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

[captain_dalan]

2 hours ago, Exorcet said:

I think it might be slightly more accurate to say it capped loft altitude. The missile can still climb at pretty aggressive angles.

I can't tell if it's more accurate or not, as even if i had real data to base my claims on, i would not be allowed to disclose that data here. What i can tell however, base on relatively objective (ignore the oxymoron here) metrics is that the current capped loft, gives off statistically inferior missile performance, both in range/altitude and in terminal properties at range/altitude. Is it realistic? If it is, i have a rhetorical question to the engineer that designed it: "why, oh why did you gimp your own missile mate"?

2 hours ago, Exorcet said:

 

Are you hitting at 50 nmi? I find that 40 is just about max range against Ace AI that I'm testing against. I tried a few shots at 60 just now. 40k no loft, 40k 30 deg loft, and 50k no loft. Nothing hit, but the latter two were 1.5 nmi from target at 1000 knots while the first shot was 3.6.

Previous testing at 40 nmi isn't a 100% hit rate but that where the AI starts getting killed and I have had the most success getting as high as possible and firing level.

Depends on how willing i am do "meta" the AI. But, i used to get even more kills with less meta with the previous loft profile. Which again, begs the question, why would you program your missile to act in sub-optimal manner? 40 mile shots at angels 40 are not a good number for this missile. AI or not. Under the current circumstances, the missile is shoehorned to match an AMRAAM level terminal dynamics at given ranges, with much more inferior maneuvering properties due to extra mass and drag. Again, is this realistic? If it is, someone in the RnD messed up. Big time. 

BTW, not to be taken the wrong way, i do not advise loft assists. Yes, they do improve missile performance right now, but if (in our lifetimes) the lofting logic gets adjusted to previous values or if (even less likely to happen in our lifetimes) the missile gets switched to the new API, the end-result will be a lot of bad habits and messed up muscle memory. Not to mention, completely invalid time tables. Personally i just stay away from the missile and and missions that involve it, except for testing purposes. Stick to FOX-1 Cold War servers or dogfighting servers for now. Or if you are like me, practice CASE I's on Bankler's training mission (new personal best at 66 as of last weekend). But haven't went online in 3 weeks, nor do i plan to. And if the missile stays as it is, best to just shelf it. 

[RoyalWaffles]

I need help understanding something, SO I got on the GS server and decided "you know what, i havernt flown f14 in a good bit lets try it again" So i went up and being aware aim 54 is no longer the god missile i decided to climb to an altitude of 50000 feet to try to rattle off some aim 54s, i open fire with 4 C mk47s at high altitude (angels 30+) targets at around 60-70 miles Only to look at f6 and see my aim 54s.... not really lofting at all? Like they only went to an altitude of about 60000 feet and started a stupidly shallow dive towards said targets and lost all their speed before getting half way there, Did Hb change something AGAIN? where are those nice 45 degree lofts we used to have?

[Callsign JoNay]

25 minutes ago, NotAFlanker said:

Did Hb change something AGAIN?

Read the last few pages of this topic.

[WarthogOsl]

14 hours ago, Exorcet said:

Just looking at the images those lofts seem a bit shallower than mine, but if you have the max pitch angle of the missile that would give us a number to compare.

Went back and checked this.  Initial launch angle was actually only 20 degrees.  A second or two after launch the missile increased to around 27 degrees nose up.  It then gradually raised it's nose to a max angle of 44 degrees.

At 10 miles from the target, the missile was at Mach 2.54, and on impact at 27,000 feet, it was going Mach 1.65.

[Clunk1001]

It's been about a month since I played DCS.  I jumped into the Tomcat yesterday, and literally couldn't hit anything with the AIM54.

A month ago I was probably 90% hit rate and would be flying large campaigns against multiple groups of Mig29 Ace AI quite happily  (possibly a little too easy - but now it's gone the other way).

Here is a Tacview of what I'm generally experiencing now. 1 vs 6 Rookie Mig23.  30000ft, 30 miles, all six AIM54 miss.  (A month ago, I wouldn't have even played this scenario because it would be too easy, even with just 4 AIM54).

I have many simple training missions like this one, but even the simplest of mission now seems full of near invincible AI against the AIM54.  I can't go anywhere near my Liberation campaigns in a Tomcat.

Someone mentioned about 're-learning' to use the Pheonix.  Well, yes, I've found that I can use it like a Sparrow (7 miles wvr) very effectively.  But I don't seem to be able to use the Pheonix for what it was designed for, which means I can't use the Tomcat for what it was designed for. The Pheonix is an integral part of the Tomcat, if the Pheonix doesn't work the Tomcat doesn't work.

My questions are:

- In the holy quest for the grail of AIM54 flight modelling accuracy, is this just how the missile is going to be now?  It seems to me that the more work on AIM54 'accuracy' which is done the less 'realistic' the scenarios become in DCS.

- are there any tips for the current state of the missile? (please don't say higher and faster - I saw those successful Tacviews posted earlier, and the missile was designed to work below 48,000ft)

Thanks

Tacview-20221018-114227-DCS-test3.zip.acmi

[Clunk1001]

Here's a different example.  The bandits don't even react to the missiles at all, and the AIM54s just sail right past them.  (I'm deliberately not evading or touching the controls having fired the missiles in TWS, just to demonstrate the issue  -  so no tips on missile evasion/tactics thankyou ).

AIM54 Not Tracking.acmi

Edited October 18, 2022 by Clunk1001

[Exorcet]

6 hours ago, WarthogOsl said:

Went back and checked this.  Initial launch angle was actually only 20 degrees.  A second or two after launch the missile increased to around 27 degrees nose up.  It then gradually raised it's nose to a max angle of 44 degrees.

At 10 miles from the target, the missile was at Mach 2.54, and on impact at 27,000 feet, it was going Mach 1.65.

Thanks, so it's actually similar to my Phoenix (totally expected if we're on the same patch). 45ish degrees is about the most aggressive angle I'd want to see in the climb but I know it can go above that depending on launch angle.

[Callsign JoNay]

6 hours ago, Clunk1001 said:

Here's a different example.  The bandits don't even react to the missiles at all, and the AIM54s just sail right past them.  (I'm deliberately not evading or touching the controls having fired the missiles in TWS, just to demonstrate the issue  -  so no tips on missile evasion/tactics thankyou ).

AIM54 Not Tracking.acmi 312.01 kB · 2 downloads

 

Jeez, that's weird. You weren't kidding, no reaction at all and six misses. Can't say I've experienced that one before.

[Clunk1001]

It's okay everyone - I've worked out what seems to be a consistent method to engage a Mig23 with the Pheonix.

Sarcasm aside, no actually, wait, I have nothing left but sarcasm at the moment on this subject.

Come on, lets have some more posts about 'accuracy' and 'realism', we've only got 25 pages so far.....

Friends.acmi

[WarthogOsl]

On 10/18/2022 at 4:36 AM, Clunk1001 said:

- are there any tips for the current state of the missile? (please don't say higher and faster - I saw those successful Tacviews posted earlier, and the missile was designed to work below 48,000ft)

 

FWIW, I don't think faster is really necessary.  I typically try and launch between Mach .8-.9, but I still have success slower than that.  As for higher...well, I think that's unavoidable right now.  In the past year and half, I've steadily had to increase my launch hight from 25k, to 30k, then 35k, and finally 40/45k (I reserve the 45k for MiG-29's and Flankers).    I'm flying an F-14A, btw.

There was a recent F-14 pilot interview where he said "the Sparrow likes speed and the Phoenix likes altitude."  I've found that to be true in DCS.

[JabsDabbin]

I can't for the life of me complete* the marianas instant action mission hunting the jeff. W/o cranking the jf's sd-10 missile tends to hit just after the first aim54 goes active.

I probably killed maybe 5 JF's in a good 20-30 attempts(maybe even more lol). Missiles launched 40k ft alt at mach 1.2+, tried distances from earliest possible (+- 70nm in tws or 75nm pd stt lock, is this engagement 'hard locked' to only allow radar locks within this distance?) to 45nm but anything less than 60nm basically ensures the SD-10 will land first pushing me on the defensive.

 

* Almost completed it once but shot 3 of the 4 jf's WVR

Edited October 19, 2022 by JabsDabbin

[Clunk1001]

16 minutes ago, JabsDabbin said:

I can't for the life of me complete* the marianas instant action mission hunting the jeff. W/o cranking the jf's sd-10 missile tends to hit just after the first aim54 goes active.

I probably killed maybe 5 JF's in a good 20-30 attempts(maybe even more lol). Missiles launched 40k ft alt at mach 1.2+, tried distances from earliest possible (+- 70nm in tws or 75nm pd stt lock, is this engagement 'hard locked' to only allow radar locks within this distance?) to 45nm but anything less than 60nm basically ensures the SD-10 will land first pushing me on the defensive.

 

* Almost completed it once but shot 3 of the 4 jf's WVR

 

The Phoenix is totally phucked, phrankly.

some people here just wave an  ‘accurate Aim54 flight model’ flag, then posts some curve diagrams published by NASA, and ignore the obvious issues.

This is a DCS add-on.  If the Phoenix doesn’t work it DCS, then the Phoenix doesn’t work phull stop.

Edited October 19, 2022 by Clunk1001

[draconus]

10 minutes ago, JabsDabbin said:

W/o cranking the jf's sd-10 missile tends to hit just after the first aim54 goes active.

But you should crank for exactly that reason. You also confused km with nm. It's not a mission limitation but of the radar. Remember that you have a wingman and that it's not a fair fight. If your 54s miss you're basically forced to retreat and AI is pretty good at evading them.

[Clunk1001]

On 10/5/2022 at 7:50 PM, IronMike said:

Our job is to simulate as accurately and true to real life as humanly possible. Their "job" is to make the best of it, 

 

I have to disagree with this.

Your customers have purchased an add-on to DCS and it is not unreasonable to expect that product to work within the DCS environment.

As soon as you accept money from sale of a DCS add-on then delivering a DCS add-on is your job.  not just morally, but legally.

Whilst I appreciate difficulties with ED and the DCS environment, this does not detract from the obligation you have to provide something which works in DCS (and that includes AI missions, campaigns, and PvP)

I don’t agree with this “Accuracy over playability” attitude.

Edit: Sorry, this seems aggressive, I’m quiet passionate about it.  I think the work to-date on this module is phenomenal.  

Edited October 19, 2022 by Clunk1001

[near_blind]

30 minutes ago, Clunk1001 said:

I don’t agree with this “Accuracy over playability” attitude.

That’s antithetical to DCS’ whole spiel.

[Clunk1001]

3 minutes ago, near_blind said:

That’s antithetical to DCS’ whole spiel.

I don’t think it is.  
The simulation has to work.  The simulation has to have some resemblance to real life. Without playability you just have a pile of code and a pile of decals, but nothing to play.

[lunaticfringe]

On 10/18/2022 at 8:25 AM, Clunk1001 said:

Here's a different example.  The bandits don't even react to the missiles at all, and the AIM54s just sail right past them.  (I'm deliberately not evading or touching the controls having fired the missiles in TWS, just to demonstrate the issue  -  so no tips on missile evasion/tactics thankyou ).

AIM54 Not Tracking.acmi 312.01 kB · 7 downloads

 

You're taking 30 mile shots; as has been repeatedly mentioned, there's a "hold" region from the mid-30s down to 20 miles as expressed by RL pilots previously.  This range is forcing the missile to take a loft in a region that is too short for it to maintain valid acquisition at the top of its arc.  Compounding matters is the relative "floor" three of those MiG-23s descend to; net 15,000' elevation at 30 miles range and closing is too wide an elevation band for the AWG-9 to maintain all of these targets in TWS Auto. And as these are AIM-54As, there's no recovery of the shot if the track is dropped- because the missile doesn't know when to go active independently.

 

On 10/18/2022 at 7:36 AM, Clunk1001 said:

My questions are:

- In the holy quest for the grail of AIM54 flight modelling accuracy, is this just how the missile is going to be now?  It seems to me that the more work on AIM54 'accuracy' which is done the less 'realistic' the scenarios become in DCS.

- are there any tips for the current state of the missile? (please don't say higher and faster - I saw those successful Tacviews posted earlier, and the missile was designed to work below 48,000ft)

Thanks

Tacview-20221018-114227-DCS-test3.zip.acmi 425.8 kB · 5 downloads

 

Again, ~30  mile shots.  Only two out of six missiles had enough time and space to just get above Mach 3 because the range is too short; they're all respectively arriving at 5 miles range below Mach 2, and two of them were negated prior to going active at ten miles (your shots on Aerial 3-2 and 4-1) because they were turned behind the AWG-9 gimbal in your turn away.  Those two Floggers don't even so much as have to flinch because you've let their respective rounds arrive DOA.  Throw in the magic AI half loop on the other four to cut bait and run, and it's a geometry kill on all six shots.

 

15 minutes ago, Clunk1001 said:

I don’t think it is.  
The simulation has to work.  The simulation has to have some resemblance to real life. Without playability you just have a pile of code and a pile of decals, but nothing to play.

 

The simulation has resemblance to real life; that's why employment technique is critical.  You don't want to use it in accordance with real life terms, you're going to get similar results to real life with regards to out of parameter shots.  

[near_blind]

21 minutes ago, Clunk1001 said:

I don’t think it is.  
The simulation has to work.  The simulation has to have some resemblance to real life. Without playability you just have a pile of code and a pile of decals, but nothing to play.

Playability is subjective. Some people say the F-14 is unplayable because it doesn't have a pilot model for VR. Others say it's unplayable because it doesn't display air speed in the HUD. Others still because the AIM-54 is no longer a missile for all targets, in all seasons. 

You're shooting the oldest version of the missile with the least robust guidance capabilities in a situation where the greatest limiting factor (the radar) is going to struggle most and create the greatest potential for the missile to do a stupid. Lo, the missiles are doing a stupid. I can't find a real life example where an AIM-54 was used in such a situation, so I can't say with 100% confidence whether the sim is following reality or not. Given what I know about the AIM-54A, it looks about right. At that range, I'd opt for STT, or better yet, a Sparrow. Maybe even try the -54C

On the other hand, examples of the missile matching actual test shots have been provided somewhere beforehand. This is what DCS is supposed to do. You match the sim to performance you can prove, and refine if and when new information becomes available. 

[JabsDabbin]

1 hour ago, draconus said:

But you should crank for exactly that reason. You also confused km with nm. It's not a mission limitation but of the radar. Remember that you have a wingman and that it's not a fair fight. If your 54s miss you're basically forced to retreat and AI is pretty good at evading them.

I edited the units not sure if you responded to the initial post. But that's largely irrelevant in all honesty. The max locking distance raises eyebrows because I've been able to lock small stuff/helicopters further away. But i'm fairly new to this level of realism so a radar limitation was my second best guess. 

 

About cranking, I would agree that yes that is the most sensible thing to do but i was under the impression that cranking is a modern tactic which was not in widespread in use when the f14 entered service. Also realize that a missile with a much lower theoretical max range manages to out-speed the phoenix in this engagement and that just seems wrong. The JF's don't lock and fire until 45-55nm.

 

[WarthogOsl]

1 hour ago, JabsDabbin said:

I can't for the life of me complete* the marianas instant action mission hunting the jeff. W/o cranking the jf's sd-10 missile tends to hit just after the first aim54 goes active.

I probably killed maybe 5 JF's in a good 20-30 attempts(maybe even more lol). Missiles launched 40k ft alt at mach 1.2+, tried distances from earliest possible (+- 70nm in tws or 75nm pd stt lock, is this engagement 'hard locked' to only allow radar locks within this distance?) to 45nm but anything less than 60nm basically ensures the SD-10 will land first pushing me on the defensive.

 

* Almost completed it once but shot 3 of the 4 jf's WVR

 

I think HB has said this mission doesn't really work within the current state of things, and suggested editing it and turning down the skill level of the AI.  There's also some issue with the bandits being vertically stacked too close together.  I think they are going to revisit all the instant action missions, so I wouldn't feel too bad about it.  I've only completed the Jeff one once or twice out of MANY attempts.

3 minutes ago, JabsDabbin said:

About cranking, I would agree that yes that is the most sensible thing to do but i was under the impression that cranking is a modern tactic which was not in widespread in use when the f14 entered service.

 

Yeah, but the Jeff didn't exist in those times either.  I think they were cranking by the time Flankers and Flanker variants appeared.

[JabsDabbin]

7 minutes ago, WarthogOsl said:

I think HB has said this mission doesn't really work within the current state of things, and suggested editing it and turning down the skill level of the AI.  There's also some issue with the bandits being vertically stacked too close together.  I think they are going to revisit all the instant action missions, so I wouldn't feel too bad about it.  I've only completed the Jeff one once or twice out of MANY attempts.

Yeah, but the Jeff didn't exist in those times either.  I think they were cranking by the time Flankers and Flanker variants appeared.

Thank you, that would make the most sense, it's great practice in its current state though 

 

[draconus]

35 minutes ago, JabsDabbin said:

The max locking distance raises eyebrows because I've been able to lock small stuff/helicopters further away.

Show me.

The AWG-9 is arguably best radar in DCS range wise atm.

[Baco]

yes the "acuracy" people are simulating the rivets.. but the system does not work as it should. You can throw as many tech data as you want.  The simulation of the weapons systems: F 14-A + AIM 54c is not represented accurately. And that is the main fact. you can yadayada all you want... 

In teh world of Simulations numbers are not all that matters, since the experience is different than reality. The flight model of teh F 14 feels alive, i dont know it its morer accurate thna the viper o Hornet one, but it feels better... in the same manner, the Phoenix feels wrong. period.

[Clunk1001]

1 minute ago, Baco said:

yes the "acuracy" people are simulating the rivets.. but the system does not work as it should. You can throw as many tech data as you want.  The simulation of the weapons systems: F 14-A + AIM 54c is not represented accurately. And that is the main fact. you can yadayada all you want... 

In teh world of Simulations numbers are not all that matters, since the experience is different than reality. The flight model of teh F 14 feels alive, i dont know it its morer accurate thna the viper o Hornet one, but it feels better... in the same manner, the Phoenix feels wrong. period.

I agree.

if the Phoenix had so many limitations it wouldn’t have been strapped to an aircraft in the first place - and all you get here is “you fired from 31 miles - everyone knows it won’t work at 31 miles”, “you fired at 49 miles - everyone knows it won’t work at 49 miles”, “you fired at 20 miles, that will never work”, “you fired at m0.92 - this graph shows it won’t work at m0.92”, “you fired at an aircraft - everyone knows you can’t fire them at aircraft”, “you used the wrong version”,  “you need to fire them above 48000ft”.  “Higher”, “faster”, “the bandit farted … that throws off the aim54 - look at this NASA study on flatulence induced guidance mishaps”

sounds to me more like unrealistic limitations being applied to the missile to make the PvP experience more “playable”.  And if it’s not usable for AI and single player - blame it on the user.

[KlarSnow]

If you think it is unrealistic, find a source on the AIM-54 that proves the current implementation is incorrect and send it to heatblur. Saying you feel this or you feel that will not change anything without something to prove your assertion. Every change or implementation they have done has matched all available data about the missile quite well.

Since the initial thrust reduction patch in september there was another patch that has broken the lofting algorithm, this is known about and is being worked on. At this moment, today in Open Beta the missile is underperforming compared to as intended in many situations because the lofting algorithm is broken.

Everything else is as intended.The AIM-54 is one of the most complex missiles out there with a lot of limitations and issues. The AWG-9 is one of the most powerful but also weakest computationally radars out there. All of the issues you are running into are realistic to the extent of hard quantifiable data that is available about either system. If you have a source to disprove this then do so.