Aircrafts visible 15 NM away without zoom and model enlargement

[Bushmanni]

You can see objects from way too long distances now even with wide FOV. Although the object is only one pixel wide it makes use of radar unnecessary, especially against skylined objects. You can also easily see the missile objects themselves from beyond 10NM. I know some people will rejoice with this but it renders BVR combat unrealistic as you can find enemies too fast and easily using just eyes instead of radar.

 

I haven't tested exact ranges yet but obviously seeing F-15 from 15 NM away with default FOV and also the falling droptanks is way too good visibility. Default FOV should amount to about 2-4 NM visibility (off-center vision) and max zoom to about 10 NM (central fovea vision) for fighter sized targets according to literature.

[Paradox]

In a clear sky 15nm is probably on the longer side of realistic to spot a large plane to be honest

[atsmith6]

You can see objects from way too long distances now even with wide FOV. Although the object is only one pixel wide it makes use of radar unnecessary, especially against skylined objects. You can also easily see the missile objects themselves from beyond 10NM. I know some people will rejoice with this but it renders BVR combat unrealistic as you can find enemies too fast and easily using just eyes instead of radar.

 

I haven't tested exact ranges yet but obviously seeing F-15 from 15 NM away with default FOV and also the falling droptanks is way too good visibility. Default FOV should amount to about 2-4 NM visibility (off-center vision) and max zoom to about 10 NM (central fovea vision) for fighter sized targets according to literature.

 

What are your model enlargement settings at?

[Kippy]

Bushmanni, what monitor setup are you playing the game with?

 

I play with a small 1280 by 1024 monitor and have trouble even visually acquiring aircraft 4 miles away. Meanwhile my buddy in the back seat of my L-39 was spotting contacts at absurd ranges. He was playing on a 34 inch TV.

 

How are you suggesting we fix this? Reduce the visibility range throughout the entire game, making the situation for players like me even worse?

[0xDEADBEEF]

Bushmanni, what monitor setup are you playing the game with?

 

I play with a small 1280 by 1024 monitor and have trouble even visually acquiring aircraft 4 miles away. Meanwhile my buddy in the back seat of my L-39 was spotting contacts at absurd ranges. He was playing on a 34 inch TV.

 

How are you suggesting we fix this? Reduce the visibility range throughout the entire game, making the situation for players like me even worse?

 

With all due respect, I dont think Ed should spend a lot of effort optimizing DCS for players who play on, excuse me, ancient hardware. This is nothing personal, development should go ahead instead of backwards.

 

I play on 1080p on a 24 inch and I can see contacts 50km/25nm away frontal aspect with my eyeball Mk1 very easily! While I do like how that new system looks, ranging and visibility through clouds absolutely need some more work. I do like it much better than the old imposter system I have to admit.

[AtaliaA1]

Humans..............We are never satisfied.

 

I must go with the advances in technology must be followed or we dare go backwards.

[Fri13]

I think they just need to little bit optimize it, as larger models you can spot from far, like a A-50, over 30km. But just spot it against clear sky (default weather). At about 15-20km you can see F-15C size targets, again against clear sky.

 

But then comes F-5E II or Mig-21Bis and you lose it easily from sights just in 2-3km ranges. Even F-16 is difficult to spot over 5km range.

 

I haven't tested deeply that anymore since 1.5.5 was released but at least the ground units are little too easy to see as from about 15km I can spot them. I should be do that at about 10km range when they are moving clearly uncovered. But stationary it should be closer to 1.5-2km to spot stationary vehicle that isn't camouflaged but isn't parked on the field or streets/roads. And if concealed then just really not to easily spot it at all.

[Kippy]

With all due respect, I dont think Ed should spend a lot of effort optimizing DCS for players who play on, excuse me, ancient hardware. This is nothing personal, development should go ahead instead of backwards.

 

I play on 1080p on a 24 inch and I can see contacts 50km/25nm away frontal aspect with my eyeball Mk1 very easily! While I do like how that new system looks, ranging and visibility through clouds absolutely need some more work. I do like it much better than the old imposter system I have to admit.

I have to say DEADBEEF, I think you are correct. I agree that ED should spend time working forward instead of making sure people with older hardware can play. But also understand that not everybody can afford the newest monitors and hardware. We risk potentially losing existing customers. That's not something we want. I'm no expert, far from it, but I would like to hear from somebody with the experience whether or not ED could peruse an option to make the visibility ranges stay for smaller res displays yet decrease them for larger displays.

 

DCS and it's modules are all "equal opportunity" and are independent of wallet size. There is no 20$ upgrade package for the AGM-65s that increases their effectiveness. Visibility range however.. :lol:

 

Jokes aside, I'd like to see this fixed with both playerbases in mind. Either way, I'd buy a new monitor if it meant that I'd have to quit DCS otherwise.

[Bushmanni]

Rendering of objects only few pixels wide is much better now but the maximum distances where the dot is visible are way too long. I use FullHD 46" TV on which a single pixel dot is well visible but I know that it's not that hard to see on a 24" FullHD monitor either that I used to use.

 

Human vision has a high acuity area in central fovea and less acuity around it. This creates the situation where you can see object clearly from far distances if you know where to look but usually are able to spot them only about 1/4th of the max visible distance. If you scan an area with the central vision you can spot an object within that area from the maximum visible distance but it takes time and work to do that. What the rendering system should do is replicate this physiological phenomenon. The way DCS handles this is giving better spotting distance when using more zoom. IMO I find this the best method so far that I have come across.

 

There's a paper named "Visual Search in Air Combat" that has a graph that shows maximum detection ranges (central vision) for different fighters in different aspects. The farthest visible plane in the graph is F-14 in belly view with a visible range of about 12 NM. Shortest range is for MiG-21 head-on view with 2.5 NM maximum visible distance. The average spotting distance with off-center vision which should be about 1/4th of those said distances would make MiG-21 visible only about 0.8 NM away. I couldn't see any info of the exact conditions of the environment where these spotting distances should be happening but it's supposedly some generic typical situation where a fighter pilot would be looking for the bandit and is in line with other sources.

 

Essentially the rendering engine should somehow use the FOV as determining factor for object visibility. If you have larger FOV, single pixel sized objects should be less visible. As we are mostly concerned with sub pixels sized objects, the in-game visibility is a function of contrast of the pixel with it's surroundings. So the single pixel sized objects contrast should be a function of FOV. This would greatly even the playing field between different resolutions as the pixel visibility is tied down to FOV rather than monitor resolution. The exact way how to make this happen is not that easy to figure out though.

 

I think for some time there won't be a monitor or display device that can surpass human vision so we are safe to assume the object is reduced to sub pixel size well before it should vanish from view. When the object is single pixel sized it's contrast with the background would be altered so that it satisfies the expected visibility of said object at the current FOV. Basically it would remain at full nominal contrast for some distance and then start losing contrast so that it's about maybe 20% visible at the nominal max visible distance and then keep fading linearly until it disappears. 20% contrast at max spotting distance might be still too much but you can alter it to a value where it is reliably but just barely visible.

[Diemos]

I don't know where all these ranges are coming from, but anecdotally, I can stand in my backyard and watch aerobatic planes flying around in their designated zones, and they are easily visible at 20 - 25 miles... Much smaller than a fighter...

 

Then again, my wife can barely see them...

 

Think of it as the difference between people's RL vision... :thumbup:

[Art-J]

I think it's the first ever thread on these forums where someone suggests lowering target spotting distance :D. Numerous threads on the subject are full of contrary opinions, and for a good reason I personally would say, flying mostly WWII/Korea planes on a 24" 16:10 monitor, where loosing contacts within 1-2 km is annoying norm rather than exception.

 

I agree, however, about Your FoV-related suggestions. There has to be a better way than current imposters system, which kind-of-fixes a few problems, while creating twice as many others.

[Nefaro]

<..>

flying mostly WWII/Korea planes on a 24" 16:10 monitor, where loosing contacts within 1-2 km is annoying norm rather than exception.

<..>

 

 

I use the same size monitor/ratio. Feel like I'm blind past a mile, maybe two if I'm lucky.

 

But it's not only the displayed size of the contact, but the color, when they're close, that makes it even more difficult.

 

I suppose it's the gfx engine that makes the various colors drab & washed-out. Too similar. Makes visual spotting of aircraft difficult because they all look equally washed out, almost the same color, when viewed against the terrain. No camo required. :cry:

[Bushmanni]

I took it upon myself to explore object visibility more and see if I can figure out some kind of pixel visibility system using FOV as parameter. I first started with determining how well you can actually see objects with naked eye. I calculated how far away you need to be from 24" monitor for the monitor to have better angular resolution than average eyes (about 1.5m). Then I took a screenshot of F-15 from DCS and calculated how far it would be when resized to certain size and looked from 2m distance on a 24" monitor. I then made a picture that has the F-15 depicted on several different distances and an angle scale that you can use to determine how well you can see the planes with your off center vision. I will refer to these images as super acuity images.

 

You need to look these images at 1:1 scale on a 24" monitor from 2.0m distance for the objects to be in correct scale and for the angle scale to show correct angles.

 

Head on

 

Side

 

Top

 

Device comparison

Leftmost is super acuity, then FullHD 24" monitor, Oculus Rift with 1.0 PD and far right Oculus Rift with 2.0 PD. 24" monitor pixel size assumes in-game FOV is set to same as actual monitor FOV (55 degrees). For the Rift I assumed 94 degree FOV. Angle lines are meaningless in this image, just forgot to remove them.

Device comparison with ground background

From left: super acuity, FullHD 24", Oculus Rift 2.0PD

 

Observations:

Maximum visibility distances were somewhat longer than you would expect based on the paper I referred before. This could be explained by not simulating atmospheric haze that causes the object to have more lighter color the farther it is. I might also have just much better vision as average person. The off-center vision spotting distances were as expected or even shorter. You can use the angle scale to see how far you can see the plane with different amounts of off-center angle. I would consider 10 degrees off as a typical value for spotting when randomly looking around. When systematically scanning, you would expect to spot the target with much smaller cone but probably still not with the central vision.

 

I also tested how simple "perfect antialiasing" method would work for spotting using pixellate filter on image editor. Against sky or other single color backgrounds it seems to work quite well giving similar contrast for the single large pixel as for the super acuity image (when looked from 2m distance). But when you change the background to ground, this doesn't apply anymore as the single pixel gets lost in the cacophony of other pixels. If I darkened the pixel in Rift image sufficiently the plane could be seen with similar contrast as in the other images. You definitely need some kind of extra algorithm besides high quality anti aliasing for working out the pixel color to get a correct contrast with different backgrounds.

 

Using FOV as visibility parameter doesn't seem very easy task but I'll explore this more.

[Schmidtfire]

Spotting is very subjective as it is dependant on eyesight/color sight and hardware like monitor/VR, graphics settings, resolution etc. It's difficult to find a sweetspot that works for everyone, but I think ED is working on a new spotting system, the impostors seems to be gone with the 1.5.5 patch.

 

Btw. Giora "Hawkeye" Epstein of the Israeli Air Force had a very good eyesight and could spot aircraft at a distance of 24 miles...

[Bushmanni]

Yep. I think a good goal for our virtual eyes would be average fighter pilot, ie. above average but nothing super exceptional.

[Haukka81]

Yep. I think a good goal for our virtual eyes would be average fighter pilot, ie. above average but nothing super exceptional.

 

 

 

+1 for this

 

And it would be nice to get zoom more real, now stock is quite overdone. Feels like super good binoculars.

 

 

 

 

Sent from my iPhone using Tapatalk