line flickering in and out of existence. Odd effect to note is about once or twice every quarter second or so the target will flash to twice its normal size and will become significantly darker than normal. If this effect was not present the max tracking range would be significantly less than here as it's the “flashing effect” that’s making it easy to track at this distance.

4.5-4nmi - Flashing effect becomes inverted making it harder to see than before due to this. By inverted I mean: The flashing effect continues but it has reversed polarity jumping between normal sized and smaller with the smaller size also becoming far less opaque. Also the rate it is occurring at appears to be slower than the former flashing effect.

- image captured half of the normal size seen here

https://i.imgur.com/9SqDz40.png

2.5nmi - Inverted flashing effect doubles (if not more) the rate at which it is occurring.

- half-sized target seen here, it's not much bigger than at 8+nmi.

1.0-0.8nmi - there is no doubt that the aircraft at this range is just about as easy to track as it was at 4.0-5.0 nmi. The inverted flashing effect continues jumping between normal-sized and smaller with the smaller size also becoming less opaque it is my opinion that it may actually be easier to see the target at 4.5-5.5 nmi than at this distance due to this inverted flashing effect. Note picture is of normal size not the flash effect in progress.

0.3-0.5nmi - able to tell its an A10

https://i.imgur.com/HmuVIDj.png

2560x1440, 24” monitor, MSAA 4x, SSAA off, NVIDIA sharpening and AA off

~32nmi - first visual in-game with max zoom; one or two pixels only which are flickering in and out of existence rapidly.

https://i.imgur.com/PIsoRot.png

~11.5nmi - In DCS first visual with 90° FOV single-pixel only that is flickering in and out of existence.

6-6.5nmi - I would say at this distance the A10 in-game is at the range i’d call max possible range for ‘easy’ tracking.

0.7nmi - able to tell its an A10

https://i.imgur.com/A7t84eY.png

2560x1440, 24” monitor, MSAA off, SSAA 2x, NVIDIA sharpening and AA off

~22.1nmi - first visual in-game with max zoom; one or two pixels only which are flickering in and out of existence rapidly.

https://i.imgur.com/zJGKu3A.png

~11.2nmi - approximate maximum possible detection range irl in DCS target is invisible to 90° fov and still incredibly hard to see zoomed all the way in.

~7nmi - low-end max uncued spotting range aka max range where spotting the dot and tracking it was easy. In DCS first visual with 90° FOV, singular pixel only flickering in and out of existence. Would say irl bit easier to see at max range than this largely due to color differences; in-game the aircraft is too blue. I would say at this distance he is about as visible as the real A10 moments before it disappeared IRL although the A10 irl I would say had more contrast with the background and more of a nebulous dot. Low opacity on the outside and increasing towards the center.

~2.5-2.8nmi - I would say at this distance the A10 in-game is at the range i’d call max possible range for ‘easy’ tracking. Although the IRL A10 i’d say was a bit easier largely due to it having more contrast with the background and being a bit darker + bigger. In game the A10 is still only a single line of pixels with most of the pixels at the edges of the line flickering in and out of existence.

~1.7nmi - About as visible as the A10 was even at 5-7miles irl. Very hard to see due to color and lack of contrast. IRL at 1.7 nmi irl I was clearly able to tell it was an A10 could see double engines and tell it was a dual tail aircraft not so in-game.

~0.4-0.5nmi - about the range in game where I can tell for sure its an A10 About a ⅓ if not less than the range I could irl

https://i.imgur.com/VrCy9wG.png

1920x1080, 24” monitor, MSAA off, SSAA off, NVIDIA sharpening and AA off

37nmi - first visual in game with max zoom; one or two pixels only which are flickering in and out of existence rapidly.

https://i.imgur.com/ntgl8xx.png

11.5-11nmi - In DCS first visual with 90° FOV single-pixel only that is flickering in and out of existence.

6.5nmi - I’d call this range max easy tracking range.

0.3-0.4 - about the range in game where I can tell for sure its an A10 about a ¼ if not less than the range I could irl.

https://i.imgur.com/qkUZplr.png

1920x1080, 24” monitor, MSAA off, SSAA 2x, NVIDIA sharpening and AA off

~25.0nmi - first visual in-game with max zoom; one or two pixels only which are flickering in and out of existence rapidly.

https://i.imgur.com/b6Ozegl.png

11.2nmi - low end max spot range irl in DCS target is invisible to 90° fov and still incredibly hard to see zoomed all the way in.

6.5nmi- In DCS first visual with 90° FOV single-pixel only that is flickering in and out of existence. Note much like at 1440 target is flashing to twice its normal size and will become significantly darker than normal. If this effect was not present the max detection range would absolutely be less than it is here as it's the “flashing effect” that’s making it easy to see the target at this distance.

4.5nmi - Target flashing effect reversed itself much like at 1440. But unlike 1440 this happened before I’d say the target is at a range I’d call max easy tracking range. Target is now almost entirely invisible. Now it is much harder to see than at 6.5nmi.

2.5-2.0nmi - Target flashing has mostly subsided and normal target tracking has returned. I’d call this range max easy tracking range.

~0.4-0.5nmi - about the range in-game where I can tell for sure its an A10 About a ⅓ if not less than the range I could irl.

https://i.imgur.com/yicr3Cs.png

1920x1080, 24” monitor, MSAA 4x, SSAA off, NVIDIA sharpening and AA off

~38.0-35.0nmi - first visual in-game with max zoom; one or two pixels only which are flickering in and out of existence rapidly.

https://i.imgur.com/WoFsvMJ.png

~11.2nmi - approximate max spot range irl in DCS target is invisible to 90° fov and still incredibly hard to see zoomed all the way in.

~10.5-10nmi - In DCS first visual with 90° FOV single-pixel only that is flickering in and out of existence.

~7.0-6.5nmi- I would say at this distance the A10 in-game is at the range I’d call max possible range for ‘easy’ tracking.

~0.4-0.5nmi - about the range in-game where I can tell for sure its an A10 About a ⅓ if not less than the range I could IRL

https://i.imgur.com/GXNtTOx.png

Conclusions:

-MSAA 4x is the closest to what I saw vision wise and if we could selectively apply something

like this just to aircraft models and not the whole screen this would help out with vision a lot. Still, though, it needs work.

a)From 1nmi to around 3nmi the target is too small and not dark enough. This actually lines

up quite well with what Serfross talks about in his paper that:

which I fully agree with now and after witnessing the differences between rlf and DCS I

cannot stress enough how accurate this paper is.

b)Targets from very close to far ranges are too "blue" it seems that there is pixel

color bleed here making objects blend in FAR more than they actually do.

-1920x1080, in general, is more accurate to what spotting is like than 2560x1440

-3d objects load in too far out and a limit on this should be placed

-Lack of glint or color changes when at an angle to the sun hamper visibility. Glint allows the detection and tracking of aircraft much further out than aircraft that aren't glinting. In one particular case, I was able to see a Learjet at around 17 miles quite easily when I was getting the glint off of the aircraft. This glint appeared several times larger than it did at substantially closer ranges. And was easy to spot due to the extremely bright nature of the glint (bright enough to hurt my eyes and force me to squint). However, once the glint stopped the plane was completely invisible.

-Aircraft that are catching the sun appears more of a greyish white and against the blue sky are substantially easier to see.

-SSAA can cause an odd flashing effect that severally hampers visibility at close ranges (under

6nmi) but helps it greatly above that distance.

Edited December 2, 2019 by BIGNEWY
edited 1.15

draconus · November 20, 2019

Thx for doing the tests but...

To compare IRL visibility one has to set their FOV based on monitor size and distance from eyes otherwise it's disconnected from the RL angular size. 90 degrees or max zoom are just test numbers here.

Not knowing the exact speed of those A10's makes any results just a hearsay. And it's still only the one observation. You were also not sitting in the cockpit.

Uploading smaller than original images makes them useless.

a) enough with the model enlargement bs, please

b) Yes! Proper color and additional effects is the way to go for realistic visibility modeling. This is where DCS lacks most and can greatly improve in.

Edited November 20, 2019 by draconus

nessuno0505 · November 20, 2019

It would be interesting to compare your observations on monitors with spotting range in VR, using different pixel density and AA settings

nighthawk2174 · November 20, 2019

Thx for doing the tests but...
To compare IRL visibility one has to set their FOV based on monitor size and distance from eyes otherwise it's disconnected from the RL angular size. 90 degrees or max zoom are just test numbers here.
Not knowing the exact speed of those A10's makes any results just a hearsay. And it's still only the one observation. You were also not sitting in the cockpit.
Uploading smaller than original images makes them useless.

a) enough with the model enlargement bs, please

b) Yes! Proper color and additional effects is the way to go for realistic visibility modeling. This is where DCS lacks most and can greatly improve in.

On point A:

I disagree firmly that's its bs; It wasn't just this one observation i've been spending a couple hours over the last couple of days observing all kinds of aircraft and I am firmly in agreement with the results sefross lays out in his paper. His findings indicate that there needs to be scaling to make it possible to have realistic levels of aircraft discrimination (telling direction and basic shapes) and I have 0 doubts about this. Not only is it well laid out in his paper why scaling is necessary but after spending time sitting and comparing rlf to games... well games that use some kind of scaling are just so much more accurate.

Additionally, and I should have made this more clear, the tests with the monitors were largely just to see what happens in DCS with differences in resolution; how much does resolution affects spotting in game. Also, I chose 90* largely due to it being near a common FOV setting most of the people I fly with are near and that I felt it offers a good compromise in representing human vision. Where a 40-45* arc is going to be the area with the highest "resolution" (for lack of a better word) with angles beyond 90 being exceptionally low in quality compared to this initial zone.

-also of note i just realized my mistake with the images and i'll upload the full sized ones when I can.

It would be interesting to compare your observations on monitors with spotting range in VR, using different pixel density and AA settings

I personally don't have access to 4k or VR but i'm hopping to get someone else who does to help me out on this front!

Edited November 20, 2019 by nighthawk2174

Eldur · November 24, 2019

I like your research, but upping pics of a 2560 monitor shrinked down to half size doesn't add to it at all. I can't see anything on those pics unless the targets are 3nm or closer, which, however, approximately matches my own observations at 1920x1200...

nighthawk2174 · November 24, 2019

Pictures updated finally got around to it today.

Mars Exulte · November 27, 2019

Very indepth testing, however I would point out a few flaws.

#1 Unless you adjust monitor FoV so the image is approximately the same ''scale'' as it would be in real life (on 24'' monitor at normal sitting distance it would be fairly zoomed in) you're going to get skewed results as objects very much will be ''smaller than they really are''. Just something to control for. This is also a leading contributor to spotting difficulties as people widen FoV for SA and immersion, inherently sacrificing some of their long range spotting (however flawed or not it is is irrelevant as it's going to have a similar effect regardless of the conditions).

#2 Assuming you're using the same monitor for both tests and simply lowering the resolution, you should be aware this will also skew results as monitors don't perform well at non-native resolutions. I.E. a 1440 screen downscaled to 1080 will be noticeably inferior compared to a native 1080 of the same size. Just something else to control for. It will tend to make small objects fuzzy, or soften edges usually.

That said, you spent a looooot of time experimenting and there is some very interesting information here.

nighthawk2174 · November 28, 2019

Very indepth testing, however I would point out a few flaws.

#1 Unless you adjust monitor FoV so the image is approximately the same ''scale'' as it would be in real life (on 24'' monitor at normal sitting distance it would be fairly zoomed in) you're going to get skewed results as objects very much will be ''smaller than they really are''. Just something to control for. This is also a leading contributor to spotting difficulties as people widen FoV for SA and immersion, inherently sacrificing some of their long range spotting (however flawed or not it is is irrelevant as it's going to have a similar effect regardless of the conditions).

#2 Assuming you're using the same monitor for both tests and simply lowering the resolution, you should be aware this will also skew results as monitors don't perform well at non-native resolutions. I.E. a 1440 screen downscaled to 1080 will be noticeably inferior compared to a native 1080 of the same size. Just something else to control for. It will tend to make small objects fuzzy, or soften edges usually.

That said, you spent a looooot of time experimenting and there is some very interesting information here.

Thank you,

1) I used 90* as it was both a good compromise between human FOV and what you can see on a monitor and was close to what I've heard is in use by those I fly with. So I felt seeing how visibility here compares would be important. In the end though any vision system imo should factor in FOV and seek to maintain how visible a target would be irl no matter the FOV.

2) Different monitors, upon testing just a downscaled 1440 monitor I can see what you mean with the blury edges but imho its not extremely noticeable and in the end the results were roughly the same.

BuzzU · November 29, 2019

A-10 cruising at 400 mph?

nighthawk2174 · December 1, 2019

TAS; based off of what Habu (our resident A10 piolt) told me at around 7000ft (approx altitude of the jets) and ~55F (12.8C) - on the low end - you should see around 305-310kts on the hud. Which when converted to TAS and MPH is 400mph.

BIGNEWY · December 2, 2019

I have edited the first post rather than delete, please remember our 1.15 rule here

thank you

David OC · December 2, 2019

We know all this nighthawk2174 and it's been around the forums a few times.

Conclusions and Recommendations

Quote

"Certainly, such a method could also be used to “magnify” simulated

ground images or other targets that a pilot needs to observe or “target” from long distances.

Furthermore, this should reach beyond the realms of flight simulation to other types of simulation, whether it is a simulated medical surgery or a consumer video game.

When the resolution of a system cannot provide the necessary detail for an object of a given size, it appears that, up to a point, size can be enhanced to compensate for this, yet the realism of the simulation need not be negatively affected. There are certainly limits to how much size enhancement can occur and still appear realistic."

End Quote

This paper is talking about fixed training simulators and adjustments for them and that setup.

Now how does ED implement a solution here?

How many different setups and resolutions are we talking about here?

How does this work with competitive multiplayer and VR?

That's what you also have to do here. Take a step back and look the bigger picture and how it can be implement without being gamey. From 1080p to 4k 24 inch to 4K tv, projector and VR.

ED's right in the middle ground here now perhaps? If you go anyway more, some will throw up there arms because "It's way easier if you have a 65 inch 4K tv" etc. I don't have 2k to spend on a tv to be more competitive, example etc.

Edited December 2, 2019 by David OC

draconus · December 2, 2019

If you go anyway more, some will throw up there arms because "It's way easier if you have a 65 inch 4K tv" etc. I don't have 2k to spend on a tv to be more competitive, example etc.

You can't stop the future of the display hardware in the name of competitivness, which DCS is not at all about. We will finally reach the eye resolution limits and no one will blur the image just to stay on the same level with low end displays. Better hardware brings us closer to simulated reality as it should, just like having i5 instead of core2.

Tippis · December 2, 2019

This paper is talking about fixed training simulators and adjustments for them and that setup.

Now how does ED implement a solution here?

How many different setups and resolutions are we talking about here?

How does this work with competitive multiplayer and VR?

That's what you also have to do here. Take a step back and look the bigger picture and how it can be implement without being gamey. From 1080p to 4k 24 inch to 4K tv, projector and VR.

The methodology itself is pretty much platform-agnostic. There might be some need to dampen the scaling curve as (angular) resolutions approach the absolute limits of even far-above-ordinary eyesight, but this runs up against the wall of diminishing returns — why would such a display device be needed?

But as the thesis points out, higher resolutions does not necessarily erase the need for this methodology since what it compensates for is not just size in relation to the display's capabilities but also things like minute movements, contrast, glint etc. Things that would require not just new hardware (not just “HDR” but proper over-brights as well as a much more precise colour differentiation) but a whole new rendering pipeline.

Figuring out a baseline scale for different displays is pretty simple — it's a bit of trigonometry plus reading the hardware settings from the display and adjusting the curve to yield the same real-world size at a specific FoV. The main complication — and it's not even a universal one — is figuring out the distance between display and viewer. This could either be a player setting, an assumed “sensible” number based on the display size (or type, in the case of VR), or just fixed. If left to the player, they could certainly input some bogus number to get a bit more magnification, but that would also screw up their zoom: they would get an advantage in one area in exchange for disadvantage in another.

And that's really the more difficult hurdle for implementation. Arguably, for such a solution to work, zoom just has to go — it's the zoom that is the gamey feature, not the adaptive scaling. And that's when the pitchforks and torches would be brought out… :D

SharpeXB · December 2, 2019

You can't stop the future of the display hardware in the name of competitivness, which DCS is not at all about. We will finally reach the eye resolution limits and no one will blur the image just to stay on the same level with low end displays. Better hardware brings us closer to simulated reality as it should, just like having i5 instead of core2.

If you had to choose, It would honestly be better that the sim favored higher end hardware over lower. Would anyone want a game that actually encouraged lower resolutions when the trend in displays and HMDs is moving to higher? Think 2019 and towards the future, not trying to keep with solutions for the past. A system which encourages players turn down the resolution of their displays in order to be competitive is not a good solution.

SharpeXB · December 2, 2019

I had a 4 ship of A10's pass right (and I mean RIGHT) over my house and I did a little test.

So you felt the need to do this test because you think ED has never seen a real A-10 and you feel compelled to educate them? :unsure:

This actually lines
up quite well with what Serfross talks about in his paper that:
...
which I fully agree with now ...
cannot stress enough how accurate this paper is.

So this post is just another in a long series of requests for “smart scaling” which is just a feature from another game.

Is it too much to ask that people stop posting this silly paper on the forums and stop asking DCS to imitate other games?

This paper is not god’s holy truth about flight simulation graphics & displays. And I seriously doubt the method described was ever used in any real simulator because changing the size of the target aircraft would be a terrible idea for real training.

Some useful quotes from it:

“The views expressed in this article are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government.”

More likely the whole paper was done with the goal of implementing this concept into video games, it’s mentioned several times.

“It should also apply to other fields of simulation where fine target detail is needed but is not readily

achievable, ranging from medical simulations to video games.”

“Furthermore, this should reach beyond the realms of flight simulation to other

types of simulation, whether it is a simulated medical surgery or a consumer video game.”

“Furthermore, whether it is flying-squadron-level “low-fidelity” flight

trainers or a scaled-down portable version of a video game, there has always been the

market for lower cost, more portable versions of otherwise expensive simulations.”

I do with ED would state something firmly about “smart scaling” so people would stop asking repeatedly for it and move on.

Edited December 2, 2019 by SharpeXB

Tippis · December 2, 2019

More likely the whole paper was done with the goal of implementing this concept into video games, it’s mentioned several times

It is expressly done with the goal of improving professional simulations to provide a more realistic learning environment. So no, that is not more likely.

You don't have to project your own wishes or inject any incorrect guesswork when the author straight out tells you what the goal is. This also applies to your assumptions about the goal the OP.

Edited December 2, 2019 by Tippis

nighthawk2174 · December 2, 2019

So you felt the need to do this test because you think ED has never seen a real A-10 and you feel compelled to educate them? :unsure:
[/Quote]

Well compared to irl spotting in DCS is quite bad so that's why I made this and wanted to see for myself the differences between 1440 and 1080 and see how could it be improved when compared to irl. May I remind everyone (agree or disagree) of what 9L has said in the past "We are competing with real-life..." well in the realm of vision DCS ain't doing so hot.

So this post is just another in a long series of requests for “smart scaling” which is just a feature from another game.
Is it too much to ask that people stop posting this silly paper on the forums and stop asking DCS to imitate other games?

This paper is not god’s holy truth about flight simulation graphics & displays. And I seriously doubt the method described was ever used in any real simulator because changing the size of the target aircraft would be a terrible idea for real training.
[/Quote]

Interesting description for sure... but do you got anything to prove he's wrong? Because until you do I'm just going to assume everything you say about this paper is "...wilfully ignorant and uninformed off-topic trolling."

Some useful quotes from it:
“The views expressed in this article are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government.”

“It should also apply to other fields of simulation where fine target detail is needed but is not readily
achievable, ranging from medical simulations to video games.”

“Furthermore, this should reach beyond the realms of flight simulation to other
types of simulation, whether it is a simulated medical surgery or a consumer video game.”

“Furthermore, whether it is flying-squadron-level “low-fidelity” flight
trainers or a scaled-down portable version of a video game, there has always been the
market for lower cost, more portable versions of otherwise expensive simulations.”

More likely the whole paper was done with the goal of implementing this concept into video games, it’s mentioned several times.
[/Quote]

I don't quite get how you get the conclusion of "the goal of implementing this concept into video games" from these. Maybe you just didn't notice he used words such as and/or/to to describe how scaling can be applied to all levels and types of simulation...

SharpeXB · December 2, 2019

Interesting description for sure... but do you got anything to prove he's wrong?

They’ve seen this silly paper posted on here hundreds of times...

Maybe rule 1.10 should apply to this.

And it’s only “right” for a specific setting and hardware and has tons of problems if you tried to use in in a game like this. Mainly that it would just look ugly.

Edited December 2, 2019 by SharpeXB

nighthawk2174 · December 2, 2019

https://forums.eagle.ru/showpost.php?p=3992128&postcount=97 - except it's easy to modify it to account for resolution and FOV

and well you can try it out for yourself:

https://why485.itch.io/smart-scaling-demonstration

and it is far from ugly, the fact you say that tells me either 1 of 2 things

1) you've never actually tried the above program

2) your definition of ugly is, looks fine-great to normal people...

NineLine · December 3, 2019

There is a lot of improvements to be done before we have to resort to smart scaling, better draw distances on all visual effects, including shadows, better contrast, better lighting, etc. Before we start making modules unnaturally larger than they should be, which as I have said in the past could be used as a form of cheat or advantage, we need to exhaust all other options, I have said most of this stuff before nighthawk, so I would prefer if you want to dredge up old discussions on all your favourite topics, you please look for the answers already given, especially if you are going to quote me.

Tippis · December 3, 2019

And it’s only “right” for a specific setting and hardware and has tons of problems if you tried to use in in a game like this. Mainly that it would just look ugly.

As always, this is demonstrably false as has been proven to you on numerous occasions. You have already admitted that you don't know what it looks like and that you haven't actually read the paper where your imagined concerns are already discussed and dismissed.

Just to reiterate what “ugly” effects we're talking about:

You keep clinging to this notion that it has “tons of problems” but you have never been able to actually specify what those problems are supposed to be and backed it up with any kinds of demonstrations or facts.

So the question remains the same as in those other thread: why are you so adamantly against improved realism and a more balanced and equitable across different hardware?

Edited December 3, 2019 by Tippis

nighthawk2174 · December 3, 2019

There is a lot of improvements to be done before we have to resort to smart scaling, better draw distances on all visual effects, including shadows, better contrast, better lighting, etc. Before we start making modules unnaturally larger than they should be, which as I have said in the past could be used as a form of cheat or advantage,

Resort too? I *do* agree that better shadows, contrast, and lighting are needed (I even make note of this in my conclusion in particular glint) but the reason you make things larger than they are is that it is more accurate to how your eyes perceive it. As sefross points out " The solution chosen here involves artificially enlarging the size of the simulated target to make the necessary cues more visible. It was believed that the target could be enlarged enough to do this, yet not be so large as to appear unrealistic or give the pilot a false sense of how the target would look in real combat".

Additionally, I'd like to know how you think it could be used as a cheat or an advantage? Especially if in its implementation you take into account FOV and screen resolution/size. For an example of this just load up:

https://why485.itch.io/smart-scaling-demonstration

put it on the lowest res and use Snapat V1 (clamped) it basically clamps the scaling factor at really odd resolutions to prevent any weirdness. And for the rest use Snapat V2 or Chihirobelmo which both take into account FOV and screen resolution to maintain how visible a target should be no matter your resolution or screen size.

please look for the answers already given, especially if you are going to quote me.

The most you've ever really said is "ED is looking into it"... and 'its a solution on an old game' (even if the solution is newer than the game) and imo that isn't much of an answer just a deflection that's often given for things that are broken but can't/aren't being worked on at the moment (missiles for example). Besides to quote chirhirobelmo:

"Hi Nineline,

Labels make aircraft easy to spot in all situations, but Serfoss Magnification does not.

Plane size continues to decrease by distance. It's just not got small rapidly as without scaling. Therefore spotting possibility decreases by distance, and becomes invisible at a certain distance. Decreasing contrast against background should play for 3d model, but labels don't.

It's not coming from 20+ year old game but coming from 2003 study used 1600 * 1200 = 1,920,000 pixels resolution for there research. It's not that different to today's standard 1920 * 1080 = 2,273,600 pixels resolution.

Maybe it works to good for 1440p or 4k monitor, I understand it will not fit for multiplayer balance, but I'd like to have it as an option for single-player or just closed multiplayer server so that it helps 1080p monitor user.

Another thing Serfoss scaling helps is a smaller monitor user. 24inch monitor user might see things 1:2 scaled but 42-inch monitor might see things 1:1 scale. magnifying 3.0nm object 2 times helps 24-inch monitor user.

So Serfoss Scaling is not as arcadish as label but just helps low res / small monitor user. user can enable or disable option considering their setup.

Yes it will not balance open public multiplayer server but I wish ED consider singleplayer and closed multiplayer server.

regards,"

Edited December 3, 2019 by nighthawk2174

NineLine · December 3, 2019

I don't agree that better shadows, contrast, and lighting are needed

But they are. These improvements are needed to improve visibility because the moment the graphics don't look like the real thing is the moment visibility queues are not the same. The visual distances for say the dogfighting does not need scaling, they need better contrast such as you might see against a blue sky or against the ground, flashes or reflections, etc are all better ideas than resizing. Remember, we have real pilots on the team, Mr. Grey spends a lot of time peering at WWII birds in their air.

ED isn't looking into smart scaling, we tried something like it a while back, it wasn't great, of course, you read all this already, cuz I know you would make sure to look at everything that has been discussed already.

I run 2880 x 1600 windowed on a huge monitor, I have zero issues with spotting except when expected (and yes, there are times you should have trouble spotting because most things in combat are made to be hard to see). If I ran this resolution on a small monitor, I am sure I would have more issues. There is so much more that goes into it that simply making objects bigger.

And honestly, much like gun dispersion, this has all been talked about before, and like the other threads, people will start fighting with each other, then someone on the mod team will have to lock it.

nighthawk2174 · December 3, 2019

duplicate post whoops

Edited December 3, 2019 by nighthawk2174

Visibility: 1920x1080 vs 2560x1440

Recommended Posts

Link to comment

Share on other sites

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Popular Posts

Tippis

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Recently Browsing 0 members