|
|
#41
10-27-2011, 11:14 AM
|
|||
|
|||
Quote:
If you are afraid abt the low range value (1.6 +-0.1 NM), keep in mind that speed greatly affect detection range. The merge speed here being probably over 1000knots 
|
|
#42
10-27-2011, 11:34 AM
|
||||
|
||||
|
Tomcat Insuber is right; for the sight range see the documents that we found about it wrote by the U.S. Navy.
As simple example i would remember you that a runway of an international airport is from 1,6 nautical miles (3000 m) and 2,2 nautical miles (4000 m). If for seeing a contact with for example the sun behind your back you need to barely enter in formation with him we are really unlucky without radar.. About the contact visualization of moving objects it is simpler, cause the human brain (and of the animal in general) is made for enlight the moving objects. But i need to say that the image that you posted is really interesting because add to the discussion something that we lacked: info about the contact seeing in one of the worst situations, the one with the sun behind the contact.
|
|
#43
10-27-2011, 12:02 PM
|
|||
|
|||
|
Using the document let’s see the other factors who take part in the sim's "detection" algorithm, starting for the function of the curve above.
1)  In simulation: in CloD there should already be a model for these atmospheric/weather features, haze is reproduced on our screen, clouds are not there yet but they are working on it (and I hope that the dynamic weather is calculated before the mission as an event map, having the priority on the CPU only in that moment... it would be crazy to simulate that in real time!... I think Falcon 4 does something similar)... any way if you can reproduce it on the screen it means that you have the position of the modeled effect and its values. Did airplanes disappear behind a cloud already? Anyway oil smoke, sand, fog HAVE to matter on the windshield who have to be a factor too! These values can be used to weigh the atmospheric variable on the visibility algorithm. 2)  Yesterday I was talking with our military pilot about that: in fact it's true that in tactical formation (3km between aircrafts) with a not uniform background (low altitude) it’s difficult for him to detect his wingman... this is because he and his wingman are on the same vector at the same speed and the eye doesn't catch that airplane because practically "it's not moving". If his wingman accelerates or changes vector (some degrees of pitch are enough) he can be found again. I've read in WW2 pilots' books that sometimes they've lost the enemy because of this. In simulation: we are already living with this problem since in game our eyes don't focus on moving objects (all we see is moving pixels)... so many time we are going to lose a contact and here the fisheye camera (70 FOV) is helping us a lot! Anyway what's the problem in simulating relative motion in the sim? We already have speed and vector of the planes... it's all in the sim: we need only to weigh this factor in the visibility algorithm as difference between the observer speed/vector and that he has around him. 3)  In my opinion every model of plane in the game (not every instance of that plane in the sky) needs a visibility array that includes pre-calculated contrast values based on its aspects. Example: 1. A new skin is applied to the 3D model (you change skin on the plane's setting page) 2. The algorithm starts and calculate the average color of the model, the average brightness at EVERY aspect (30° or 45° differences are enough I think). Sure it should be better to include not the average but more ratio values. 3. The application save that array in the system.  Build an abstract map like the one above (but with not aspect index) for the terrain... it’s a big matrix and having the full map you can easily create an algorithm to calculate this value... more or less detailed (USE THAT RAM!)... Use these in game to see the contrast and you have a simulated visibility that doesn't base itself on "pixels" and better camos are now working (sure not at 100% as camos are intended but sure better than now). It's better if CloD allows the player to use his custom skin only if it's been supervises by 1C itself... like RoF does. No more BS skins please. In this case the pre-calculated array can be stored in the server and no more recalculation is need at every change of skin. 4)  In simulation: do we have light sources? I think it's enough... no more black dots looking in the direction of the sun: "Beware The Hun In The Sun!", even an average combat simulator should take care of that. As the document about camouflage says, light and reflections is a important factor on target visibility (as TomcatViP says too above, thanks for the addition!) and they need to be variables. As above if every plane has an array of surface reflection by aspect we are near to the solution. Light and its color its a multipler that puts a strain on the "contrast" value of that aspect... 6)  In simulation: let’s talk about vapor trails... I hope that they don't disappear suddenly like in the old IL2... did you see them? So there is a plane on its apex… 7)  In simulation: I’ll put it on the atmospheric model… Sure there are not real numbers for these factors. It’s all approximation and a tweaking matter. It’s not that we need target visibility to pass from 20% to 100% fidelity with reality. Like the realistic bombing/torpedo Mod of IL2 1946: we were at 40% with the stock 1946 and thanks to the mod we passed to the 80%... some were claiming that this feature was not complete (fixed 2 seconds for the activation), sure, but it was still more realistic than the stock one! We need to see on our monitor what the average fighter pilots see (one day maybe we could also set his visibility skill fatigue ect…) : we are not supposed to hunt pixels. IMO at the moment CloD is only a collection of beautiful pictures (Oleg is a photographer)... the sounds are coming, but this is intended as flight sim and COMBAT sim. The procedures that you are forced to follow are still not many. CEM is not a great thing neither: it can be difficult to be managed by the ones who come from IL2, not by players used to racing simulators for example. - If I want to see amazing pictures I watch at a Discovery Channel documentary at 1080p - If I want to have a great flight simulator I fly with Condor - If I want to do long and realistic procedures I fly with DCS A10 or Falcon 4 - If I want an air combat simulator I'll play with both games above and probably IL2 1946 (I want to be clear here: my love for it started to decrease the day I bought a 24" monitor and ultra detailed maps were released) I’ll post something that I think could be a valid solution (without taking account of resolutions, setting configuration and player's own eyes). It will be with labels BUT I assure you they will not be invasive and will keep a nice flight immersion.
__________________
 A whole generation of pilots learned to treasure the Spitfire for its delightful response to aerobatic manoeuvres and its handiness as a dogfighter. Iit is odd that they had continued to esteem these qualities over those of other fighters in spite of the fact that they were of only secondary importance tactically.Thus it is doubly ironic that the Spitfire’s reputation would habitually be established by reference to archaic, non-tactical criteria. Last edited by 6S.Manu; 10-27-2011 at 12:50 PM.
|
|
#44
10-27-2011, 12:33 PM
|
|||
|
|||
|
Quote:
 Regarding human brain. Detection of mvmt Ok but tracking no. WHat I mean is that for a positive identification you need to hve your eyeball locked on the moving object and this is more difficult when the speed is increased. For example, fighter pilots flying fast jet at low alt experience a tunnel vision. It the same phenomena but inversed. So IMHO it wld be translated that way in the sim: once target in POV, plane image is blurred to the point that we see only that there is "something" there. Then the intensity of the blur decrease progressively to a neat image depending on target range and time on focus (the time the pilot head look in that specific direction) It's a bit different to what Manu excellently said but I am more focused on the merge case
Last edited by TomcatViP; 10-27-2011 at 12:36 PM.
|
|
#45
10-27-2011, 12:42 PM
|
|||
|
|||
|
guys,
your disquisition is cleverly conducted with high technical competence, you have my cheering and regard,..but my question is...would be relistic in CloD and contemporary flight sims reproduce a convincing target spotting considering that a contact, spotted 5 miles away, (considering at the common resolutions we actually play) is rendered by no more than 4-5 pixels? I mean, if you consider 4-5 pixels (but even with 15-20) with all the permutations they can be arranged in and color gradients they can assume, can be possible differentiate a me109 from a spit or mustang or F16; and reproduce all of them in all brightness conditions? I think the res is the limit. Cheers
|
|
#46
10-27-2011, 12:56 PM
|
|||
|
|||
|
Quote:
The solution is a background function that give you informations about that the virtual pilot see even if you (the player) are not able to see anything in your monitor (in reality you can not see the 3D model or its lod if you don't detect it). We need to show to the player those informations: it has not to be something arcade as tha hated IL2's F6 key or full time labels.
__________________
A whole generation of pilots learned to treasure the Spitfire for its delightful response to aerobatic manoeuvres and its handiness as a dogfighter. Iit is odd that they had continued to esteem these qualities over those of other fighters in spite of the fact that they were of only secondary importance tactically.Thus it is doubly ironic that the Spitfire’s reputation would habitually be established by reference to archaic, non-tactical criteria. Last edited by 6S.Manu; 10-27-2011 at 02:43 PM.
|
|
#47
10-27-2011, 01:26 PM
|
|||
|
|||
|
For me would be satisfactory a linear scaling of LODs, with no holes approaching, and ensure the visibility of the target by a clever usage of contrasts (balancing the contrast of the background and A/C based on color/camo and brightness).
I consider natural to sniff the contact to ascertain the nature of it before consider a reaction; when my opponent is Human, whatever is the condition of advantage or disadvantage( in terms of tactical position or visibility), I'm assured by the fact we both have equal footing. Against AI, and this make me crazy in all sims, we are eternally in disadvantage, AI knows all of us since 15 miles away.GRRRRR.... and it never lost sight of us (behind a cloud...ops.. in CloD clouds are not modeled yet eh eh..another thread...or sun blinded) Ciao, thanks
|
|
#48
10-27-2011, 01:28 PM
|
||||
|
||||
|
pupaxx indeed that is the problem: the limit of the rendering-monitor and the real life.
So there are people that want not a real life sight system because an icon system is hurting their feelings. I understand perfectly that the icons are terrible (at least as they are in Il2 and Cod), invasive, immersion killing etc etc, but i felt also with manu that there were things not realistic in the way that the contact were simulated. But for beeing sure and stop all the "i think that you are wrong" we searched for documents wrote from reliable sources. The collecting is still in going and the scheme about the f-16 contact seeing is interesting ( where did u take it?). At the end i think that, once a reasonable and verified data collection is made, we should test out the results on Cod and, if the results as it seems now are different from the reality ones, start to think about how to solve that problem in a way that is affordable from the community and implementable from the programmers. Obviously nothing is assuring us that the programmers will not throw everything off as junk or simply not reading at all about it, but our goal is to have a correct simulator to fly and to enjoy, without pretending to be aces because we see contacts 100 km away or to be blind because we see barely at 3 Km.. Last edited by 6S.Tamat; 10-27-2011 at 02:36 PM.
|
|
#49
10-27-2011, 02:06 PM
|
|||
|
|||
|
BTW with the new settings, in ATAG the dots are now visible (sometimes) from far away, but they still tend to disappear suddenly.
Farther dots are grayish, a nice effect imho, and closer dots are darker. Maybe as a stopgap measure it's just enough to improve and debug the present dot system (contrast, visibility at all distances, LOD) as pupaxx was suggesting, and adding the radio vectoring system for historical and playability reasons. Then the long term solution could be further evaluated, but with a more playable (and sellable) game. Cheers! Last edited by Insuber; 10-27-2011 at 02:09 PM.
|
|
#50
10-27-2011, 03:19 PM
|
|||
|
|||
|
rHey guys, I work as an airline pilot and I read this post before I went to work this morning and thought that I will try this out today while on cruise altitude. We have a so called Multifunction display on the flight deck that shows all the traffic around you as you fly.
This is my conclusion (My eyesight is perfect... so far...) At 5nm (9.2km) I can usually easily spot an AC and i can even tell what type is it is. However at thi distance its very difficult to tell if is a 737 or an Airbus as the share similiar shapes. Smaller planes are very difficult to ID at this distance. You can usually tell, if its a Kingair (due to its T-tail) or smaller, but to know exactly is almost impossible. Remember that a 737 or Airbus are alot bigger than the planes we talk about in game!! Also these planes are white against a blue or green background and that makes them alot easier to spot! many times someone is passing you, according to your instruments, at say 15km and same altitude and both of us try hard to see him but in some light conditions u see noone even though u know he is there somewhere! Obviosuly at night its alot easier with the strobes and beacons flashing. Im not so sure i would be able to even see a 109 at 10km or maybe even 6-7km and I certainly would not be able to ID such as small plane untill ALOT closer!
|
 |
|
|
Linear Mode
