Spotting Distance... again

[NineLine]

As a side note, calling each others names such as "cupcake" isn't needed, and this will be the first and last warning on this. Then I hand out warning points. Play nice, we all want the same things, even if you can't see it.

[SharpeXB]

I've been playing flight sims since before Chuck Yeager's Air Combat and SWOTL - that was way back in 1991! In fact, I used to play Apache on an Amstrad PCW 9512 before that - an interesting experience but still cool even on a green screen.

Games of that era didn’t feature anything close to realistic graphics. So they aren’t very relevant to this issue. No doubt they employed all sorts of visual help which would look awkward in today’s games.

[Mars Exulte]

I believe that was back in the day of flat terrain with pyramids for mountains. You were lucky to see anything back then :p Good times...

[bell_rj]

Hi SharpeXB, I'm not claiming that those games are relevant. I was responding to your implication that I didn't know how to play a 3d game by pointing out that I've been playing them for a long time.

 

I agree that those games didn't feature realistic graphics. The early flight sims had skies of few shades of blue. In the air, long distance targets would be represented by a single black or grey pixel that on an MCGA screen would be easy to see. On the ground, landscapes would be large blocks of green or sand with few details. Targets comprised of gray or black pixels tended to stick out like a sore thumb.

 

In DCS we are blessed to have fantastic, complicated terrain textures. So units are quite hard to see on the ground but I think the balance is about right. It's not easy but not impossible to spot them. It feels realistic. I know others believe otherwise, and FLIR issues are a different beast again.

 

Let's talk about being in the air now though - and you'll notice that a lot of, if not all, recent comments on Reddit are referring to AA combat not AG combat - I think the balance is off. It feels too difficult to spot units without resorting to labels (which I personally find a bit immersion breaking). Let's keep labels in DCS - options are good, right? Meanwhile I have read NineLine's reply above which gives me some hope that ED are working to make things better for those of us who would like spotting to be easier. I also hope that whatever changes ED make to the graphics engine that they please those who thinks spotting is currently just right for them.

[Decibel dB]

but if you haven't tried the labels just using the dots, then you really need to

 

Hi, can you elaborate on that a little? Is it an option on the settings?

[msalama]

FoV isn't even a factor in this

 

Taking RL into account, it definitely is a factor. But never mind.

[Demon_]

At 11:50

Edited March 16, 2019 by Demon_

[Tippis]

Taking RL into account, it definitely is a factor. But never mind.

The rendering pipeline is not real life. It is not a factor in what causes the problems with how units are shown (or not) at a distance.

 

 

 

Of course, IRL, eyes are pretty fixed-FoV but have all kinds of cognitive faculties backing them up to compensate, and those are the ones that so many “simulations” fail to simulate.

[Mars Exulte]

Taking RL into account, it definitely is a factor. But never mind.

 

I really don't have a problem with people requesting improvements. I have a problem with people requesting 'fixes' for 'problems/flaws' that

 

A. They very clearly don't even understand the factors resulting in their perceived 'problem' (it's like asking a plumber to work on your car)

 

B. They do not recognise that it's not actually a problem but a very logical result of those factors, i.e. it's 'working as intended' they just don't like the results

 

It isn't about whether I think spotting can be improved or not, or whether or not it's hard, or whether I care whether features are added or not. It's about not wanting people who don't even know what they're talking about lobbying for game changes.

 

 

So, to make a list in one location:

 

Factors that influence spotting off top of my head are as follows

 

#1 Lighting and weather conditions (duh) Some conditions are more conducive to spotting than others

 

#2 Relative speeds and distances (your perceived ease of spotting with slow moving nearby WWII fighters compared to jets moving faster at greater distances) Stuff like Il-2/War Thunder etc are naturally easier to spot in because everyone is closer and slower moving meaning you have less opportunity to lose them once acquired. Combined with #3 this can result in it being muuuuuch easier

 

#3 Presence or absence of ''gameification'' tricks like dynamic modeling (which BMS uses to make it ''easier''), labels, etc

 

#4 FoV. Yes, this is a MAJOR contributing factor in ALL games, in fact it is a core principle in video game design across all genres, it's THAT big of a factor. Anybody who can't recognise that should stay out of technical discussions altogether.

 

#5 Hardware. As frequently mentioned, monitor size and resolution, framerates, graphic settings, etc, are all also major factors. If you are four feet from a 24'' 4k monitor you will have a harder time noticing stuff than if you're four feet from a 65" 1080p monitor. No way around this, it's going to vary with every person you talk to.

 

#6 Training and personal habits. I frequently in these threads see people say they can't even fly in formation with other people, that's outright bs lol Or losing track of them while in WVR. This is the L2P factor. Yes, it's frequently a bs argument, but it actually IS a factor in many cases, and as somebody who would regard himself as being in the top 10-20% of pilots, I am quite comfortable in saying it is a factor in nearly every grievance barring blatant bugs or issues. Even in DCS, the ''average'' pilot is of questionable competency and tend to cry foul unjustly, just like with EVERY OTHER GAME EVER :p It's NEVER their fault

 

 

 

''Solutions'' presented are thusly:

 

#1 Magic. Somebody unspecifically demands somebody make it better because reasons. No real plan is presented or reasoning utilised. Just 'fix it'. Usually something like ''make it more visible somehow'' At least they're honest about not having a clue.

 

#2 Dynamic LoDs/modeling. Adjusting the size of objects to appear larger than they really are. Yes, it enables you to use your unnaturally wide FoV without sacrifice. It is also the gamiest solution as it involves physically distorting an object so you can see it easier. It won't make any difference if your problem is hardware or user related.

 

#3 Labels. Can be a very gamey option depending on the level of ''labeling'' utilised such as full name and distance brightly colored depending on faction. Can also be a simple speck over/near another speck, in which case its impact is minimal across the board. This is imo the best ''compromise'' for people who absolutely cannot see stuff. It is an 'aide' but relatively discreet, and not so ludicrous as enlarging aircraft at a distance to the size of a 747, when the people crying the loudest probably STILL won't see it.

 

#4 (my camp) There is no solution because there is no problem. This is what you get when you render objects at true scale. You're ok with magically enlarged aircraft and ''shoot here'' labels, but using your scroll wheel to adjust to your FoV to something reasonable for specific conditions is from Satan? The following analogy is applicable :

 

You have a radar A2A fighter, with an adjustable scale for displayed range, like any of the FC3 aircraft for example. It scales from 5 to 100. You CAN lock it in to that longer display, but you lose granular detail, especially on closer targets. So what do you do? You demand ED fix their f'ing game and add dynamic scaling of radar contacts to compensate for your own incompetence and/or stubbornness. Jk, you lower the scale when you need to under certain conditions.

 

Visuals are the same. At close range it is 100% a competency issue, within a few miles you can see any aircraft in DCS under any condition, any FoV, no exceptions. Anybody claiming otherwise is spouting bs. If you're not seeing it it's because you're not seeing it. It's there.

 

The ''semi-legitimate gripe'' is for pre-merge initial contact. And THIS is where your FoV is a dominant factor. If 50-60 degrees is ''scale'' but you're zoomed out to 100+ so you can see all/most your instruments without moving, then logically any objects in the distance will be at least half their ''apparent'' size as a result (probably more like 1/4 size if my reasoning is correct). This is a clear, logical result of YOUR chosen view setting. It will affect dynamic scaling as well, if implemented, labels however will not be affected. So, even with the dynamic models, to see properly you will STILL need to ''zoom in'' some because you're too far back.

Edited March 16, 2019 by zhukov032186

[Tippis]

#4 FoV. Yes, this is a MAJOR contributing factor in ALL games, in fact it is a core principle in video game design across all genres, it's THAT big of a factor. Anybody who can't recognise that should stay out of technical discussions altogether.

 

#5 Hardware. As frequently mentioned, monitor size and resolution, framerates, graphic settings, etc, are all also major factors. If you are four feet from a 24'' 4k monitor you will have a harder time noticing stuff than if you're four feet from a 65" 1080p monitor. No way around this, it's going to vary with every person you talk to.

 

Again, no. FoV is not a factor in the underlying problem here because it does not alter how large pixels are. There's no need to do backhanded snipes about people just because you're not talking about the same problem they are.

 

Hardware matters, yes, but in the opposite way you're thinking, where worse hardware gives better results because that changes the size of pixels (and the aliasing – or lack thereof – between them).

 

You can compare different circumstances to yield different results, sure, but that's not the issue. The issue is that, like for like as far as the viewport into the world goes, higher resolutions and higher-quality graphics settings make spotting harder (and note, I'm not arguing that it should make it easier either — I'm saying it shouldn't matter). FoV is not a factor because (VR zoom aside) FoV is universally applicable and part of that like-for-like comparison. Two players at max zoom level will have very different opportunities to spot long-distance aircraft due to how better graphics hardware makes them harder to see.

 

Gitting gud does not solve this problem, because it does not change how the rendering works

Gitting betta hardware does not solve this problem — it actually makes it worse, again due to the rendering process.

 

Only a solution that tries its hardest to balance out what you see in a like-for-like situation does, and there are indeed methods for doing this (and without unduly affecting high FoV). The only really tricky part is determining what the baseline should be, and withstanding the barrage of whinging for people above and/or below that baseline. :P

 

This is what you get when you render objects at true scale

And that's the problem in a nutshell: true scale is an inaccurate simulation of visual perception. Eyes are not CCDs; vision is not pixel-based. To properly convey vision, you have to translate those cognitive processes into pixels, not the other way around.
Edited March 16, 2019 by Tippis

[Mars Exulte]

Tippis, FoV does not change the size of pixels, no... It changes how many pixels are used to render an object. 10' FoV = object great big and four inches wide, obviously many pixels. 100' FoV object perhaps only a single pixel in diameter, therefore much harder to see. FoV is a factor in all games, it's why the ships in Mass Effect have 30ft vaulted ceilings to accomodate the 3rd person wide FoV. If the ship was built 'to scale' the camera would be clipping into the ship or into space.

 

No, hardware is not 'opposite' what I said. That's exactly what I indicated. A larger lower resolution monitor has larger pixels than a smaller higher resolution monitor, and therefore an easier time spotting objects rendered as a ''few pixels''. We are in agreement. The closer you get however, the advantage shifts back to higher resolution as the objects become more clearly defined, although at this point it's not really a significant factor anymore anyway.

 

 

 

True scale is by definition an ACCURATE SIMULATION. Ffs, that doesn't even make sense to say what you just said o.O ACCURATE SCALE combined with INaccurate FoV is the issue, as the object is rendered at the correct size, but your FoV to allow greater situational awareness causes it to be smaller than it should be if you were in cockpit.

 

We can argue about what if any solution to apply, but there's no debate about the CAUSE. It is 100% FoV related, except where glitchey or incorrect LoDs etc are a factor. My ''snarky'' tone is because this is an extremely obvious fact that people are arguing about. It's like people telling me the sky isn't blue. Yes... yes it is. We might not like it, but it is indeed blue.

Edited March 17, 2019 by zhukov032186

[bell_rj]

Hi NineLine,

 

Thank you for your interesting reply. I appreciate that you have addressed the issue and not tried to dismiss it as a small number of others have.

 

The insight you have given us some insight into EDs attempts to improve visual cues is very interesting and it would be great to hear more on this as soon as ED are able to share updates.

 

I enjoy the challenge of DCS - I don't want a dumbed down version necessarily - although having options like easy avonics and easy flight model is great for helping less proficient fliers overcome what can be a very steep learning curve. As they move up the curve they can turn off the easy mode options. I see spotting in the same way - it can be difficult and practice helps, the right technique helps, setting FoV correctly helps, labels help, etc. Starting with labels is a good option to start with, but I believe that the spotting learning curve is a bit too steep after that and I'm hoping for more improvements in the middle of the learning curve - like the improved visual cues you mention.

 

As you recommended, I'm trying labels again. Especially now I know how to edit view.lua to customise label draw distance.

[imacken]

Did you guys see the images I posted here https://forums.eagle.ru/showthread.php?t=235606

Even at max zoom, the aircraft cannot be seen on my 4K monitor. They can be with labels, so they must be within 5 miles.

Interesting comparison.

[Jethro11]

I would like to see improvements in this too. Especially in VR. I played a mission the other day and jester was saying the plane was right in front of me and I never saw it. I even opened the map to see where it was and it’s altitude and went back in game and still never found it. I went outside view and panned around but never spotted it. I just gave up and closed the game down. It just wasn’t fun for me, which is to bad because it’s a great game in so many ways.

 

Hope they fix spotting soon, or give us some options to adjust size.

[samich]

Thanks nineline. Appreciated you weighing in on this. I'd love to hear about eds work towards vulkan and anything they're working on in a news letter at some point down the line.

Things like reflection, engine smoke density, contrails would all really help with visibility but I hear what you're saying about having to work within the constraints of the majorities hardware and I certainly wouldn't want to lose the depth and accuracy that makes us all love dcs. this is principally why I think slight aid is necessary even if it must be artificial.

 

Labels solve all issues, but things like being able to see them through the aircraft frame really break the immersion so I avoid them.

[Tippis]

Tippis, FoV does not change the size of pixels, no... It changes how many pixels are used to render an object.

And that is the heart of the problem: the number of pixels, which allows the same target to appear at different sizes on different hardware, making it harder to spot on (what's commonly thought as) better hardware.

 

The reason FoV — in the sense of user-adjusted viewport — isn't part of the issue is that translation from world size to screen size would (or should) nullify that part anyway. Yes, there would still be the issue of how well you can see the screen, but the part of the equation that decides what's on the screen can still be controlled and unified.

 

No, hardware is not 'opposite' what I said. That's exactly what I indicated. A larger lower resolution monitor has larger pixels than a smaller higher resolution monitor, and therefore an easier time spotting objects rendered as a ''few pixels''. We are in agreement. The closer you get however, the advantage shifts back to higher resolution as the objects become more clearly defined, although at this point it's not really a significant factor anymore anyway.

I don't think we disagree on the mechanics of it — we disagree on how it relates to the problem of how distant objects are rendered. You seem to claim that it's not an issue that higher resolutions and better graphics results in harder spotting; I say that this should not be the case and that it can indeed be solved.

 

True scale is by definition an ACCURATE SIMULATION. Ffs, that doesn't even make sense to say what you just said o.O

True scale is not how you would accurate simulate vision, because that's not how the cognitive process of perception works. It may be mathematically accurate in terms of angular coverage, but it is not the be-all-end-all answer to what can be seen. Indeed, if anything, that angular coverage should be able to provide a lower limit at how small something can be before it disappears for everyone, regardless of screen resolution and zoom setting.

 

The key distinction here is the notion of what we are simulating. We are talking about spotting aircraft. It is a discussion of simulating perception. As trite as the adage may be, perception is not reality. Physical, actual size is a factor in that perception, but it is far from the only one so true scale cannot be an accurate simulation. You have even listed some other factors already — contrast and movement in particular. Model scaling is often decried as unrealistic, when in fact, that's the best way of replicating the effects of such cues (at least until monitor technology gets a major overhaul).

 

We can argue about what if any solution to apply, but there's no debate about the CAUSE. It is 100% FoV related, except where glitchey or incorrect LoDs etc are a factor.

Well, that's just untrue. There is debate about the cause. It is 100% unrelated to FoV. It is a function of how there is no unified bound for how small (or large, for that matter) an on-screen item can be at a given distance.

 

Put another way, the cause of the issue is that raw resolution is allowed to dictate the size, as opposed to angular sizes adjusted by other visual cues, and that no attempt is made to unify how those are translated evenly into on-screen smudges.

 

 

 

…and of course, there's also the issue that ED has hinted at, that not only does display resolution affect spotting, but it also affects sensors such that better hardware yields worse sensor performance. That alone — if true — warrants a total overhaul of how unit sizes are being handled.

Edited March 17, 2019 by Tippis

[SharpeXB]

Model scaling is often decried as unrealistic, when in fact, that's the best way of replicating the effects of such cues (at least until monitor technology gets a major overhaul).

Model Enlargement was already tried in DCS and turned out to be terrible. When you draw objects out of scale and larger than their surroundings it enhances their visibility too much. Trucks are the size of battleships. It doesn’t work. And yet even at the largest setting people still complained that they couldn’t see stuff. I doubt ED is going down that road again.

[wilbur81]

Model Enlargement was already tried in DCS and turned out to be terrible. When you draw objects out of scale and larger than their surroundings it enhances their visibility too much. Trucks are the size of battleships. It doesn’t work.

 

Very much agreed. My only issue with 'spotting' in DCS is MP related: A person playing at 4K is at a disadvantage playing against someone with 1080p. Again, I wanted to buy a large 1080p TV with low input lag, but couldn't find anything over 50 inches. Everybody had already gone 4K. So, I bit the bullet, accepting lower performance. Then, in my first MP session after having upgraded from 1080p to 4K, I was shocked at how blind I was. Before 4K, I had thought I was pretty good at WVR ACM/BFM. :-)

[Tippis]

Model Enlargement was already tried in DCS and turned out to be terrible. When you draw objects out of scale and larger than their surroundings it enhances their visibility too much. Trucks are the size of battleships. It doesn’t work. And yet even at the largest setting people still complained that they couldn’t see stuff. I doubt ED is going down that road again.

It works just fine if done correctly, which is something quite different from what you're describing. Coincidentally, a truck scaled up to the size of a battleship would have to be at such a distance that you would not be able to spot it anyway so you wouldn't see it regardless. :P

 

Per ED's comments, the reason it turned out terrible in DCS was because of how it affected the sensor detection code, which is a separate issue (or… rather, it should be a separate issue, and the fact that they're supposedly related means there is something in desperate need of fixing irrespective of the whole draw scale discussion).

[SharpeXB]

It works just fine if done correctly

There’s no way to do it correctly. The size of an object has only one correct definition. If you scale objects up artificially it creates more problems than it solves. It has the effect of actually enhancing far away contacts. Already been tried and failed. No reason to waste time going down that road again.

I should find the really goofy screenshots back from then where you could see other aircraft taking off from bases 80 miles away.

Edited March 17, 2019 by SharpeXB

[dawgie79]

The size of an object has only one correct definition.

Not on a monitor it hasn't. On a monitor, screen size in relation to resolution comes into effect. Ergo, the physical size of the pixels used in the screen. They are smaller in a 32" 4K than in a 32" 1080p. Hence more difficulty spotting on a 32" 4K.

[Tippis]

The size of an object has only one correct definition.

No, it has a whole bunch of different definitions, starting with such a simple one as “what measure of size are we talking about?” Non-relativistic expanse in 3D-space is no more correct than any other, and especially not when we want to discuss the matter of perception — an area where there numerous subfields of study dealing with various optical illusions that put any such notion to shame.

 

If you scale objects up artificially it creates more problems than it solves.

…such as…?

 

It has the effect of actually enhancing far away contacts.

That's if you do it wrong. You can also choose to do it right and not have the draw scale be in any way related to sensor scale. Just because it was once done wrong does not preclude it from being done right.
Edited March 17, 2019 by Tippis

[SVgamer72]

Pixel densities 1080p vs 4K

https://www.noteloop.com/kit/display/pixel-density/

 

The solution is to scale based on resolution so contacts are rendered at the same relative size regardless of resolution.

 

No scaling:

 

With scaling based on resolution:

 

 

Sent from my iPhone using Tapatalk

[wilbur81]

I'm not the brightest bulb in the knife drawer, but this makes a lot of sense. ^^^

[SharpeXB]

Not on a monitor it hasn't. On a monitor, screen size in relation to resolution comes into effect. Ergo, the physical size of the pixels used in the screen. They are smaller in a 32" 4K than in a 32" 1080p. Hence more difficulty spotting on a 32" 4K.

I’m talking about the “size” in the 3D world, not the angular size on a screen. And if you’re talking about objects the size of a single pixel, a single black 4K dot would equal 4x 25% grey pixels on a 1080p, so larger but softer and less defined. You’re better off with more resolution.

One reason the Model Enlargement imposter system was flawed was that, yes limiting sprites to a certain number of pixels made them bigger on lower res screens.

 

No, it has a whole bunch of different definitions,

No it’s really simple. A truck is 25’ long in the 3D world so it stays 25’ long no matter how far away you are from it. If everything else shrinks in perspective except the truck, eventually the truck is rendered the size of a battleship. Model Enlargement where objects stopped shrinking with perspective is gone from DCS. It’s bonkers. It was tried, failed and now it’s gone.

Edited March 17, 2019 by SharpeXB