AIM-54 Hotfix PSA and Feedback Thread - Guided Discussion

[lunaticfringe]

5 minutes ago, Strider21 said:

Care to explain what exactly I am doing wrong here?

2/9 Firing at 30 NNM 35000' MACH 0.98. All 7 misses AMI-54 had plenty of energy and pull G but passed in front and above of the target. 

Tacview-20220216-144458-DCS-AIM54C Test .zip.acmi 1.03 MB · 0 downloads

Stop flying straight at the bandit with a radar missile out.  Whereas a crank with a semi-active radar missile is purely for controlling closure, with an active missile shot in TWS the cut forces their hand.  If they continue to press, they've got to turn at you, which aids your missile late- either he turns back and through the missile's flight path roughly 100 degrees or so to try and defend (and when it's the AI, especially "Ace"- that means absolutely perfect notching, which was happening consistently here, but the extra delay caused by maneuver you make is where the missile can often catch them based on timing), or he turns away from you to go for the notch angle, which gives you easy access to tail aspect and zero SA.  Compounding the problem for the bandit in some fashion always helps your current shot out, or the next one you take. 

Also, don't loft the missile yourself; let the Phoenix guidance do its thing.  It has work to be done, but forcibly pitching it further right now is messing with the lofting gain. 

[Strider21]

59 minutes ago, lunaticfringe said:

Stop flying straight at the bandit with a radar missile out.  Whereas a crank with a semi-active radar missile is purely for controlling closure, with an active missile shot in TWS the cut forces their hand.  If they continue to press, they've got to turn at you, which aids your missile late- either he turns back and through the missile's flight path roughly 100 degrees or so to try and defend (and when it's the AI, especially "Ace"- that means absolutely perfect notching, which was happening consistently here, but the extra delay caused by maneuver you make is where the missile can often catch them based on timing), or he turns away from you to go for the notch angle, which gives you easy access to tail aspect and zero SA.  Compounding the problem for the bandit in some fashion always helps your current shot out, or the next one you take. 

Also, don't loft the missile yourself; let the Phoenix guidance do its thing.  It has work to be done, but forcibly pitching it further right now is messing with the lofting gain. 

I should have mentioned these were shots were tests to specifically look at the performance. I didn't crank to limit the change that TWS would loose lock. Either way crank or not should not have changed the terminal homing which seems to be the issue in these shots. 

Likewise the first 3 shots were done in level flight, the next 6 were done with a 30 deg pitch. A friend had asked whether doing a 30 deg pitch improved performance. Based on the terminal energy being nearly the same when fired in level flight and with a pitch up I don't think it makes a difference. The level launch shots had the same issues as the 30 deg pitch up. That is the missile takes a huge lead and passes in front of the target. . If you look at the terminal maneuver by the AIM-54s that  they are pulling 15-18g with plenty of energy MACH 2+ but pass in front of the target. 

Edited February 18, 2022 by Strider21

[Clunk1001]

2 hours ago, Frostie said:

All well and good but I was trying to replicate your profile that you claimed R27 was superior, I had very little detail to work with such as altitude of bandits and your loadout so figured carrying 6 gave me the best results on seeing the phoenix at different regimes. I was very loose with missile shots.

 

I don't know how more I can try and help you understand what you need to do. Whenever something is highlighted you move the goal posts.

I put a perfectly fine tacview up earlier of 4 Su33 vs 1 F14 with 4*54,2*7M,2*9M. 

 

 

 

 

 

I'm not suggesting the R27 is superior (that really is my bad flying).  I was trying to focus on a consistent test with the Phoenix (regardless of whether I survive or not - I just wanted to test the Phoenix).  I was getting frustrated that at 30,000 I see a very high miss rate.   My frustrated (apologies if I've taken any out on forum members) is at dying after I've seen 8 Phoenix sail past a target....and that everyone seems to think this is normal.

[draconus]

25 minutes ago, Clunk1001 said:

My frustrated (apologies if I've taken any out on forum members) is at dying after I've seen 8 Phoenix sail past a target....and that everyone seems to think this is normal.

You have to understand that any A2A missile's Pk is higly dependant on the employment, situation and enemy actions. In any case you cannot expect them to hit. You look for the outcome of course but until you're sure you act like your missile did not connect - either keep pressing or defend when too close and outnumbered. The last part is about not dying.

When you keep doing the same test the same way and it doesn't work out well don't expect different results. That what I think is normal. Do it differently, include different situations and you will have different (and more real) view on the missile's Pk.

As for the guidance - this is what Mike keeps saying - it's not their job, but they keep working together with ED to make it (and new API) better. Don't act like it is final perfect version - it's not - good tests are still welcome.

Edited February 18, 2022 by draconus

[IronMike]

2 hours ago, Strider21 said:

Care to explain what exactly I am doing wrong here?

2/9 Firing at 30 NNM 35000' MACH 0.98. All 7 misses AMI-54 had plenty of energy and pull G but passed in front and above of the target. 

Tacview-20220216-144458-DCS-AIM54C Test .zip.acmi 1.03 MB · 6 downloads

They are being notched in their final lead turn. You can see how the AI does exactly the same maneuver over and over again, being omniscient as it is. For the Su-30, you need to fire either further out, or much closer in (mind the r77 though). I posted tacviews of how to deal with 2 of them in an earlier comment. If you do the head-on thing their very effective defense will simply be the same over and over again. So you need to make them react to something else upfront, which offsets this ultra-notch of theirs. Look at the tacviews below, they are the same ones I mention. You can see how in the long range shot and the Cmk47 shots they do it as well, and one out of 2 defends a missile. That's ok, he's been put defensive and now I can press and finish him. A missile does not need to always hit to be effective, the "game" that you play with your bandit, extends beyond that.

Let's do a very stupid example. FPS pvp, you advance on an enemy held house: you see a player popping his head out of a window, you aim, shoot, but just as you fire he ducks behind the wall. Your bullet misses. Do you blame the bullet for missing him? Or do you proceed to chuck a grenade through the window instead? Now, forgive me, it is a very stupid comparison I shall admit. But point remains: don't blame the bullet for your enemy taking cover. Don't blame the missile for your bandit notching it. Switch your game on, and take advantage of him getting his off of the offensive for these brief moments.

There are guidance issues remaining, but a missile being notched isn't really a guidance issue. It is a successful defense which you have to counter.
 

2xSu30_Amk60.acmi 2xSu30_Amk60_long.acmi 2xSu30_Cmk47.acmi

Edited February 18, 2022 by IronMike

[Clunk1001]

I appreciate my simple head-on test may not be 'realistic', and there are better ways to engage, but it's a test.

I did the same test at 30,000 and 40,000 against the 3 flankers. Launching at around 60nm

At 30,000 all 7 Phoenix missiles miss  (are avoided)   0% missile hit rate.   This is pretty much what I'm seeing consistently below 40,000  (because I fly below 40,000.....or rather I used to fly below 40,000!).   I know some of you seem to be seeing other results.

Same test at 40,000 all flankers destroyed,  42% missile hit rate.

The main difference seems to be the missile speed at reaching the target.  around M1.8 for the missiles fired at 30,000, and almost M2.8 for the missiles fired from 40,000.  (At top of curve, the difference again is about 1M).  There seems no difference in the AI reaction or notch timing to the missiles, simply that they cannot avoid the faster ones.

Now, I know what's coming....I know the air is thinner - I have a commercial pilots licence, I know about thinner air - my point still stands that the hit rate of the phoenix (miss rate, AI avoidance rate, whatever you want to call it) feels wrong below 40,000.  

Someone mentioned earlier about buffs/cheats, and to 'learn' the game -  popping up to Angels 40 because I know that the game will make my phoenix kill things if I'm just that few feet higher, feels like a bit of a cheat to me.  I would expect to see a drop off in performance below 40,000 consistent with air density (not just missile performance, but the results of that performance).   I just wouldn't expect to see 7 phoenix miss like this at 30,000.

@IronMike   I believe I can guess how high you are in those engagements because of the speed of your missiles  (have you tried that at 30,000?).  I do like your analogy about the guy ducking his head back into the window when shot at (I am a huge fan on analogies).   And my advice for that scenario is that you are obviously shooting from prone or kneeling, you should simply press x on the keyboard to make your player stand up when you shoot (being that few feet higher seems to kill them every time in the game for some reason ).    I think I'd agree with you that this doesn't look like a guidance issue.   

Thanks for all the great work btw, really appreciate it!  And really appreciate the help on here too.

 

Tacview-20220218-073921-DCS-Dogfight.zip.acmi Tacview-20220218-074233-DCS-Dogfight.zip.acmi

Edited February 18, 2022 by Clunk1001

[Clunk1001]

3 hours ago, lunaticfringe said:

Three veteran Su-33s loaded for bear and maneuvering/evasion capable (Fighter Sweep/Evasion on), one F-14A with four AIM-54 Mk60s.  Three for three. 

All three shots taken at or above Mach 1 and 30,000', each roughly a minute apart from ranges of 72, 57, and 40 miles.  Basic 30 degree crank turns to control closure, and their one return R-27ER shot was trashed by my third Phoenix catching the shooter by the difference in effective missile range.

It's how it's being used.  We've got folks shooting all out of parameters, from slow at SL to crazy round dumps at middle altitudes and below the Mach, and we've got people putting consistent steel on target by using even the most basic applicable technique. How does this conversation proceed when people don't even recognize the clues for why they're having trouble with the weapon and why it's not working for them is in the data they're presenting as evidence?  

Tacview-20220218-001928-DCS.zip.acmi 396.61 kB · 5 downloads

Hi,

Thanks for this.  

I have a question,  my 'parameters' seem the same as yours, but 30 seconds after launch your 54A is busy accelerating to M4.33 while at the same time my 54C is slowing down from it's top speed of M3.3. Same launch height, I'm actually launching faster than you  (I'm at M1.07, you launch at M1.03).  Missile speed at target is about 1M difference (and your's hit).  As I mentioned above, this speed at target seems to be an important factor when it comes to AI avoidance.  Is the speed difference here an A vs C thing?   Or is there  some update to missile performance which I'm missing?

Edited February 18, 2022 by Clunk1001

[draconus]

3 minutes ago, Clunk1001 said:

Is the speed difference here an A vs C thing?

Yes, he specifically said AIM-54 Mk60 - longer burn time means more range and higher top speed.

[Clunk1001]

1 hour ago, draconus said:

Yes, he specifically said AIM-54 Mk60 - longer burn time means more range and higher top speed.

So, going back to my simple tests, comparing the two variants....  

At 40,000 head on (same profile/test/launch parameters as lunaticfringe, and my previous test), both seem to perform about the same.

But at 30,000 head on, I'm finding the 54A will generally kill at least 2 of the 3 bandits and often all 3 bandits, whereas the 54C will generally miss all 3 bandits.   

I've seen a lot of tracks with people saying "look how easy it is to kill 3 flankers, it must be your tactics".   That's great - but they seem to be either using the A, or using the C at 40,000ft.   And that's not what I'm talking about - I can do the same with the A (thanks lunaticfringe for pointing that out to me) and I can do the same with the C at 40000ft.  Just like the rest of you.   My point is the C at 30,000 does not seem right (in my test scenario).  If the motor on the phoenix is modelled accurately then the guidance may be off, if the guidance is modelled correctly then the motor may be off   (all subject to the DCS AI of course).

The newer 54C missile seems to perform significantly worse below 40,000  (not by a little bit, but by a massive amount, sometimes with 100% kill rate for the A vs 0% kill rate with the C).  

Anyway, I'm off to fly some mid 80s campaigns now, before the C model was introduced.   

Edited February 18, 2022 by Clunk1001

[mettschnitte]

vor 2 Minuten schrieb Clunk1001:

The newer 54C missile seems to perform significantly worse below 40,000  (not by a little bit, but by a massive amount, sometimes with 100% kill rate for the C vs 0% kill rate with the A). 

The three missiles share different motors, the mk.47 and the mk.60 The mk.47 is a shorter range, shorter burning motor than the mk.60
They may seem to perform similar, but the long range performance (>30nm) differs greatly between the two motors

[draconus]

7 minutes ago, Clunk1001 said:

The newer 54C missile seems to perform significantly worse below 40,000  (not by a little bit, but by a massive amount, sometimes with 100% kill rate for the C vs 0% kill rate with the A).

You meant the other way around... but it's obvious that you need different employment parameters with different missile/engine verison.

[Clunk1001]

14 minutes ago, draconus said:

You meant the other way around... but it's obvious that you need different employment parameters with different missile/engine verison.

Yeah, just noticed and corrected that.

"you need different employment parameters with different missile/engine verison"   Do I?   Are saying that the C genuinely performed that bad at 30,000 feet?!   And that pilots flew with 2 C phoenix knew that they had little chance employing them successfully at/below 30,000 in a head on?

My point is there seems to be a massive discrepancy between deployment of the C at 30000 vs 40000,  and it doesnt feel like it's anything to do with 'employment parameters' ….but people keep jut saying use different tactics….but I’m saying somethings not right.

Edited February 18, 2022 by Clunk1001

[Lurker]

18 hours ago, Clunk1001 said:

Everyone seems to be saying it's 'tactics' , or 'geometry', or the wrong missile.  I am always open to suggestion/help/criticism.  But can anyone help and offer a solution to the above scenario that doesn't involve using 12 phoenix against 3 Flankers?  Or are we saying that it is a realistic expectation to require 4+ Phoenix missiles per rookie flanker (in which case I'll shut up and assume the missile is just fine  )?

 

Yes. Don't fire all of them at once. If the missile makes the aggressor turn around, it's doing it's job. You can't expect it to fly 25 miles and then chase down the target. If you're expecting that, then you're playing the wrong game. Fire a missile each, crank and see what happens. If they go cold on you, press. Once they turn hot, fire again, crank. If they keep pressing they will die. If they somehow manage to avoid the second wave, and live, then you are the one who has to turn cold, or risk getting killed. 

Edited February 18, 2022 by Lurker

[Clunk1001]

39 minutes ago, Lurker said:

Yes. Don't fire all of them at once. If the missile makes the aggressor turn around, it's doing it's job. You can't expect it to fly 25 miles and then chase down the target. If you're expecting that, then you're playing the wrong game. Fire a missile each, crank and see what happens. If they go cold on you, press. Once they turn hot, fire again, crank. If they keep pressing they will die. If they somehow manage to avoid the second wave, and live, then you are the one who has to turn cold, or risk getting killed. 

 

That’s doesn’t make sense with TWS.  Fire one missile, see how it goes? - but my next missile is going to target no2 regardless of what missile 1 or bandit 1 does.

in my tests,  tws on 3 targets the timing between release (3 seconds or 10 seconds) makes little difference (at 30,000).

am I wrong?

Edited February 18, 2022 by Clunk1001

[some1]

I did a few practice engagements against an AI Flanker at 35'000 feet recently, and did not really see much guidance problems, most often a single mk.47 missile launched from 30 nm scores a hit. 

It may have something to do with how AI responds to my actions, as in my case it's pressing with attack, and fails to beam or outrun the missile. 

Tacview-20220217-210410-DCS-bvr_F14_LR.zip.acmi Tacview-20220217-210658-DCS-bvr_F14A_LR.zip.acmi

[Clunk1001]

14 minutes ago, some1 said:

I did a few practice engagements against an AI Flanker at 35'000 feet recently, and did not really see much guidance problems, most often a single mk.47 missile launched from 30 nm scores a hit. 

It may have something to do with how AI responds to my actions, as in my case it's pressing with attack, and fails to beam or outrun the missile. 

Tacview-20220217-210410-DCS-bvr_F14_LR.zip.acmi 146.51 kB · 0 downloads Tacview-20220217-210658-DCS-bvr_F14A_LR.zip.acmi 154.14 kB · 0 downloads

And at 30,000 with a C ?   How does that work out for you?

Edited February 18, 2022 by Clunk1001

[Lurker]

20 minutes ago, Clunk1001 said:

That’s doesn’t make sense with TWS.  Fire one missile, see how it goes? - but my next missile is going to target no2 regardless of what missile 1 or bandit 1 does.

in my tests,  tws on 3 targets the timing between release (3 seconds or 10 seconds) makes little difference (at 30,000).

am I wrong?

 

No, if you're facing three bandits, then fire one missile each. That would be 3 missiles in the air at a time, maximum. You still have 3 missiles left. Order your wingman to hold fire (although with the AI in DCS the way it's working it's questionable if you can even control him properly). 

Ideally you should be testing 1 v 1  or 1 v 2 just so you can have more missile volleys per sortie. My point is (and the point everyone else is trying to make here) is that you can't ever expect a certain PK, not even against AI, there are too many factors in play. You're testing missile kinematics and guidance, and for those tests to be as accurate as possible you need to use proper tactics when engaging. 

[Clunk1001]

26 minutes ago, Lurker said:

No, if you're facing three bandits, then fire one missile each. That would be 3 missiles in the air at a time, maximum. You still have 3 missiles left. Order your wingman to hold fire (although with the AI in DCS the way it's working it's questionable if you can even control him properly). 

Ideally you should be testing 1 v 1  or 1 v 2 just so you can have more missile volleys per sortie. My point is (and the point everyone else is trying to make here) is that you can't ever expect a certain PK, not even against AI, there are too many factors in play. You're testing missile kinematics and guidance, and for those tests to be as accurate as possible you need to use proper tactics when engaging. 

I disagree,  I only need the tactics to be consistent across all the tests for a test to be valid (even if the tactics are wrong).

My test was - crank 30, go head on and launch 1 missile each (TWS) at 60 mile, then crank 30 degrees, wait, then re-engage as required.   

My result is that with the C at 30,000 I'll often watch them all miss.  With with the A in the same engagement I'll watch them hit most if not all of the bandits.   

I can expect these same results, again, and again, and again, and again against the AI.   And I can prove it with all the tracks I have.   

My conclusion is that the C isn't performing as it should at 30,000. 

And as I've mentioned, every track I've seen posted to justify it's all about use of tactics (yours included) does not show a C being employed at 30,000.   They all seem to be either an A at 35, or a C at 40, or an A at 30 (all of which work fine)

Anyone else want to join the firing a C at 30,000 club and see what I'm talking about?   Or shall I await another track showing an A fired at 40,000ft along with an 'it's all about tactics' comment?    

 

Edited February 18, 2022 by Clunk1001

[Lurker]

13 minutes ago, Clunk1001 said:

My result is that with the C at 30,000 I'll often watch them all miss.  With with the A in the same engagement I'll watch them hit most if not all of the bandits.   

 

When you say A you mean the A mk60 with the longer burning motor? Yes it appears that the C seems to be the worst performing variant kinematically, which is something that has been confirmed by Heatblur to be still in investigation. (I.E. whether thats the way it was in real life or not). 

I don't know whether that's intentional or not but the C has a smokeless motor, and is meant be used more as a poor man's AMRAAM and at similar ranges, it sacrifices reach for stealth (no launch indication) and better guidance and terminal performance. I have no idea how it was in real life, but it's probably a fair bet to say that in DCS World this is not simulated for the C at all. In other words we get all of it's disadvantages and no advantages, except for the smokeless launch, which is wasted on the AI anyway. 

Edited February 18, 2022 by Lurker

[Clunk1001]

6 minutes ago, Lurker said:

When you say A you mean the A mk60 with the longer burning motor? Yes it appears that the C seems to be the worst performing variant kinematically, which is something that has been confirmed by Heatblur to be still in investigation. (I.E. whether thats the way it was in real life or not). 

THANK YOU!!!!

that’s exactly the info that I was looking for!

Sorry folks if I was late to the party on that bit of info -  I obviously didnt articulate my point well enough to get that reply/info sooner.  
Anyway, I have found a lot of the info very useful - thanks again.  Now…off to the 80s for me…

Edited February 18, 2022 by Clunk1001

[draconus]

14 minutes ago, Clunk1001 said:

My conclusion is that the C isn't performing as it should at 30,000.

And how it should perform in your opinion? What's with you and 30,000? You can launch even at SL if you must, just expect much shorter effective range. There is nothing special about 30,000 - just don't expect it to perform the same as higher up or C the same as Amk60.

[Clunk1001]

45 minutes ago, draconus said:

And how it should perform in your opinion? What's with you and 30,000? You can launch even at SL if you must, just expect much shorter effective range. There is nothing special about 30,000 - just don't expect it to perform the same as higher up or C the same as Amk60.

Good question.  As mentioned earlier I would expect to a degradation of performance with decreasing density altitude (which we have);  the results with the C seems okay at 40,000 (nobody really knows).  The reason I’m highlighting 30,000 is that this is where the C seems to break.  It seems to go from 85% kill ratio at 40,000 down to around 10% at 30,000 (in my head on test).  And it’s that inconsistency which I’m highlighting.
 

I’ve no idea how it should actually perform,  but I’m pretty sure that the density altitude variation between these two heights would not account for such a dramatic shift in usability.

 

Edited February 18, 2022 by Clunk1001

[draconus]

43 minutes ago, Clunk1001 said:

Good question.  As mentioned earlier I would expect to a degradation of performance with decreasing density altitude (which we have);  the results with the C seems okay at 40,000 (nobody really knows).  The reason I’m highlighting 30,000 is that this is where the C seems to break.  It seems to go from 85% kill ratio at 40,000 down to around 10% at 30,000 (in my head on test).  And it’s that inconsistency which I’m highlighting.

There's no particular altitude at which the missile kinematics break. It's just one of many launch parameters. Want to launch at 30k and still hit a target like it did from 40k? You have these options: go faster at launch, lob it a bit, launch closer. Obviously lower alt means higher air density which causes more drag, slower acceleration, the missile gets lower max speed, gets to the lower lofting alt and then have lower energy in terminal phase.

[Clunk1001]

8 minutes ago, draconus said:

There's no particular altitude at which the missile kinematics break. It's just one of many launch parameters. Want to launch at 30k and still hit a target like it did from 40k? You have these options: go faster at launch, lob it a bit, launch closer. Obviously lower alt means higher air density which causes more drag, slower acceleration, the missile gets lower max speed, gets to the lower lofting alt and then have lower energy in terminal phase.

“There's no particular altitude at which the missile kinematics break.”

Not in real life, no.

But in software simulation, this is entirety possible.  “Break” is not the correct term - becomes noticeable to the point where it seems to be incorrect. 
 

Edited February 18, 2022 by Clunk1001

[Strider21]

6 hours ago, IronMike said:

They are being notched in their final lead turn. You can see how the AI does exactly the same maneuver over and over again, being omniscient as it is. For the Su-30, you need to fire either further out, or much closer in (mind the r77 though). I posted tacviews of how to deal with 2 of them in an earlier comment. If you do the head-on thing their very effective defense will simply be the same over and over again. So you need to make them react to something else upfront, which offsets this ultra-notch of theirs. Look at the tacviews below, they are the same ones I mention. You can see how in the long range shot and the Cmk47 shots they do it as well, and one out of 2 defends a missile. That's ok, he's been put defensive and now I can press and finish him. A missile does not need to always hit to be effective, the "game" that you play with your bandit, extends beyond that.

Let's do a very stupid example. FPS pvp, you advance on an enemy held house: you see a player popping his head out of a window, you aim, shoot, but just as you fire he ducks behind the wall. Your bullet misses. Do you blame the bullet for missing him? Or do you proceed to chuck a grenade through the window instead? Now, forgive me, it is a very stupid comparison I shall admit. But point remains: don't blame the bullet for your enemy taking cover. Don't blame the missile for your bandit notching it. Switch your game on, and take advantage of him getting his off of the offensive for these brief moments.

There are guidance issues remaining, but a missile being notched isn't really a guidance issue. It is a successful defense which you have to counter.
 

2xSu30_Amk60.acmi 165.39 kB · 5 downloads 2xSu30_Amk60_long.acmi 166.56 kB · 3 downloads 2xSu30_Cmk47.acmi 247.63 kB · 5 downloads

 

Thank you for you response and examples. Prior to the CFD update and HOTFIX for guidance when the missile was Notch'd it  would go straight and level and pass behind the bandit. It now consistently passes in front of the bandit while pulling lead. Is this part of the guidance update?  

So 30 NM is not a good range? Further out where should I be firing and closer in where should I be firing? How do you figure out the range to fire at various speeds and altitudes?

Edited February 18, 2022 by Strider21