DCS: F-14 Development Update - AIM-54 Phoenix Improvements & Overhaul - Guided Discussion

[lunaticfringe]

5 minutes ago, LanceCriminal86 said:

Never, because he was still wrong in most of his assertions and provided abusive and harassing DMs. There's a reason he's not here.

 

Somewhere he's still getting slapped by A's, so it all balances out. 

[CarbonFox]

From tests. Haven't noticed too much of difference in kill probability between the Mk 47 and Mk 60 motors against non-maneuvering targets (Bombers) over 65nm out. Tried the same against an Su-27 in two tests where each were launched at about 42nm from target at 39k going Mach 1.4 against the target at 35k. The Flanker turns and runs into thicker and none of missile hit. The AIM-54C Mk. 60 got within rock-throwing distance before the Su-27 walked away. The new 54C is still a nasty thing with the new guidance as it kept on the Flanker regardless of chaff, forcing it defensive for awhile. Enough time for me to close in an tag it with a Sparrow.

Edited September 6, 2022 by CarbonFox

[Noctrach]

The graphs being that close is actually pretty crazy. Damn.

[Meteorlover]

It's amazing that the terrible performance degradation of the Aim-54 was ultimately driven by a single maniacal obsession over a long period of time.
Hb this acceptance.
After all, is HB's judgment that Phoenix is 100% in line with reality and that the previous Phoenix of this patch is a 'fictional missile'?
I'm very curious about why one person's data has the credibility to have a top priority over another.

Looking at the process immediately after Early Access and this year's process, HB is very confused about the performance of the AIM-54.The AIM-54 feels particularly confused compared to other third-party weapons such as the SD-10.

[Naquaii]

52 minutes ago, Meteorlover said:

It's amazing that the terrible performance degradation of the Aim-54 was ultimately driven by a single maniacal obsession over a long period of time.
Hb this acceptance.
After all, is HB's judgment that Phoenix is 100% in line with reality and that the previous Phoenix of this patch is a 'fictional missile'?
I'm very curious about why one person's data has the credibility to have a top priority over another.

Looking at the process immediately after Early Access and this year's process, HB is very confused about the performance of the AIM-54.The AIM-54 feels particularly confused compared to other third-party weapons such as the SD-10.

 

49 minutes ago, AirMeister said:

Yes indeed. Imagine being trolled for years while being right all this time. 
Unfair. 
On the other hand, his ban would not prevent him to read the thread i guess so if he still plays he's probably doing so with a smile of satisfaction on his face. 

He was never right, what he advocated was way worse than this, he'd have the AIM-54 be a nerfed Sparrow as in his mind the AIM-54 should just be unable to kill fighters period. I'm not even gonna talk about how he behaved on these forums while ranting about this.

That said none of this influenced our research or decision-making. Is it so hard to believe that we continuously research and refine our models?

That's also the reason for the implementation of the "active on it's own"-AIM-54C taking time. That we wanted to research and verify the information regarding this as the publically available information wasn't enough.

Believe that if you want or not but we have really no reason to not be honest about this.

Edited September 6, 2022 by Naquaii

[LanceCriminal86]

Gotta love the forgetfulness that the AIM-54's aero and motor performance had to be "adjusted" in order to even remotely perform like the real missile due to the old guidance and missile API (or really lack of the API) for the long shots. Or assuming that the values provided for the missile at launch were supplied in a vacuum with no review or input from ED. Even with those physics and guidance there were complaints the missile was underperforming at long/mid range because the guidance logic would burn so much energy off maneuvering.

With the current physics and old, non-API guidance the missile would probably have been far worse than any iteration so far seen in release or open beta. Without the added values to adjust loft, lead calculation, etc. we'd be seeing a much shorter range missile, burning off what less energy it has now, with terrible loft profiles. And of course the inability to model differences in the A and C. Let's go back to where doing barrel rolls makes you unhittable, and missiles burning up all their energy before they get anywhere near the target from course corrections and maneuvering.

And the most entertaining bit of course are the new and super low-post accounts showing up lately while the folks that had hounded the previous AIM-54 threads with criticism seem to be mostly silent or in agreement that the latest changes are making sense.

Edited September 6, 2022 by LanceCriminal86

[Meteorlover]

Don't be too conceited.
We know that everything about DCS is stacked on top of LOMAC's old, unrealistic script, and that certain missiles are excessively stupid and certain missiles remain unchanged with untested performance.
The same goes for unrealistic notching
In other words, most of them are ED's fault
But if a particular person wins a long debate and accepts it and creates something that affects everyone, then the rationale and the process of judgment must be clarified That's the basic.

[Naquaii]

2 hours ago, Meteorlover said:

Don't be too conceited.
We know that everything about DCS is stacked on top of LOMAC's old, unrealistic script, and that certain missiles are excessively stupid and certain missiles remain unchanged with untested performance.
The same goes for unrealistic notching
In other words, most of them are ED's fault
But if a particular person wins a long debate and accepts it and creates something that affects everyone, then the rationale and the process of judgment must be clarified That's the basic.

If by winning you mean arguing and whining for something and then declaring victory when given something completely different than what you asked for, sure. I'd argue that you're reading far too much into something that is really, really simple. We change our modelling when we have new data, that's basic and easy to understand.

And none of what you listed above about old code precludes the possibilty of refining and improving what we have.

Edited September 6, 2022 by Naquaii

[captain_dalan]

10 hours ago, The_Tau said:

I actually managed to beat that mission but I must say, 54A seeker is really a gamble here and you must be lucky. Even if you do everything correctly 54A seeker may just eat the chaff and there it went. 

Basically I made my wingman in trail 50-60 miles and I go alone, climb!! (50k), fire 3 54s (save 1 for later pitbull shots) at 50-60 nms, split S, wait for JF17 to reorganize after first attack, send wingman to engage, then turn around to support him and hope that 54As actually hit something.

I turned MLC off in RIO pit, we fight over water so AWG9 should have much easier job maintain track

ECM makes JF17s fire later so I use that on wingman as well

I am adding tacview

https://we.tl/t-yYdPwE2J0Y

 

(This was my 8th take on this , clear example why C is superior to A, as it should be)

 

Definitely doable with Charlie Mk60's, even only on my second attempt, the AI even defended quite aggressively on all launches. Nothing fancy on my part either (like say exploiting the wingman like i did on the previous run to make him sponge enemy missile fire). I'll upload the tacview bellow. I didn't bother with Charlie Mk47's, as to be quite honest i've had about enough of this mission for a while! Besides, i stand by my initial conclusions, except for extremely long shots, Mk60's are better. They accelerate faster, they maintain higher glide velocity and the fact they burn less, means that even with the smoke, the bandit isn't gonna spot the plumes from the distance you will usually launch these things. 

Short description of the hop: climbed to angels 46-47, coasted to mach 0.9 (took zone 4 just to stay there, yeah, the F-14 is practically subsonic with 2 bags and 4 Phoenixes above 36000ft) indicated, fired most of the missiles between 50 and 60 miles distance. I don't know if it helped, but i tried not to have the AWG9 track more then 2 bandits while i had non-active missiles in the air. I splashed one and the wingman splashed another (the one that notched my first shot - not so easy to nothc two mach 1.6 missiles at the same time). This meant that by the time we closed in, it was now 2 on 2, a much more manageable situation. I dragged one of the SD-10's into the pitch black ocean and let it's launcher's wingman eat 90pds of TNT before checking in with the Poseidon's locker, while my wingman took care of him. I wanted to go for the enemy tanker as well, but i was hitting bingo by this time, and the AWACS insisted there were to tankers around to refuel.  

4 hours ago, CarbonFox said:

From tests. Haven't noticed too much of difference in kill probability between the Mk 47 and Mk 60 motors against non-maneuvering targets (Bombers) over 65nm out. Tried the same against an Su-27 in two tests where each were launched at about 42nm from target at 39k going Mach 1.4 against the target at 35k. The Flanker turns and runs into thicker and none of missile hit. The AIM-54C Mk. 60 got within rock-throwing distance before the Su-27 walked away. The new 54C is still a nasty thing with the new guidance as it kept on the Flanker regardless of chaff, forcing it defensive for awhile. Enough time for me to close in an tag it with a Sparrow.

 

The kill probability for non-maneuvering targets won't be that different indeed. Buch the check the mach-airspeed numbers before impact. Mk60's carry a small but NOT insignificant amount of kinetic energy more then the Mk47's. This energy can mean (and it often does) the difference between a successful Split-S and a fireball in the sky.

2 hours ago, Meteorlover said:

Looking at the process immediately after Early Access and this year's process, HB is very confused about the performance of the AIM-54.The AIM-54 feels particularly confused compared to other third-party weapons such as the SD-10.

I find it odd how some people forget that during the 4 years we've had the F-14 and the AIM-54, both at times actually underperformed when compared to now...... and in many aspects still do, compared to how they should..... selective memory is a bliss! 

20 minutes ago, Бойовий Сокіл said:

I feel like the 54 is in a good spot now, probably as good as it will ever be, apart from some smaller changes to fine-tune it in certain regimes. The fact it resembled official flight test data this closely speaks volumes. 

Agreed. It did "seem" that way, right from the start of my tests. Aside from some loft-guidance logic peculiarities, but i don't think those will change the overall performance that much if/when implemented. 

Tacview-20220906-200132-DCS-September update F-14A_IA_Marianas_BVR_JF17 Mk60C.zip.acmi

Edited September 6, 2022 by captain_dalan

[IronMike]

I just want to re-iterate that none of these changes have anything to do with CSGO-yeah. Pardon me if I got the username partially wrong. 

Not once did he present any data or factual reference that would have contained something we could act on. Instead he continuously expressed his feelings and opinions about the Tomcat, the Phoenix and us. His basic reasoning was very simple: "It was developed against bombers, so it should not work against fighters." and: "It was old, so it should be useless." and of course: "Heatblur made it 'op' to sell more modules." On top of that he used his comments to a) express his hate agains the phoenix, b) his hate against the Tomcat and Tomcat community and c) his accusations against us - which continued to show in his impulsive and negative behavior towards both us and everyone around. I use the word "hate" specifically, because he himself used it repeatedly in conversations with me/us/here. This is certainly not the kind of input we listen to, yet we have been patient with him for more than 2 years, despite repeated ban requests by forum users due to his continued negative behavior, which ultimately culminated in a barage of absolutely insulting and inacceptable PMs that led to his ban, after being warned several times, and given several chances, both by us and ED. His opinions had nothing to do with him being banned.

The entire process of us refining the phoenix started already in autumn 2019, long before anyone made any such reference or claims towards what it should or should not be, based on their "impressions". And with this I would like us all to close the off-topic convo about CSGO, and return to the issues at hand. Thank you.

When simulating such complex matters as missiles, data is what you can go by, and trying to make it achieve what (little or limited) is known from real life that could be achieved. We had to over-power it at first, which btw left it still underpowered at the time. With the missiles in DCS changing in the recent years, this offset its performance such that it became eventually overpowered, upon which we continued to dive in deeper, and with the help of your feedback (factual feedback by so many of you), the help of SMEs and by now years of research, we finally managed to learn what we did not know before, or only knew in a more limited way. But the goal was since the beginning: to make it perform as close as possible to known IRL shots. As @Naquaii said: we change the modeling when we have new data. This has never been, nor ever will be based on opinions, balance, feelings, complaints or any of that kind. But it will of course always happen, when someone presents us with evidence that either shows us what we can improve or what we got wrong or what we can change to make it even more realistic. The impact this has on whether the missile becomes "better" or "worse", is, to put it bluntly, irrelevant. And I, too, second the idea to not over-interpret too much into the pro and contra going forth between "opposing factions". It does not influence our development.

What does influence it, is the many great input coming from you guys that for example showed the CFD being off, quirks in the guidance, in lofting, and in many many areas you have and continue to unearth. And we are, as always, very grateful for that. As mentioned at the beginning of this thread: it is a journey of learning, and without you guys being on board, we would never advance faster than we are now. And as long as we stand, I hope this journey will never end.  

Edited September 6, 2022 by IronMike

[freehand]

The HB sub forum is hilarious I all ways come here for giggles.

[NeedzWD40]

Re: BVR Hunting the Jeff Marianas mission

I'm noticing a lot of peculiarities on this one, not just with the AIM-54. Firstly, it seems that the SD-10s launched by the JF-17s go active at about 16-18nmi and will continue to ping my RWR for ~6 minutes, including actively trying to track my aircraft. On top of this, if I get within their launch range, quite often the SD-10s will grab onto AIM-54s in the air, including those of my wingman. This makes things a bit more difficult since they pack 16 SD-10s altogether, so they'll almost instantly sling another missile out if parameters are good.

The default payload seems only good for causing the initial run to break off and maybe waste one or two of their missiles. The A models really get trashed easily, but even the C models cannot keep up with the JF-17's acceleration, even if you wait until the last possible moment to launch. The wingman is real gung ho and I've repeatedly watch him fly right into SD-10s coming right at him, so he's only been good for a quick suppression. After that, he easily gets knocked down.

The AIM-7Ms by default are total trash. I've been in perfect shot parameters and all the JF-17s have to do is a small turn and the missile goes dumb. Same for the AIM-9Ms, as it seems even with an afterburner tail shot the JF-17's flares will cause an instant miss. I've frequently just ran away til the JF-17s are out of gas, then turn back and try to hit them while they're stalling out and they still manage to lose both my 9s and my 7s.

The only sure success I've been able to have is using an STT AIM-54, but that only wipes out one with certainty, leaving 3 left that have no problems running you down. AI JF-17s can easily top out at 900KTAS, even with the double racks!

The other thing is I've noticed some real bizarre behavioral changes when I change the JF-17s to simply perform the CAP tasking and not engage group. As in they quit flying high and instead get down low with more extreme separation. They can still easily dodge in this way, but they lose some aggressiveness for whatever reason. I've also tried moving them all to Ace level, but all that did was make them eat all 4 AIM-54s from 70nmi out (?!). I also removed all their countermeasures for testing purposes and they're pretty much just turning cold and leveraging their excellent acceleration to lose any AIM-54 shots.

Lastly, the F-14A seems to be better than the F-14B for this one. The lighter weight seems to help when you need to evade quickly to drag an SD-10 out.

It's peculiar because outside of this mission, the AIM-54s have been working incredibly well for me. It seems more like the JF-17/SD-10 combo is simply a much more modern threat and probably pushing the limits of what the F-14 can tackle in a disadvantageous situation.

[WarthogOsl]

FWIW, I've had issues with the Hunting the Jeff mission with every incarnation of the Phoenix since Marianas was a thing.

Which leads to my next question:  What I call "the Phoenix head-nod."  I've seen this with every incarnation of the Phoenix as well, including the most current one.  This is where the missile seems right on track for a good intercept, but when it gets right near the bandit, it pitches up 10-15 degrees, diverging from the intercept course, and going dumb.   I don't know if this is because of chaff, or notch, because I never see other missiles exhibit this behavior.  For example, if a Sparrow gets chaffed, it just seems points back along the bandits flight path, presumably looking at the chaff.  The Phoenix, with its pitch up, seems to be very deliberately doing....something with its guidance.  Is there some default behavior that says "if you lose an active radar contact, pitch up?"

[captain_dalan]

1 hour ago, WarthogOsl said:

FWIW, I've had issues with the Hunting the Jeff mission with every incarnation of the Phoenix since Marianas was a thing.

Which leads to my next question:  What I call "the Phoenix head-nod."  I've seen this with every incarnation of the Phoenix as well, including the most current one.  This is where the missile seems right on track for a good intercept, but when it gets right near the bandit, it pitches up 10-15 degrees, diverging from the intercept course, and going dumb.   I don't know if this is because of chaff, or notch, because I never see other missiles exhibit this behavior.  For example, if a Sparrow gets chaffed, it just seems points back along the bandits flight path, presumably looking at the chaff.  The Phoenix, with its pitch up, seems to be very deliberately doing....something with its guidance.  Is there some default behavior that says "if you lose an active radar contact, pitch up?"

On the couple of occasions i've seen that in tacview, it appeared to be notching. Going for CM looks the same as it does with the Sparrow, that is, the missile will start going after some artefact behind the bandit. 

3 hours ago, NeedzWD40 said:

It's peculiar because outside of this mission, the AIM-54s have been working incredibly well for me. It seems more like the JF-17/SD-10 combo is simply a much more modern threat and probably pushing the limits of what the F-14 can tackle in a disadvantageous situation.

To me it looks like the SD-10 is just bogus. Vipers and Eagles with AMRAAMs, even Foxhounds with R-33's don't come close to being that much of a threat. Or maybe the Chinese really have developed a mach 5 capable missile in an AMRAAM sized rocket.....in which case, God/s help us all....

[Spurts]

So I did my 150nm intercept deck launch with an F-14A dragging 6xAIM-45C-60.  My previous climb profile (accel at 30k) will not work (1.23M peak after a long run), so I had to gain altitude (to 44k @.99M) and then dive (to 33k @1.4M).  I was able to hold 1.4+M up to 46k where I fired.  On to the missile.

Target was Tu-22 head on at 40k @ 1.86M.  Range was 66nm (I forgot to equip the Phoenix to shoot any sooner.  Max pitch of 35deg 10s after launch.  At motor burn out pitch was down to 22deg.  68k@3.88M with a slant range to target of 52nm.  Appogee at 80k @ 3.37M with 35nm slant range.  At 10nm slant range it was 24deg nose down diving from 57k @ 3.07M.  Impact 14s later still doing 2.75M (target did not evade).  I pulled power and dove but did not crank, I was 22nm from target at impact at 37k and 1.31M.

No complaints on this easy test.  3.0M+ average speed and 2.75M terminal.

[WarthogOsl]

1 hour ago, captain_dalan said:

On the couple of occasions i've seen that in tacview, it appeared to be notching. Going for CM looks the same as it does with the Sparrow, that is, the missile will start going after some artefact behind the bandit. 

So, lets assume it is a notch...I still wonder if the pitch up is intentional behavior by either HB or the ED API.  Because often it seems like if the missile just continued with no further guidance, it could potentially still intercept rather then flying over the bandit.  I dunno, maybe the API is so limited that the only way to simulate a Fox 3 notch is to make the missile swerve hard enough that it loses lock and doesn't have time or space to reaquire, but I am curious.

[DSplayer]

1 hour ago, captain_dalan said:

To me it looks like the SD-10 is just bogus. Vipers and Eagles with AMRAAMs, even Foxhounds with R-33's don't come close to being that much of a threat. Or maybe the Chinese really have developed a mach 5 capable missile in an AMRAAM sized rocket.....in which case, God/s help us all....

The SD-10 actually had an incorrect nozzle_exit_area value that gave it a little more extra thrust and speed at higher altitudes up until the recent patch after I reported it. It should be pretty reasonable in terms of speed atm. You just gotta keep in mind that the SD-10 has a ~20mm larger diameter and ~100mm longer than the 120C-5 with probably a different internal layout. AIM-120C-7 for DCS when ED smh /s.

 

29 minutes ago, WarthogOsl said:

So, lets assume it is a notch...I still wonder if the pitch up is intentional behavior by either HB or the ED API.  Because often it seems like if the missile just continued with no further guidance, it could potentially still intercept rather then flying over the bandit.  I dunno, maybe the API is so limited that the only way to simulate a Fox 3 notch is to make the missile swerve hard enough that it loses lock and doesn't have time or space to reaquire, but I am curious.

Seems like its an older API/scheme thing along with maybe induced lift and drag? The SD-10 on its pre-INS datalink AMRAAM scheme doesn't have that issue and flies in a straight line until it regains lock and the AIM-120s with the INS datalink basically tries to follow predict where you'll be next when it loses track.

 

[lahsoldier]

On 9/5/2022 at 1:20 PM, IronMike said:

@IronMike "I can try some on veteran if you like" 

 If you don't mind, can you do one at that altitude verse veteran Mig 29s(s) and see if they split S on you.   If not watch what happens when they do.  If they do its almost a guarantee miss at least for me. Bad guys 30k set on whatever setting they are default with besides veteran. You 46K, 35 mile M1+ TWS shot.    I have done this full burner, 46k, 35 mile  tws shot and so far have hit 1 out of 10 and he for some reason decided not to split S.

  I would be very interested in your opinion and how (if you were able to get relatively consistent kills, how.   Thanks for your time.

 

 

[WarthogOsl]

Here's an example of the "head-nod" that just happened.  Was the missile that missed notched?  It seemed to be tracking okay well past when the bandit was at a 90 degrees course to the missile.  In this case, I actually wonder if the missile was putting so much lead on the target that it went past it's gimble limits, as the bandit is just about at the missile's 9 o'clock when it suddenly pitches up.  Can that happen?  And again, why does it pitch up when it happens?

 

Tacview-20220906-231924-DCS-Through_The_Inferno_MI_v1.zip.acmi

[Meteorlover]

12 hours ago, IronMike said:

I just want to re-iterate that none of these changes have anything to do with CSGO-yeah. Pardon me if I got the username partially wrong. 

Not once did he present any data or factual reference that would have contained something we could act on. Instead he continuously expressed his feelings and opinions about the Tomcat, the Phoenix and us. His basic reasoning was very simple: "It was developed against bombers, so it should not work against fighters." and: "It was old, so it should be useless." and of course: "Heatblur made it 'op' to sell more modules." On top of that he used his comments to a) express his hate agains the phoenix, b) his hate against the Tomcat and Tomcat community and c) his accusations against us - which continued to show in his impulsive and negative behavior towards both us and everyone around. I use the word "hate" specifically, because he himself used it repeatedly in conversations with me/us/here. This is certainly not the kind of input we listen to, yet we have been patient with him for more than 2 years, despite repeated ban requests by forum users due to his continued negative behavior, which ultimately culminated in a barage of absolutely insulting and inacceptable PMs that led to his ban, after being warned several times, and given several chances, both by us and ED. His opinions had nothing to do with him being banned.

The entire process of us refining the phoenix started already in autumn 2019, long before anyone made any such reference or claims towards what it should or should not be, based on their "impressions". And with this I would like us all to close the off-topic convo about CSGO, and return to the issues at hand. Thank you.

When simulating such complex matters as missiles, data is what you can go by, and trying to make it achieve what (little or limited) is known from real life that could be achieved. We had to over-power it at first, which btw left it still underpowered at the time. With the missiles in DCS changing in the recent years, this offset its performance such that it became eventually overpowered, upon which we continued to dive in deeper, and with the help of your feedback (factual feedback by so many of you), the help of SMEs and by now years of research, we finally managed to learn what we did not know before, or only knew in a more limited way. But the goal was since the beginning: to make it perform as close as possible to known IRL shots. As @Naquaii said: we change the modeling when we have new data. This has never been, nor ever will be based on opinions, balance, feelings, complaints or any of that kind. But it will of course always happen, when someone presents us with evidence that either shows us what we can improve or what we got wrong or what we can change to make it even more realistic. The impact this has on whether the missile becomes "better" or "worse", is, to put it bluntly, irrelevant. And I, too, second the idea to not over-interpret too much into the pro and contra going forth between "opposing factions". It does not influence our development.

What does influence it, is the many great input coming from you guys that for example showed the CFD being off, quirks in the guidance, in lofting, and in many many areas you have and continue to unearth. And we are, as always, very grateful for that. As mentioned at the beginning of this thread: it is a journey of learning, and without you guys being on board, we would never advance faster than we are now. And as long as we stand, I hope this journey will never end.  

 

 

I understand.
I apologize for using an overly emotional and strong tone.
Then may I ask a question?
It's not just an AIM-54 phenomenon's
Missile left or right or up and down even though the target is nearby. Or even lose all your energy while rolling(Even though it was a hit if it was guided normally) I've still seen a lot of that.
It doesn't matter if it's a real missile's phenomenon, but I always think it's close to a script "error."Is this a problem with the API or is it possible to modify the anonymous script? Or there's no problem? I want to hear your thoughts
It's a pure question that I've thought about for a long time.

[IronMike]

2 hours ago, Meteorlover said:

 

I understand.
I apologize for using an overly emotional and strong tone.
Then may I ask a question?
It's not just an AIM-54 phenomenon's
Missile left or right or up and down even though the target is nearby. Or even lose all your energy while rolling(Even though it was a hit if it was guided normally) I've still seen a lot of that.
It doesn't matter if it's a real missile's phenomenon, but I always think it's close to a script "error."Is this a problem with the API or is it possible to modify the anonymous script? Or there's no problem? I want to hear your thoughts
It's a pure question that I've thought about for a long time.

There is no need to apologize at all, all good.

It definitely is a problem, just as you describe it. It's not at all a real missile phenomenon, it is a bug in the guidance, or maybe a flaw in the logic, or a limitation of the current model, which is partially still unknown to both us and ED. We are currently trinyg to chase this down - if anyone has a short track with it happening, please share! - because it makes the missile do really stupid things, like weaving above a target to match the intercept course, but like, not even pointing down to make the actual lead intercept with just 500 feet left, etc..

These are guidance specific issues, which we cannot fix on our own, but we're working with ED on all of these. We take this matter very seriously (and so does ED), btw, because we think this is one of the last remaining major issues that hampers the phoenix (and missiles in general). Also noteworthy is that due to its long range, the phoenix pronounces such issues much more than other missiles, too. Which makes it double as important to find a way to fix this for the phoenix, because it is much more affected by it than shorter ranged missiles. A very good and recent example for that was the video shared by @Soulres - a shorter ranged missile would have likely ran out of steam before making it around that mountain and a faster missile may have lead more aggressively and crashed into the mountain. This particular issue (flying around a mountain) was believed to be fixed when the "magic INS" bug got fixed couple years ago, but now we know this still exists, too. And this all boils down to these guidance issues that need to be fixed both short term and long term. 

To be transparent on how that works - strictly speaking guidance related issues now that are generally out of our hands: we collect what we can find on the issue (and have been doing so for a long time ofc), then present the case to ED, upon which they propose a change, implement a fix themselves, or guide us in adjusting certain parameters, and then rinse and repeat. It is a very difficult and unfortunately time consuming process, stemming from the complexity of coding missiles and the unreliable reproducability of these issues: we all have seen them, we know they exist, we even experience them on a regular basis, but they do not happen so consistently that it can be 100% reproduced/debugged by the developers both from us and ED. I am not saying this to excuse the admittedly very slow process btw, just to explain to you how we go about it, so you guys understand what is going on behind closed doors in general. And it is also important to know that ED is very forthcoming and helpful in these matters. It is just a very, very painful process for all parties involved, as in difficult to solve. In that sense we are very grateful for your very kind patience.

This particular phenomenon you mention is of course not all that remains to be fixed, but in our opinion it is one of the biggest remaining issues.

I also want to add to the lofting debate: while some of the lofts are excessive at short to medium ranges at times, often times they are benefitial as are (provided you do not overpitch and the missile goes to space), and if you compare them with a more shallow or non lofted profile, arrive faster at the target, as if they would when travelling through thicker air for a longer time. Again, this is not to say that there are no issues with loft, there 100% are, and these, too, we are looking to change and improve ofc.

Finally the transition to the new missile schema will get rid of some limitations the old missile API poses on guidance, missile logic, etc. and may additionally help to tackle parts of the above mentioned - which remains to be seen if and how much. 

We are not done with the phoenix, and we will continue improving it and working on the remaining issues.

 

Edited September 7, 2022 by IronMike

[Meteorlover]

2 minutes ago, Бойовий Сокіл said:

Why are you using large letters? It won't help get your point across any better.

i used default.but write by smartphone.

[IronMike]

2 minutes ago, Бойовий Сокіл said:

Why are you using large letters? It won't help get your point across any better.

That's ok, I do not mind, and it may not even be intentional, may be typing on a phone, etc... Let's stay on topic please and all a bit more open minded towards each other. 

1 minute ago, Meteorlover said:

i used default.but write by smartphone.

Ha! Sniped me.

[okopanja]

1 hour ago, IronMike said:

Finally the transition to the new missile schema will get rid of some limitations the old missile API poses on guidance, missile logic, etc. and may additionally help to tackle parts of the above mentioned - which remains to be seen if and how much. 

Is there any roadmap on missiles?

E.g. it would be helpful if you could perhaps argue with ED that publishing the list would eliminate majority of noise that comes everytime some parameters change.

[Meteorlover]

1 hour ago, IronMike said:

There is no need to apologize at all, all good.

It definitely is a problem, just as you describe it. It's not at all a real missile phenomenon, it is a bug in the guidance, or maybe a flaw in the logic, or a limitation of the current model, which is partially still unknown to both us and ED. We are currently trinyg to chase this down - if anyone has a short track with it happening, please share! - because it makes the missile do really stupid things, like weaving above a target to match the intercept course, but like, not even pointing down to make the actual lead intercept with just 500 feet left, etc..

These are guidance specific issues, which we cannot fix on our own, but we're working with ED on all of these. We take this matter very seriously (and so does ED), btw, because we think this is one of the last remaining major issues that hampers the phoenix (and missiles in general). Also noteworthy is that due to its long range, the phoenix pronounces such issues much more than other missiles, too. Which makes it double as important to find a way to fix this for the phoenix, because it is much more affected by it than shorter ranged missiles. A very good and recent example for that was the video shared by @Soulres - a shorter ranged missile would have likely ran out of steam before making it around that mountain and a faster missile may have lead more aggressively and crashed into the mountain. This particular issue (flying around a mountain) was believed to be fixed when the "magic INS" bug got fixed couple years ago, but now we know this still exists, too. And this all boils down to these guidance issues that need to be fixed both short term and long term. 

To be transparent on how that works - strictly speaking guidance related issues now that are generally out of our hands: we collect what we can find on the issue (and have been doing so for a long time ofc), then present the case to ED, upon which they propose a change, implement a fix themselves, or guide us in adjusting certain parameters, and then rinse and repeat. It is a very difficult and unfortunately time consuming process, stemming from the complexity of coding missiles and the unreliable reproducability of these issues: we all have seen them, we know they exist, we even experience them on a regular basis, but they do not happen so consistently that it can be 100% reproduced/debugged by the developers both from us and ED. I am not saying this to excuse the admittedly very slow process btw, just to explain to you how we go about it, so you guys understand what is going on behind closed doors in general. And it is also important to know that ED is very forthcoming and helpful in these matters. It is just a very, very painful process for all parties involved, as in difficult to solve. In that sense we are very grateful for your very kind patience.

This particular phenomenon you mention is of course not all that remains to be fixed, but in our opinion it is one of the biggest remaining issues.

I also want to add to the lofting debate: while some of the lofts are excessive at short to medium ranges at times, often times they are benefitial as are (provided you do not overpitch and the missile goes to space), and if you compare them with a more shallow or non lofted profile, arrive faster at the target, as if they would when travelling through thicker air for a longer time. Again, this is not to say that there are no issues with loft, there 100% are, and these, too, we are looking to change and improve ofc.

Finally the transition to the new missile schema will get rid of some limitations the old missile API poses on guidance, missile logic, etc. and may additionally help to tackle parts of the above mentioned - which remains to be seen if and how much. 

We are not done with the phoenix, and we will continue improving it and working on the remaining issues.

 

 

 

1 hour ago, IronMike said:

There is no need to apologize at all, all good.

It definitely is a problem, just as you describe it. It's not at all a real missile phenomenon, it is a bug in the guidance, or maybe a flaw in the logic, or a limitation of the current model, which is partially still unknown to both us and ED. We are currently trinyg to chase this down - if anyone has a short track with it happening, please share! - because it makes the missile do really stupid things, like weaving above a target to match the intercept course, but like, not even pointing down to make the actual lead intercept with just 500 feet left, etc..

These are guidance specific issues, which we cannot fix on our own, but we're working with ED on all of these. We take this matter very seriously (and so does ED), btw, because we think this is one of the last remaining major issues that hampers the phoenix (and missiles in general). Also noteworthy is that due to its long range, the phoenix pronounces such issues much more than other missiles, too. Which makes it double as important to find a way to fix this for the phoenix, because it is much more affected by it than shorter ranged missiles. A very good and recent example for that was the video shared by @Soulres - a shorter ranged missile would have likely ran out of steam before making it around that mountain and a faster missile may have lead more aggressively and crashed into the mountain. This particular issue (flying around a mountain) was believed to be fixed when the "magic INS" bug got fixed couple years ago, but now we know this still exists, too. And this all boils down to these guidance issues that need to be fixed both short term and long term. 

To be transparent on how that works - strictly speaking guidance related issues now that are generally out of our hands: we collect what we can find on the issue (and have been doing so for a long time ofc), then present the case to ED, upon which they propose a change, implement a fix themselves, or guide us in adjusting certain parameters, and then rinse and repeat. It is a very difficult and unfortunately time consuming process, stemming from the complexity of coding missiles and the unreliable reproducability of these issues: we all have seen them, we know they exist, we even experience them on a regular basis, but they do not happen so consistently that it can be 100% reproduced/debugged by the developers both from us and ED. I am not saying this to excuse the admittedly very slow process btw, just to explain to you how we go about it, so you guys understand what is going on behind closed doors in general. And it is also important to know that ED is very forthcoming and helpful in these matters. It is just a very, very painful process for all parties involved, as in difficult to solve. In that sense we are very grateful for your very kind patience.

This particular phenomenon you mention is of course not all that remains to be fixed, but in our opinion it is one of the biggest remaining issues.

I also want to add to the lofting debate: while some of the lofts are excessive at short to medium ranges at times, often times they are benefitial as are (provided you do not overpitch and the missile goes to space), and if you compare them with a more shallow or non lofted profile, arrive faster at the target, as if they would when travelling through thicker air for a longer time. Again, this is not to say that there are no issues with loft, there 100% are, and these, too, we are looking to change and improve ofc.

Finally the transition to the new missile schema will get rid of some limitations the old missile API poses on guidance, missile logic, etc. and may additionally help to tackle parts of the above mentioned - which remains to be seen if and how much. 

We are not done with the phoenix, and we will continue improving it and working on the remaining issues.

 

 

That's very good news
I commend ED and HB for their efforts and hard work
I think I can wait longer.
To be honest, I thought the ED was not interested in this issue.
How about notching?
At 90°, 100 percent of radar failure seems very artificial, especially when it comes to missiles Are you working on this as well?

 

번역  저장

Edited September 7, 2022 by Meteorlover