Improved FLIR Simulation

[ALT326]

D'Oh! All the time I was thinking this must have been reported a thousand times already, but I couldn't find any mention of it in the Bugs and Problems forum, so here we go: A-10C TGP Vehicles Almost Invisible.

 

Guys, if you have screenshots, videos or other evidence of invisible vehicles or other objects, please append them to the bug report to further illustrate the issue. I made a quick test mission, but I know there are much stronger examples out there. :thumbup:

 

I posted this one in the bugs list of DCS 1.5 and it has been moved to this section. It only proves to me that DCS/ED do not see this problem as a bug and will not try to solve this.

[Baxter]

Does anyone know how the "effect" works for the TGP? Could someone make hot white or hot black textures so they appear much "brighter" in the TGP? Like a replacement texture specifically for a vehicles heat map

[baltic_dragon]

I posted this one in the bugs list of DCS 1.5 and it has been moved to this section. It only proves to me that DCS/ED do not see this problem as a bug and will not try to solve this.

 

It is still in bugs section for A-10C. See here.

[Aeger]

Can we get some official response? Maybe a bug tracking number?

[Ganesh]

Selling a fabolous DLC and drum up buisness to sell her fantastic A-10C again but no response on this big issue :mad:

Sometimes ED is making me really sad :(

[Aeger]

I have a feeling their plate is really full right now (1.5, 2.0, F-18?) and perhaps they're on vacation. But yeah, critical problem for a critical component of the A-10C module...

 

Someone please acknowledge :cry:

[Ganesh]

It is not that they should fix it like a miracle.

But a word to us customers like: "We care about" would more than enough for many of us.

[baltic_dragon]

I have just flown in 2.0 and FLIR works beautifully in Nevada.

[electro_monkey]

I have just flown in 2.0 and FLIR works beautifully in Nevada.

 

It's the way better than in 1.5 but still trees are too hot

[scheffchen]

I have just flown in 2.0 and FLIR works beautifully in Nevada.

 

Not true. And you know it. Your campaign would also benefit from a working flir, so please reconsider your post..

[baltic_dragon]

Not true. And you know it. Your campaign would also benefit from a working flir, so please reconsider your post..

 

You misunderstood - worked well as opposed to Georgia. If the 1.5 one worked like 2.0 it would have been great. But id doesn't.

[sobek]

You have to understand how programmers think.

 

If it works as intended by the programmer, then it is technically not bugged.

 

In this case, for simplicity and probably implementation costs ED chose to implement FLIR as a color filter. That is (in simplified terms), the shade of green that most ground vehicles sport is mapped to white, everything that is not that shade of green is mapped to black (obviously with transition in between and the ability to switch polarity for BHOT).

 

Everything that has the same color as most ground vehicles will appear hot for that matter. That in itself is not a bug, but a side effect of the simplified design.

 

To truly emulate FLIR, one would have to retexture every 3d object in the game for the infrared spectrum. You can imagine what a herculean task that is.

 

The next chance where this *might* happen is possibly with the F-18, as it also features FLIR. Looking at the A-10s product lifecycle, it is simply too much to ask from ED to squeeze so much development effort into a product that is now on "extended support". The F-18 on the other hand is a new product and as such, has much more potential to actually recoup such a development effort.

 

To cut to the chase: Don't expect ED to retrofit a better FLIR render method inside the A-10C project. If you want to campaign, ask them to develop it for the F-18 and retrofit it to the A-10. That has at least a nonzero chance of success.

[sobek]

Sobek,

 

I would be willing to help with reskinning the vehicles if needed.

 

Please contact Sith or Wags, i'm no longer in the loop.

[Flagrum]

You have to understand how programmers think.

 

If it works as intended by the programmer, then it is technically not bugged.

From a technical standpoint, nothing has changed in the software and it works as before, i.e. "as intended". But the aircraft simulation is not about how the software works, but what it does. And that is to simulate an IR sensor.

 

Granted, there are several levels of realism possible, but whatever we had and still have, before, one could see targets with that sensor and now it is not possible. The sensor simply does not do what it should do.

[Cherrug]

As Baltic Dragon mentioned earlier, with Nevada map FLIR works with no problem. (I can confirm that too).The Nevada map is part of DCS World 2.0 which uses terrain engine T4, as Matt said earlier. Maybe that's why there is no official response or intention to solve the issue under DCS 1.5. and Black Sea (which is terrain engine T3). ED now works on converting the Black Sea map to T4, and when it's done (we hope it's gonna be soon) the issue or problem will vanish.

[Flagrum]

Yeah sure ... but that is only an assumption. And we were all told over and over again to not assume things in DCS ...

 

(an semi-official one-liner as reply in the respective threads would have been much appreciated. instead we are left to ... well, assume again of why those landed in the wishlists)

[electro_monkey]

The Nevada map is part of DCS World 2.0 which uses terrain engine T4, as Matt said earlier. Maybe that's why there is no official response or intention to solve the issue under DCS 1.5. and Black Sea (which is terrain engine T3). ED now works on converting the Black Sea map to T4, and when it's done (we hope it's gonna be soon) the issue or problem will vanish.

 

Unfortunately it looks like that a new terrain engine won't solve anything, target visibility in Nevada is so good just because of ground texture color. Take a look at FLIR in 1.5 with white ground texture:

[sobek]

Granted, there are several levels of realism possible, but whatever we had and still have, before, one could see targets with that sensor and now it is not possible. The sensor simply does not do what it should do.

 

Yes, the design is somewhat flawed to begin with.

 

I'd say if something has changed when moving the filter from the old to the new engine which impedes it's function then that is indeed a bug and does not belong in a whishlist.

[sobek]

Unfortunately it looks like that a new terrain engine won't solve anything, target visibility in Nevada is so good just because of ground texture color. Take a look at FLIR in 1.5 with white ground texture:

 

I agree, this is just a matter of the color gradient.

 

If vehicles ever get a desert camo, this implementation is in for some serious trouble. But ED don't need us to tell them that.

Edited December 28, 2015 by sobek

[Yurgon]

I guess I found a workaround, more or less by accident.

 

Applying the workaround:

 

1. Check your visibility range setting: Options -> System -> VISIB RANGE

 

2. Edit the appropriate Lua file in \Config\terrain using Notepad++:

 

C:\Program Files\Eagle Dynamics\DCS World OpenBeta\Config\terrain\{WHATEVER}.lua

(1.5 OpenBeta default path)

C:\Program Files\Eagle Dynamics\DCS World\Config\terrain\{WHATEVER}.lua

(1.5 Public/Release version default path)

(Should be one of Low.lua, Medium.lua, High.lua, Ultra.lua or Extreme.lua)

 

3. Change the noisemax value to 0.5:

 

--Old Noise 
land_noise =
{
noisemin = 0.0;
[b]noisemax = 0.5;[/b]
noise1front = 1000.0; 
noise1back = 30000.0;
noise1top = 2500.0;
noise1bottom = 2000.0;
noise1PerSquare = 4.0;
noise2PerSquare = 150.0;
}

 

4. Launch DCS and happy hunting. :thumbup:

 

As far as I can tell, the terrain goes back to looking a little mushy, like it did in 1.2, which right now appears to be the price of getting the TGP back to a certain degree of usefulness.

 

Feel free to experiment with different values and maybe try other properties as well. Please post back if you get better results! :thumbup:

 

Oddly enough, 2.0 uses the exact same values, so I guess this workaround depends on the noise-values as much as it depends on the actual terrain textures.

 

Comparison screenshots (downscaled):

 

 

 

 

Finally, a word towards ED: The TGP worked okay in 1.2. It was almost entirely broken in 1.5. This is not a wishlist item, this is a bug.

Edited January 3, 2016 by Yurgon
Adjusted paths for 1.5 Release Version

[dotChuckles]

This has been known and complained about ever since A-10C came out, and how long ago is that now?

 

It really does need looked at and a new solution found. If the whole of the Black sea map can be retextured (admittedly with third party help) why can't the vehicles be retextured in the same way to make a proper thermal image work?

 

The current implementation feels cheap compared to the high standard of other parts of the sim. It just doesn't make sense to me that a primary sensor should be so badly simulated.

 

Sorry if that sounds harsh. It just bugs me.

[Bob Denny]

For full fidelity, the buildings would have to be skin separately for thermal imaging as well. This would really be great.

[Yurgon]

This has been known and complained about ever since A-10C came out, and how long ago is that now?

 

I guess the merging of threads didn't help.

 

Most of these posts don't request a TGP that's better than it was in "the old days" (standalone, 1.2, you name it). Most of these posts request a TGP that works just as good as it did.

 

I don't know if you've flown the A-10C in 1.5, or if you've looked at some of the screenshots, but under some circumstances, vehicles and units have actually become invisible while they're right in the middle of the TGP. So the complaint is not "This is unrealistic", the complaint is "It's almost to impossible to finish some missions that worked okay-ish before".

 

And while I do seem to recall a sort of TGP degradation that came with a previous DCS update, that was nowhere near as bad as with 1.5.

 

Of course I wouldn't mind if the solution was to improve the TGP altogether, but for the time being a fix like my workaround from post #70 is pretty much all I'm asking for.

[sobek]

[...]

 

Sounds perfectly reasonable. If you turn up the noise, more parts of the terrain will intersect with the filter "criterion" leading to more noise in the TGP image.

[scheffchen]

 

Most of these posts don't request a TGP that's better than it was in "the old days" (standalone, 1.2, you name it). Most of these posts request a TGP that works just as good as it did.

 

 

You are wrong and this is not true. Most of the ppl. want what´s necessary for a SIMULATION- a truely simulated flir.

 

Since this is a flight simulator with realistic flight model, weapons, radar, navigation, and weather. Why accept a cheap trade-off for a very very very core element of modern warfare: Heat signature based air to ground resp. ground to ground combat. Re-think of it. I would gladely pay for a dcs-world 3 or a-10c version 2 with enhanced imaging! (As ppl. already did with blackshark and flaming cliffs - remember?).