Aircraft Visibility

[SkateZilla]

Yes, that's pretty much it. Only thing that will make that light eventually stop being visible is distortion due to air particles, or other bright objects (or light reflections) that overwhelm the smaller lights.

 

The NAV lights have exact same issue as with sunglare, they are just textures that reduce in size as they get further away, and withing 2-3Km they just "fade"out into surrounding terrain textures.

 

Likely lights/glares/lens flares etc would have to be turned into purely shader based and not geometry constricted using flat planes w/ textures.

 

I still think the Sub-Px Rendering is prolly the best option with shimmering/flickering glares when sunny.

Edited August 10, 2014 by SkateZilla

[NineLine]

Likely lights/glares/lens flares etc would have to be turned into purely shader based and not geometry constricted using flat planes w/ textures.

 

I still think the Sub-Px Rendering is prolly the best option with shimmering/flickering glares when sunny.

 

Hey, we just come up with the great ideas, its ED's job to figure that stuff out :D

[Wolf Rider]

here's something which will help with monitor calibration...

 

http://www.lagom.nl/lcd-test/

[theGozr]

Something that may help a bit while waiting .. try to setup your FOV to a "max" 95 degree .. if you are going too wide the lods will be to far like ..

[lunaticfringe]

This is a big deal. Color, resolution, and such all have soooo many factors... not everyone sees colors the same... so I could show you something that looks good to me, and be completely terrible for you... its a slippery slope.

 

If it was me, from my small time up front in an aircraft, along with my illustration and photographic background, I'd be suggesting the following:

 

1. Desaturation of DCS World's color palette- some have called it cartoonish, and, while I am predisposed to considering it more of a HDRI view of the world, the fact that the human eye can *see* that much wavelength doesn't mean that

 

a. it will ever discern all of that value at once, and

b. the monitor being challenged to display it is anywhere close to being up to the task.

 

These two things cause situations where different screens will fail to show objects that others will when there are large disparities across the available display gamut. It also puts the viewer into fewer situations where contrast will be obscene at range, too. As range increases, contrast should become more subtle. Another benefit is that by desaturating the available palatte, you're compressing the size of what is being displayed back into a more consistent color space (say for example- 85% of Apple RGB, which essentially any monitor made in the last 10 years which isn't a laptop screen made on the far end of that range can produce easily) in such a way that the screen response time can show it clean.

 

A lot of the issue is happening because of the wide palette. 1 v 1 over water, or over the plains from above 15k; the hard deck is far too dark. The amount of light hitting it is, by scale, entirely too small versus what should be a similarly heavily lit object (the target), which should become brighter against that dark. While more realistic lighting would cause the ocean or plain to become apparently lighter (rather than more vivid- ), the target would be even brighter than that, meaning that's how you pick him up. He's not going to be a chrome point of light, but he's going to stand out because their ultimate values will be separated in a more consistent fashion.

 

Contrast this to now- the visual background is dark, and the engine is doing everything in its power to keep that target vivid. This collapses the contrast and you can't see him, even with as little as two miles range with his plan view facing you.

 

And let me be frank- from what we've seen of EDGE, it doesn't fix this. Light falloff is too far out, maintaining its vivid nature. It's repeating this same problem. I'll hold until it's released to pass final judgement, but from what I've seen, this will be a continuing concern.

 

2. Calculated FoV lock- this can be calculated based on apparent window size, or better- system reported screen PPI. With that one piece of information, the engine can calculate apparent range to the eyeball from a user-defined distance to the screen. Now, this might not seem fair to folks who work on small monitors and zoom out in a huge way, but it will solve whatever potential complaints of "balance" that could be had. If folks are sitting their monitors to a point where they can pick out a single pixel, yet have comfortable view of the full screen, then you're making it a consistent experience across the board.

 

3. (the big one) Depth of Field locked to zoom, rather than what looks "pretty".

 

The biggest problem a pilot has with keeping sight is open airspace. The eye has difficulty in finding frames of reference with which to allow the muscles to dilate and generate his optical "zoom" for him. However, as the eyes zoom in and out, their Depth of Field changes. Eyeballs do have something akin to a hyperfocal distance, but you do lose huge portions of your peripheral, as well as everything under that range. And yet, depending on the range, the eye works less than like a hyperfocal, and more like a true zoom with normal relative Depth of Field.

 

And that right there is the key to detecting movement when zoomed- too much is in perfect focus. When the eye starts to see what appears to be seemingly immediate changes in contrast against a region it's looking at, it hunts to focus like your camera's AF; and nothing in the air causes apparent change in contrast to focus on than movement. Your brain starts saying, "hey, I'm seeing 45% against 50% gray, and that's maybe clouds and the horizon. But why's this blob of 80% in there moving left to right...?"

 

Bang. A split second later, if you're trained in the art, your eyeball is fixed on that movement.

 

So what I would suggest is that, rather than using the DoF capability of EDGE to make things look nice, use them to make the visual search challenge more realistic. If it could be defined as a zoom axis, everything under 15% is for in the cockpit (gauges, etc), and from there on out it should scale like 20/20 or 20/10 vision at range (since we're working with the baseline of folks needing aids having them on; also note- this is the *real* reason you need to lock FoV to do this properly), with appropriate shifts in what will be in focus at given ranges. And you would find that people should be detecting things with less amounts of zoom than expected, because things out on the horizon will be more out of focus, meaning what's closer will be in, or at least showing visible contrast.

 

But that's what I can come up with. Who knows if and what ED wants to do to solve the issue creatively, if at all.

[Wolf Rider]

One misconception, is that the distance from the user's eye to the screen (and all the angulars equated to there) has nothing to do with how any image on the monitor is affected, at whatever FoV that image is projected onto the screen at...

The closest photographic representation of how the human eye 'sees' is through tilt-shift lenses, and even then at that representation is still way off.

The human eye doesn't 'zoom', it just focuses.

[lunaticfringe]

The human eye doesn't 'zoom', it just focuses.

 

No, it doesn't zoom. Used here, it's a conceptual conceit to bring a closer parallel to how the eye works correlates to what is taking place in the engine. The engine only "zooms" rather than permitting realtime change in DoF (which is what the eye is doing). Thus, when I say that the eye is "zooming", it's drawing it's focus out to a farther distance, which is only comparable in the simulation as actually zooming along the visual z-axis.

 

One misconception, is that the distance from the user's eye to the screen (and all the angulars equated to there) has nothing to do with how any image on the monitor is affected, at whatever FoV that image is projected onto the screen at...

 

It doesn't have anything to do with what is presented on the monitor. It does, however, have everything to do with what is visible to the *player*, and that is based on said distance.

 

Example: A fighter-sized object (say, a MiG-29 smoking in MIL) is on the horizon. Based on 20/10 vision, say that this should be visible in a head-on aspect at 14 miles with the representation of one lit pixel.

 

If you permit the changing of FoV (which is what zooming actually does, projection of smaller fields of view in the same apparent visual area), you're changing the apparent size of said object. A pilot strapped into his seat can, at most, get his eyeballs closer to an object (disregarding the net range collapse by his velocity at the object) by, at most, 6-8 inches of slack in the harness and traverse of his neck. The FoV is changing by thousandths of a percent at that 14 mile range. What he can change is how far out his eyes are focused to pick up that microscopic object, which is more apparent based on its relative sharpness to the rest of what is in his visual field.

 

This is why locking the FoV, and tying it to the visual field of the display, is the "balancing" option. Everyone should loose the zoom function. Apparent FoV should only change with the addition of/resizing of displays to bring in more of the peripheral. In such a system, the best the player can do is lean in a bit- he's still having to hunt the same as a pilot based on applied DoF in the visual z-axis, rather than changing the projection.

 

What we have now is the reverse: bigger the display, the bigger you can get the target to eyeball.

[Galwran]

IL-2 Sturmovik did this the right way. First you saw faint gray dots in the distance. Then, a bit darker dots. I bit closer, and the dots became bigger. And then they became proper airplanes.

[9.JG27 DavidRed]

More resolution is better, found that out fighting the P-51 AI in 1366x768 then 1920x1280.

 

 

disagree here...lower resolution leads to dots beeing visible further out...of course the whole image quality is worse with lower resolution.

[NineLine]

disagree here...lower resolution leads to dots beeing visible further out...of course the whole image quality is worse with lower resolution.

 

Lower res might be better for distance, but I think higher res is better for close range and picking aircraft out with the terrain in the background...

[9.JG27 DavidRed]

^^yes absolutely!in close ranges definitely

 

cant really comment about smart scaling, as i havent played any sim where it was included in.at least to my knowledge.for me personally, i quite like how dcs models close contacts, and the only thing i think definitely needs improvement, is the distance where the dots start to be rendered, which should be increased.also it seems as if the distance where a plane transforms into a dot is too close.

Edited August 11, 2014 by 9./JG27 DavidRed

[karambiatos]

So is smart scaling, the thing shown in the video below, or does it work differently?

 

[ericoh]

disagree here...lower resolution leads to dots beeing visible further out...of course the whole image quality is worse with lower resolution.

 

 

At lower resolution objects tend to pop in later. Because there is no "small enought" pixel to render them earlier.

 

Then you would need to start smart scaling, or it would be nothing-nothing-to close.

[9.JG27 DavidRed]

all our tests within DOW suggest the total opposite....lower resolution will make dots visible further out.

 

maybe you are right with the "small enough pixel" argument, but then with native resolution the rendered pixels are that subtle and matching the background colour that good, that they are indeed invisible even if you know where to look at.(so basically they are not there, even if they might be), while with a lower resolution you can spot those pixels way further out.

[ericoh]

all our tests within DOW suggest the total opposite....lower resolution will make dots visible further out.

 

maybe you are right with the "small enough pixel" argument, but then with native resolution the rendered pixels are that subtle and matching the background colour that good, that they are indeed invisible even if you know where to look at.(so basically they are not there, even if they might be), while with a lower resolution you can spot those pixels way further out.

 

 

Did some quick tests. I understand what you mean. It seems like the objects tend to be bigger pixels BUT after a certain distance they dissapear compared to higher resolution.

 

 

1680x1050

 

1280x800

 

No higher resolution here. Pardon the small editing inaccuracys, i did this in gimp on my none editing machine (half heart attack, gimp is horrible.)

I think it illustrates the problem tho. 1680x1050 is 1 pixel blob. 1280x800 dissapeared at that point. Distance is 5.1 nm.

Edited August 11, 2014 by ericoh

[Wolf Rider]

No, it doesn't zoom. Used here, it's a conceptual conceit to bring a closer parallel to how the eye works correlates to what is taking place in the engine. The engine only "zooms" rather than permitting realtime change in DoF (which is what the eye is doing). Thus, when I say that the eye is "zooming", it's drawing it's focus out to a farther distance, which is only comparable in the simulation as actually zooming along the visual z-axis.

 

the eye doesn't zoom... it has the convex on the lens changing, meaning the conceptual use is incorrect -

 

 

It doesn't have anything to do with what is presented on the monitor. It does, however, have everything to do with what is visible to the *player*, and that is based on said distance.

 

Example: A fighter-sized object (say, a MiG-29 smoking in MIL) is on the horizon. Based on 20/10 vision, say that this should be visible in a head-on aspect at 14 miles with the representation of one lit pixel.

 

If you permit the changing of FoV (which is what zooming actually does, projection of smaller fields of view in the same apparent visual area), you're changing the apparent size of said object. A pilot strapped into his seat can, at most, get his eyeballs closer to an object (disregarding the net range collapse by his velocity at the object) by, at most, 6-8 inches of slack in the harness and traverse of his neck. The FoV is changing by thousandths of a percent at that 14 mile range. What he can change is how far out his eyes are focused to pick up that microscopic object, which is more apparent based on its relative sharpness to the rest of what is in his visual field.

 

 

in the real world, if that is your example above, there is no changing of the FoV... how the eye sees, is completely different to how camera sees.

 

I can accept you've given some thought to the 'problem' some suffer with spotting a target and your idea of locking the FoV and altering the DoF could have some merit... at least it would be good for screen shots.

 

 

Lighting, as you mentioned, and others have mentioned before, does factor in a bit - we'll have to wait and see what EDGE brings in there

Edited August 11, 2014 by Wolf Rider

[Einherjer]

Its also important to clean your monitor! Its frustraiting to hunt a black dot and then realize its just a dust particle on the screen ...

[lunaticfringe]

the eye doesn't zoom... it has the convex on the lens changing, meaning the conceptual use is incorrect -

 

Again, I don't disagree with that. The conceit is a comparison of what is happening with the eyes in a plane (focus), vs. what is happening with the eyes in DCS (zoom).

 

The eye changes its focus to see objects in the distance. The player in DCS uses the zoom function to see objects in the distance. It's different methods to achieve the same net result (a spotted target), but with different effects: in real life, apparent depth changes; in DCS, the target gets far larger than it should be.

 

If DCS represented visibility at range, and the changing of focus, I wouldn't be having to use the conceit. We could simply compare apples to apples and determine if it was represented correctly. We can't. Thus the fudge.

[9.JG27 DavidRed]

Did some quick tests. I understand what you mean. It seems like the objects tend to be bigger pixels BUT after a certain distance they dissapear compared to higher resolution.

 

for us its(or at least it was before the last patch), as i said, the opposite...probably you used MSAA?

[doveman]

I can accept you've given some thought to the 'problem' some suffer with spotting a target and your idea of locking the FoV and altering the DoF could have some merit... at least it would be good for screen shots.

 

That does sound like it would be cool, having a DoF adjust instead of zoom, so that the player can focus on different distances and scan, as he would IRL. Probably never been done in a game before though, so couldn't say if it will actually be usable in practice until someone implements it for us to try.

[leapingrodent]

Its also important to clean your monitor! Its frustraiting to hunt a black dot and then realize its just a dust particle on the screen ...

 

Yep... I've done this on more than one occasion... :music_whistling:

[Fri13]

Higher resolution clearly helps to spot targets. Now with changing from 1920x1200 to 3840x2160 resolution had made huge difference. Easy to spot individual tanks from 20km range and identify them in as well further.

 

Does make flying a planes totally different when spotting a fighter is possible from 30km range if they are moving.

 

Now it really looks that there is huge difference are under a net under tree, stationary on yellow field or moving overall.

 

Vehicle parked on asphalted area is very easy to spot. Stationary vehicle in complex textured ground is still hard to spot.

 

Now it just feels silly that I can see to 30km ranges in fighter, but none of my missiles is able to fly further than 5km from launch point. Makes BVR combat useless while now it really would had more difference.

[WRAITH]

Seriously you guys are driving me to drink.... more.

 

 

LoL :megalol:

 

Your to late on the mark i started 2009 when i joined the Forum Flight Sim World , now i have my own winery :megalol:

 

Sorry for O/T just couldn't resist :doh:

 

Now please continue as you may......................................

Edited August 11, 2014 by WRAITH

[Wolf Rider]

That does sound like it would be cool, having a DoF adjust instead of zoom, so that the player can focus on different distances and scan, as he would IRL. Probably never been done in a game before though, so couldn't say if it will actually be usable in practice until someone implements it for us to try.

 

yeah... LunaticFringe does seem to be saying like something similar to a telephoto lens type method (where the aspect doesn't change, but distance does)?

careful though, not to misunderstand; the human eye doesn't behave like a telephoto lens either

 

 

 

 

 

~

Now it just feels silly that I can see to 30km ranges in fighter, but none of my missiles is able to fly further than 5km from launch point. Makes BVR combat useless while now it really would had more difference.

 

 

 

yes, there is something in your mention of Beyond Visual Range (BVR)

Edited August 12, 2014 by Wolf Rider

[RagnarDa]

Just want to check in and say that this feature (smart scaling) is community moddable. For example, open up DCS World\CoreMods\aircraft\F-86\f-86f.lods and replace the text with something similar to this:

model={
   lods={
       {"f-86f.edm",2000.0}; 
       {"A-10A_L06.EDM",3000.0};
       {"C-130_lod3.EDM",4000.0};
       {"il-78m_lod04.edm",30000.0};
       {"ld_tec.edm",50000.0};
   };
   collision_shell="f-86f-collision.edm";
}

There should be a more seamless transition so ideally I believe a change of size about every 500m or so would be better. If I had 3DSMax installed I could have created better models. You really lose perception of distance though and its hard to tell if the target is approaching or flying away.