Aircraft Visibility

[Falcon_S]

Websites have previously adjusted and tested on the monitors of 15inch to bigger, in other words the site had a good look at 800x600 - thats minimum in web design in that time. Had to rely on the hardware of users who visit the site although there were a lot of users with 1024x768 (around 25% in year 2000) and higher. As of today, 99% of visitors have a screen resolution of 1024x768 pixels or higher and today only 34% of visitors have higher res. than 1920x1080. So, webdesign still take care of users 1024x768.

 

http://forums.eagle.ru/showthread.php?t=126222&highlight=graphic+card Excellent survey shows many and fair to the players, according to the percentage of time to forget the NV GT430 or ATI HD5550. This would be also in this topic but visibility are tested at higher resolutions, not listed minimum to us and now we have a problem. I believe that here is a small percentage users of the higher resolution than 1920x... Perhaps the DCS should be based on lower resolutions.

 

However, this is a game and the game sales hardware

[159th_Viper]

Perhaps the DCS should be based on lower resolutions.

 

This is 2014. A monitor with default res of 1920x1080 should be eminently affordable for all flight sim enthusiasts.

[Falcon_S]

Perhaps the DCS should be based on lower resolutions.

This is 2014. A monitor with default res of 1920x1080 should be eminently affordable for all flight sim enthusiasts.

 

I mean for testing.

 

And only 32.89% have primary resolution 1920x1080. http://store.steampowered.com/hwsurvey

 

What I'm saying is that it must not be the minimum resolution for testing. Just that.

[RagnarDa]

If you are looking for research and RL data instead of just speculation, then have a look at my first post in this thread, and try to download and go through the whole paper. It's generally not easy to find hard data on this matter, and by going through the various charts one can appreciate how complicated the issue of aircraft detection is IRL.

 

http://forums.eagle.ru/showthread.php?t=119885

 

btw, i hope that the visibility issues are improved after EDGE regardless of the method adopted. Right now DCS is the only sim where i find it extremely hard to spot contacts, even at distances of 1-2 miles.

I created a little mission and script to compare spotting distance with the data. You'll notice that at some angles the a/c doesn't even get drawn. Please play around with this and report your results.

Visibility test.miz

[159th_Viper]

I mean for testing.

 

And only 32.89% have primary resolution 1920x1080. http://store.steampowered.com/hwsurvey

 

What I'm saying is that it must not be the minimum resolution for testing. Just that.

 

A Steam poll is not really indicative of the Flight SIM Community and DCS now is it. And yes, one must draw a reasonable line somewhere. LockOn/DCS has always been known for pushing the boundaries of graphics fidelity - that's no surprise.

 

As said, one must draw the line somewhere and in this day and age I dare say 1024x768 is definitely not it as far as DCS is concerned. Then again, that's just my opinion. What we need is a survey of DCS resolutions but then again, not many members are forumites so even those results will have to be taken with circumspect.

[Falcon_S]

I agree Viper.

 

That's it as it. Leaving me to do nothing but get better monitor and this is the answer to this topic. (Maximizing plane visibility distance?) :)

[kazereal]

I mean for testing.

 

And only 32.89% have primary resolution 1920x1080. http://store.steampowered.com/hwsurvey

 

What I'm saying is that it must not be the minimum resolution for testing. Just that.

 

Remember that flight sims are not web pages.. Same minimum level should not be expected for both.

 

Also simulators have always been very resource hungry and pushing the boundaries of what can be achieved on any given hardware. Designing for minimum level does not advance simulations at same pace that hardware keeps improving.

 

Also current plans are focused more on efficiency of rendering (plus more control in the application side) so even if hardware does not take leaps the next generation of 3D API are much more advanced in this regard (meaning DirectX 12 and OpenGL-NG are taking note of AMD Mantle).

 

If it takes 4-5 years for new version of graphics engine emerge the minimum of rendering power also has time to increase by huge amounts. It makes more sense to design simulation for current high-level which is soon mid-level performance and poor performance by the time next engine iteration is complete..

[Falcon_S]

Remember that flight sims are not web pages.. ...

 

:huh: It was supposed to be parallel thinking about users.

 

Websites...

 

...However, this is a game and the game sales hardware

 

And, again... I must get better monitor. :crash:

[159th_Viper]

In-SIM @ 1360x768

 

 

 

 

In-SIM @ 1024x768

 

 

 

 

Again, Good or Bad?

[Wolf Rider]

 

~

 

 

~

 

Here's a quick comparison between my default resolution (3440x1440) and a reasonable default in this day and age I would have thought (1920x1080).

 

Target in view is a F-15 with B-52 at 5km:

 

Default View

 

~

~

~

~

 

Good or Bad?

 

 

 

:thumbup:

Edited August 19, 2014 by Wolf Rider

[Falcon_S]

Thanks for the effort Viper, it`s good of course. :) I would be happy to be the same but something is wrong with my monitor definitely.

[Falcon_S]

Is it possible to somehow balance between FOV and ZOOM. Real pilot not zooming when he move head forward-back and not lose to much FOV? If I expressed properly?

[159th_Viper]

Is it possible to somehow balance between FOV and ZOOM. Real pilot not zooming when he move head forward-back and not lose to much FOV? If I expressed properly?

 

That's the problem with the hardware we have and the penalty for sitting behind a monitor vs the Real World: Certain concessions have to be made and the necessity to zoom is unfortunately one of them.

 

Real-Life pilots, airline and otherwise, in the current test team are all of the opinion that it's harder to spot targets in real life than it is in-SIM currently. Again, when making the comparison it is imperative that one keeps in mind the distinction of sitting behind monitor vs in a real plane and accepting that we will never have the same view in both instances: Concessions have to be made, ie zooming.

 

All things considered, I'm not saying that by posting the above screenshots that all is well within the SIM: I don't know, hence the screenshots I posted of a F-15-sized target.

 

If there is info out there that the sizing/visibility is wrong then pls let me know, keeping in mind that zooming is essential to bridge the inherent shortcomings of our flight SIM equipment.

[kazereal]

:huh: It was supposed to be parallel thinking about users.

 

I'm not sure where you were aiming at but desktop resolution is not same thing as the resolution used within games or simulations since they are independently configured..

 

Edit: Steam hw-survey includes only Windows "desktop" screen resolution and most people (I think) run games in higher than that.

Some (few people I've met) keep Windows desktop resolution lower to keep fonts readable in programs such as browser.

My sample size is too small to determine anything from it actually..

 

Edit2: Some of those run games in second "screen" with video projector and only Windows desktop on "main" screen

but that's not too common perhaps..

Edited August 19, 2014 by kazereal

[Falcon_S]

You are right, to display games we need resolution from keyboard. :doh:

 

I have nothing to add more too... :pilotfly:

[kazereal]

Is it possible to somehow balance between FOV and ZOOM. Real pilot not zooming when he move head forward-back and not lose to much FOV? If I expressed properly?

 

You mean moving cockpit "camera" forward/backward?

(Not zooming in that case, FOV should stay same)

[airdoc]

Alright, here are the results of a quick test I did, comparing the apparent target size of a real life object to an object we see in DCS through our monitors.

 

The test evaluates only the "apparent size" of an aircraft, which essentially means the degrees of visual field it occupies (and reflects the size of the retina that processes it).

 

I used the P51 as a reference aircraft, which has a length of 9.83 meters. Let's assume that in reality one sees a P51 from the side, at a distance of 300meters (this distance is chosen for easier calculation purposes).

 

Have a look at this graph (sorry for poor drawing, it was done in a hurry in paint).

 

 

 

At point A is the location of the human eye. The triangle base E-G marks the location of the P51. Theta is the angle at the top of the triangle that reflects the apparent size of the target.

 

At 1 meter from point A, i have hypothesized the location of the monitor (i think it is a fair compromise).

 

If an object in real life at 300 meters distance evokes theta angle for its apparent size, then, if a simulator is to recreate the apparent size precisely, it should draw it at such dimensions that the same theta angle is formed at point A, i.e. E-G (9.83 meters) should be transformed to B-D.

 

Doing some quick math :

- tan (theta/2) = 9.83/(2*300)=0.01638. Using a tan calculator theta is found to be 1.876 degrees of apparent size.

 

-Now for BD : tan(theta/2)= BD/(2*1)=0.01638=> BD=0.032m(3.2cm)

 

So, in essence, someone who sees a P51 in reality at 300 meters, should see it appear approximately 3.2 centimeters long at 1 meter distance (and about 1.6 centimeters at 50 centimeters distance).

 

I went back to DCS and set 2 P51s at 300 meters distance, flying vertically to each other and measured the size at my monitor at different FOVs:

 

 

1. At the Standard Zoom : the P51 appears to be about 9 milimeters long.

 

 

 

 

 

2. Increasing the zoom to get a P51 size of 1.6 cm :

 

 

 

3. Increasing further the zoom to get the actual size of 3.2 cm :

 

 

here is a pic of the map showing the distance of the 2 airplanes at 0.19 nm (300m) :

 

 

 

 

In essence, what I get at my settings (1080p, 24 inch monitor) is roughly 3 times smaller targets than I would see in reality, using the standard zoom. I have to use a higher zoom to get results that resemble RL, but then the FOV, becomes pretty narrow and impacts situational awareness. Even if i adjust my seat to move closer to the monitor, at 50 centimeters, then I still get about half the image size I should be seeing compared to reality.

 

You can test the pictures yourselves in your own monitor to see how you do on your rig. My guess is that one would need about a 40 inch monitor to get close to real life. It would also be interesting to hear real pilot's opinion on these 3 pictures. I think that it seems more reasonable to assume that at 300 meters, one would look at a P51 and make out the details I see at the 2nd and maybe (for 20/20 vision or higher) 3rd picture.

 

I realise that it is hard for the developers to find a solution that fits everyone, but currently the way the engine works seems to make it excessively hard for people with 22-27 inch monitors to spot targets; hence the need for some kind of scaling.

 

And this test focused just on object size and not contrast which many members of the community have pinpointed as an area that needs to be worked for better spotting.

 

PS : this was done pretty quick, so if you find a problem with the math, please let me know. :)

Edited August 19, 2014 by airdoc

[kazereal]

PS : this was done pretty quick, so if you find a problem with the math, please let me know. :)

 

I *may* be wrong here so correct me if I'm not :).

 

According to Wikipedia human FOV (horizontal) in real world is nearly 180 degrees.

(Vertical is around 135 degrees.)

 

What is the FOV in your screenshots?

 

To me it seems that the image from DCS is "compressed" to render wider area to narrower screen to fit what you would see actually there.

 

This would mean that the angle in point C is actually different from point A (not a linear projection), right?

 

So we need to change FOV in DCS to wider angle to fix length of P-51 in your example case, right?

 

Edit: actually that would worsen it since more area would be "squashed" into narrow screen..

Edited August 19, 2014 by kazereal

[Wolf Rider]

@ airdoc...

wow... haven't seen that for a while :music_whistling:

 

the problem isn't your math, it is in the application and the fact that you need to be looking at this from the other side of photography - that being flat screen projection, of a 3d world and FoV distortions (distortion when altered from the 'designed for' default FoV). The distortions are 'pincushion' and 'barrel'.. or if you like, concave and convex.

 

 

iEdit - from your screenies, that's a very dark cockpit you're running... is your monitor contrast set correctly?

Edited August 19, 2014 by Wolf Rider

[airdoc]

@ airdoc...

wow... haven't seen that for a while :music_whistling:

 

the problem isn't your math, it is in the application and the fact that you need to be looking at this from the other side of photography - that being flat screen projection, of a 3d world and FoV distortions (distortion when altered from the 'designed for' default FoV). The distortions are 'pincushion' and 'barrel'.. or if you like, concave and convex.

 

 

iEdit - from your screenies, that's a very dark cockpit you're running... is your monitor contrast set correctly?

 

i realise that getting somewhat technical may make some people a bit weary. As Bucic already said, a lot of stuff posted is just speculation, and i tried to provide some facts about apparent size in DCS. Unless there is some mistake in the math, then, DCS draws the aircraft at a size close to the real one only for monitors above 32 inches (with standard zoom). I agree with you that there are other parameters involved but some, like the one i mentioned can actually be measured and are very important.

 

My 2 cents having served as an airforce academy physician : most modern military sims take the notion of apparent size into account. To do this, the projectors (or monitors) are of a standard size, with a standard FOV, and a standard location for the pilot's seat. In sim tests, even though pilots are looking at a simulated 3d world through 2d screens, the size of the targets is roughly the same as it would be in reality. Other factors such as motion blur, resolution, luminance, etc are also very important - but the apparent size should be standardised.

 

I 'm not arguing that DCS can go to the level of a military sim, i m just saying that it is optimised for very large monitors as far as apparent size is concerned.

 

P.S. : i m using sweet fx, which darkens a bit the pit in the P51.

 

cheers

[airdoc]

I *may* be wrong here so correct me if I'm not :).

 

According to Wikipedia human FOV (horizontal) in real world is nearly 180 degrees.

(Vertical is around 135 degrees.)

 

What is the FOV in your screenshots?

 

To me it seems that the image from DCS is "compressed" to render wider area to narrower screen to fit what you would see actually there.

 

This would mean that the angle in point C is actually different from point A (not a linear projection), right?

 

So we need to change FOV in DCS to wider angle to fix length of P-51 in your example case, right?

 

Edit: actually that would worsen it since more area would be "squashed" into narrow screen..

 

I 'm using standard FOV in the 1st pic, and zoomed in at the next 2 until i got the mentioned dimensions. Sorry don't know how to estimate the FOV in each one.

 

The monitor position is a bit confusing in the pic. At 1m i drew the small triangle just to measure the apparent size (BD) for this distance, and then compare it to the dimensions of the target at a monitor placed at the hypothetical 1m (and 50 cm). The real FOV (outer base) is not in direct linear continuum with the monitor's FOV, but this does not affect the estimation of the apparent size and the point made in the post.

 

What it's essentially saying, is that, if one sees a P51 at 300 meters in reality, then it's apparent size at 1m distance from the eye should be 3.2 cm. If one sits behind a monitor at 1m and that monitor draws a P51, then it should be 3.2 cm length if it is to match the real one. If in your set up you measure it it be i.e. 1.6 cm, then you get the apparent size of a real P51 at 600m, etc..

Edited August 20, 2014 by airdoc

[Wolf Rider]

 

 

My 2 cents having served as an airforce academy physician : most modern military sims take the notion of apparent size into account. To do this, the projectors (or monitors) are of a standard size, with a standard FOV, and a standard location for the pilot's seat. In sim tests, even though pilots are looking at a simulated 3d world through 2d screens, the size of the targets is roughly the same as it would be in reality. Other factors such as motion blur, resolution, luminance, etc are also very important - but the apparent size should be standardised.

 

I 'm not arguing that DCS can go to the level of a military sim, i m just saying that it is optimised for very large monitors as far as apparent size is concerned.

 

P.S. : i m using sweet fx, which darkens a bit the pit in the P51.

 

cheers

 

iirc, the DCS: A-10C was designed for the military...

 

 

just out of curiosity, what do you suggest the size of the "very large monitor" is?

Edited August 20, 2014 by Wolf Rider

[airdoc]

iirc, the DCS: A-10C was designed for the military...

 

 

just out of curiosity, what do you suggest the size of the "very large monitor" is?

 

i don't know details about the A-10C, since my country did not have it, but i presume that the military A-10C sim used for pilot training included the software with a replica of the cockpit and muliple monitor setup.

 

i 'd say anything over 32 inches is very large :)

 

would be interesting to see a poll about monitor sizes in sim users though.

[Wolf Rider]

i don't know details about the A-10C, since my country did not have it, but i presume that the military A-10C sim used for pilot training included the software with a replica of the cockpit and muliple monitor setup.

 

i 'd say anything over 32 inches is very large :)

 

would be interesting to see a poll about monitor sizes in sim users though.

 

 

32" monitors weren't around all that time ago when the current sim engine was developed though

 

 

yes, that's good thought there... a survey for user's monitor specs

[SharpeXB]

The problem isn't just monitor size, it's resolution. You can put your nose right up close to a 1080p monitor and it's still only 1080p. That resolution represents such poor vision you'd flunk fighter pilot school. Personally I look forward to 4K more than Oculus Rift.