investigating ALL IR guided missiles FLARE rejection inconsistency

[GumidekCZ]

I would like to report very long known bug in DCS about Flare rejection system of IR seekers in DCS.

Now the flare rejection is more than anything else dependant od distance/time to reach target, rather than ccm koeficient defined for each IR missile in Lua - together with amount of flares ejected by defedning aircraft (now after new patch, also with FLARE type (bigger/smaller)).

I did attached 5 files, 2 missions: FLARE_TEST.mizFLARE_TEST_9X.miz  - everybody can try to use flares as shown in my 2 tracks (one with Early FLAREs and one with Late FLAREs -dumped all I had): _IR_BUG_EARLY_FLARE.trk   _IR_BUG_LATE_FLARE.trk - both flown on IDLE power!

LATE track shows, how only single pair of FLAREs or two can confuse IR missile just after its release... even the mighty X!

5th file PL-5EII_always_first_to_shoot.acmi - shows, that DEKAs Chinese PL-5EII is always first to be released - thats very strange, I would expect to be launched as a last one - but this could be topic for another bug report.

MiG-29 dumped all 30 flares to finaly end up in fire, same as Hornet with all 80 2x (5F x 0,25s x8) small flares - both hit by AIM-9M at that time. My next try, it was R73/R550/PL5E.. so it doesnt matter what missile it was.

 

Edited May 27, 2021 by GumidekCZ

[GumidekCZ]

Bumping up, to be more visible to ED devs and Admins. Hopefully, they will notice some day, and someone will take look into it.

[Cmptohocah]

On 5/27/2021 at 2:36 PM, GumidekCZ said:

... both flown on IDLE power!

 

 

 

Unless something changed recently, there is no diff. in IR signature in DCS between 0% and 100% power setting (no AB).

[GumidekCZ]

with connection to FLARE rejection, I found interesting scientific paper about small and big flare effectivness released from probably idle AMX jet and C-130. On top of that, the facts mentioned in that study just confirms the above mentioned bug.

https://www.reddit.com/r/hoggit/comments/rwn11i/effectivnes_of_big_and_small_flares_in_scientific/?utm_source=share&utm_medium=web2x&context=3

Seems, that the effectivness of small FLARE as is in DCS (according to tests around 50% of the bigger brother) is unrated.

Also according to papers from manufacturer the big one Peak IR output: 20kW/sr nominal and the small one has Peak IR output: 12kW/sr nominal - 60% of the big one.

Big: 51.82mm x 24.64mm x 200 mm

Small: 24.64mm x 24.64mm x 206 mm

Edited January 5, 2022 by GumidekCZ

[GumidekCZ]

DCS Flare rejection simulation is still BIG FAIL to me. Im glad that I can hear birds singing when standing be a tree on ground, but the simulation of DCS countermeasures - by which I mean all three of them Flares , Chaffs, ECM ... it is on level of other fighter planes simulations, may be even worse. For example the simulation here is so simple, that it even doesnt matter if your engine are at IDLE or MIL(Buster).

In attached tracks I used all of 80 Flares in one single program to decoy IR guided missile ( I was out of burner and pulling into a missile or close to a beam aspect to create high angular speed to missle seeker ), but without success. I tried more variations of released Flare time sequences, but it didnt have any noticable effect.

null

80-FLARES_not_enough_1.trk 80-FLARES_not_enough_4.trk 80-FLARES_not_enough_3.trk 80-FLARES_not_enough_2.trk

[Raviar]

I ran multiple test in Tiger, Viper and Hornet against multiple SAMs and Missile, However seems the flare is not effective at all, while preamptive flares are effective, the flare is not much effective once missile is tracking , the attached TacView shows R-60 can get kill while the blue is moving against the sun direction while the trottle is ilde and releasing as much as flare as possible 

On the other hand the A-10 can easily dogde the missile by flaring, the size and the amount of flare and heat signature is matter, no doubt on it, BUT my guess is the whole DCS logic and architecture is kind of base on A-10, the missiles, logics ... didnt recieve enough update while ED continusely is launching new modules! however I might 100% wong !

The topic is not the skills and how to dodge, I know I need to dive towards the ground with high G manuver, but my concerns are Flare in DCS vs real life
 

as most of you probably know the 9X deafted by the flare! (please dont give me the flare size/type thing)

https://www.popularmechanics.com/military/aviation/news/a27094/su-22-dodge-aim-9x-sidewinder/

Tacview-20220907-173131-DCS.zip.acmi Tacview-20220907-172615-DCS.zip.acmi Tacview-20220907-172505-DCS.zip.acmi Tacview-20220907-174340-DCS.zip.acmi Tacview-20220907-174302-DCS.zip.acmi Tacview-20220907-174340-DCS.zip.acmi Tacview-20220907-174302-DCS.zip.acmi

Edited September 7, 2022 by Raviar

[Raviar]

I have the same issue

This topic needs way more attention

@BIGNEWY please help us   

[BIGNEWY]

Posts merged

On 9/3/2022 at 9:15 AM, GumidekCZ said:

DCS Flare rejection simulation is still BIG FAIL to me. Im glad that I can hear birds singing when standing be a tree on ground, but the simulation of DCS countermeasures - by which I mean all three of them Flares , Chaffs, ECM ... it is on level of other fighter planes simulations, may be even worse. For example the simulation here is so simple, that it even doesnt matter if your engine are at IDLE or MIL(Buster).

In attached tracks I used all of 80 Flares in one single program to decoy IR guided missile ( I was out of burner and pulling into a missile or close to a beam aspect to create high angular speed to missle seeker ), but without success. I tried more variations of released Flare time sequences, but it didnt have any noticable effect.

null

80-FLARES_not_enough_1.trk 162.25 kB · 2 downloads 80-FLARES_not_enough_4.trk 144.26 kB · 2 downloads 80-FLARES_not_enough_3.trk 229.57 kB · 2 downloads 80-FLARES_not_enough_2.trk 278.47 kB · 2 downloads

I will ask the team but I dont think there is any issue here, when deployed effectively they seem to be working, in some of your example a hit was inevitable no matter how many flares

@Raviar please include track replays and not just tacviews

 

thanks

 

[firimar]

Ok, flares sometimes just ignored by IR missiles, no matter how many you pop them and when and what maneuver you make.

But what about hills efficiency against IR missiles? =D

 

[GumidekCZ]

May be some of you already noticed my other report about FLAREs and IR seekers, ingnorign FLARE untill the missile is launched:

@BIGNEWY  You said: "when deployed effectively", what I was out of burners all the time, I poped 80 Flares and turned hard into missile to increase aspect and hide my engine exhausts. I could have done different evasive maneuver, but the result would be almost the same.

I know that FLARE program, release timing, maneuver is matter of IR guidance sensor performance (generation of IR seeker with it technology) but even me, Im not expecting to be modeled in DCS any time in future of DCS sim. But just simple airplane aspect, IR emission coefs for Gate/Mil engine setting, size of flare - all packed up by just some dice rolling chance of success - I thing that sim game like DCS deserve more than that.

[Exorcet]

Heavy flare use doesn't guarantee decoying a missile:

 

And while DCS does use a simple chance formula for decoys, the method itself isn't terrible. It should mean that more flares = better chance at decoying. It could probably be improved by being made a bit more dynamic. For example the decoy chance increasing with number of flares in the missile seeker FoV or decreasing with age of the flare/distance from the deploying fighter.

As for the plane heat model, AB on/off is probably more significant than throttle percent in mil:

 

[GumidekCZ]

36 minutes ago, Exorcet said:

Heavy flare use doesn't guarantee decoying a missile

It depends on seeker design and processing unit software (generation of IR missile), or if flares are just very old ones or more recent multispectral or even more recent with small fins and own propulsion.

On the IR vid from show, there is sadly no plane with engine at idle to compare with MIL.

[firimar]

Can't even imagine how those things are coded in DCS backend. Is it just some random generator, separate for each client? =D

Took 2 trackview files from same multiplayer mission and found that same missiles fly different paths for different players, they go to different flares, or not going at all, that's for the same missile. Example, this is same missile launch seen from different players (and, btw, helicopter was destroyed by this missile):

Edited September 19, 2022 by firimar

[okopanja]

24 minutes ago, firimar said:

Can't even imagine how those things are coded in DCS backend. Is it just some random generator, separate for each client? =D

Took 2 trackview files from same multiplayer mission and found that same missiles fly different paths for different players, they go to different flares, or not going at all, that's for the same missile. Example, this is same missile launch seen from different players (and, btw, helicopter was destroyed by this missile):

Yes, you are not imagining it. I fired once 77 then ET. On my PC 77 hit him, on server ET hit him, while 77 passed near cockpit. Shooter's side decide the outcome.

 

[Raviar]

On 9/16/2022 at 2:46 PM, Exorcet said:

Heavy flare use doesn't guarantee decoying a missile:

 

And while DCS does use a simple chance formula for decoys, the method itself isn't terrible. It should mean that more flares = better chance at decoying. It could probably be improved by being made a bit more dynamic. For example the decoy chance increasing with number of flares in the missile seeker FoV or decreasing with age of the flare/distance from the deploying fighter.

As for the plane heat model, AB on/off is probably more significant than throttle percent in mil:

 

Heavy!
1st you are talking about specific varient and type of AIM-9 and the most advacned one which is 9X!
2nd regular Flare suppose to do that! I compared R-60! not even M 

https://www.popularmechanics.com/military/aviation/news/a27094/su-22-dodge-aim-9x-sidewinder/

If the Su-22 Flare can confuse the AIM-9X! same flares should confuse 9J/P5/9M/R-60/R-60M/R-73 !

[Exorcet]

13 minutes ago, Raviar said:

Heavy!
1st you are talking about specific varient and type of AIM-9 and the most advacned one which is 9X!

Yes I realize that, but the point stands. Just dumping a large number of flares doesn't mean a missile will be decoyed. It may or may not, it depends on a number of factors.

13 minutes ago, Raviar said:

2nd regular Flare suppose to do that! I compared R-60! not even M 

I'm not really sure what you tried to say here.

13 minutes ago, Raviar said:

https://www.popularmechanics.com/military/aviation/news/a27094/su-22-dodge-aim-9x-sidewinder/

If the Su-22 Flare can confuse the AIM-9X! same flares should confuse 9J/P5/9M/R-60/R-60M/R-73 !

This event is difficult to apply to the discussion. A single launch doesn't provide much data and it's not even clear that the flares had any effect on the missile. There are other reasons for a miss. Even if we want use it as a relevant example, what does it tell us about modeling the flare rejection of other missiles? Not very much. You can't extract rejection rates from an anecdote.

[Noctrach]

20 hours ago, Exorcet said:

Yes I realize that, but the point stands. Just dumping a large number of flares doesn't mean a missile will be decoyed. It may or may not, it depends on a number of factors.

I'm not really sure what you tried to say here.

This event is difficult to apply to the discussion. A single launch doesn't provide much data and it's not even clear that the flares had any effect on the missile. There are other reasons for a miss. Even if we want use it as a relevant example, what does it tell us about modeling the flare rejection of other missiles? Not very much. You can't extract rejection rates from an anecdote.

The point is that DCS doesn't model seeker type at all. It's pure RNG.

There's no spoofing the seeker lock-on, because the countermeasure simulation is only done on the missile. So only once the missile is off the rail you're left with % chance with ccm value, binary IR signature (two values, AB or not-AB) sun modifier and front/aft hemisphere modifiers, where small flares are just half as effective as the large ones. So you just need to dump twice as many to have the same statistical effect.

It doesn't take into account R-73 or AIM-9X having entirely different seeker technology than R-60 or AIM-9P, other than modifying the % a little.

Seeker decoying irl is reasonably deterministic, nothing works everytime but a particular seeker will get oversaturated with a particular flare release patterns for a particular jet with sufficiently reliable consistency. It's why the ECM-ECCM arms race exists in the way it does.

You really notice this especially if you run the comparison with real life flare techniques. Try any real world flare pattern in DCS and you will notice it is an absolutely pointless waste of countermeasures. A programmed flare every other second will have you get hit by Igla-S nearly every single time. A one-time release of 8 small flares or 4 big flares in quick succession will reduce this Pk to 10% or less. It makes self-protection flaring programs for ground strike mostly pointless compared to waiting for a missile to be fired and spamming the button once the missile is in the air. It's just a diceroll, more dice wins every time.

The RNG seeker is not a bad approach per se, but the practical effect is just where the realism of the simulation completely falls apart.

Edited September 20, 2022 by Noctrach

[firimar]

19.09.2022 в 12:01, okopanja сказал:

I fired once 77 then ET. On my PC 77 hit him, on server ET hit him, while 77 passed near cockpit. Shooter's side decide the outcome.

Okay, I see, done some googling and found many reports about such issues, some of them more than year ago.

Looks like we can't expect accurate combat simulation in Digital Combat Simulator then...