AGM-114L and you: Why shouting "I cast magic missile!" doesn't work

[Floyd1212]

Another video showing what is happening in a busy MP environment.  I hop to the front seat and fire on a group of units that are relatively out in the open, from a hover.

I have also been noticing two explosions happening in slightly different locations, and sometimes with the timing off, where one explosion will happen and then you see the missile enter the screen from above and a second explosion happens.

 

[Hobel]

vor 2 Stunden schrieb Floyd1212:

Another video showing what is happening in a busy MP environment.  I hop to the front seat and fire on a group of units that are relatively out in the open, from a hover.

I have also been noticing two explosions happening in slightly different locations, and sometimes with the timing off, where one explosion will happen and then you see the missile enter the screen from above and a second explosion happens.

 

would be interesting to see what your INU says

[skypickle]

@NeedzWD40 please clarify this:

"LOBL is the longest range mode due to the nature of moving targets, and will be utilized when targets are moving or attempted when the range is less than 2.5km. Even though LOAL might be indicated <2.5km,"

So LOBL is used for targets <2.5 km away. So why would you say 'even tho LOAL might be indicated'

[Floyd1212]

3 hours ago, Hobel said:

would be interesting to see what your INU says

Another video, also with poor results.

Fired 8 missiles and shacked 4(?) targets at under 5km. I checked INU alignment several times during the engagement and it showed .011, .012, .012, and .003.

 

[NeedzWD40]

7 hours ago, Floyd1212 said:

Another video showing what is happening in a busy MP environment.  I hop to the front seat and fire on a group of units that are relatively out in the open, from a hover.

I have also been noticing two explosions happening in slightly different locations, and sometimes with the timing off, where one explosion will happen and then you see the missile enter the screen from above and a second explosion happens.

 

Firstly: Thanks for taking the time to make some videos. It's helped me try to replicate and theorize what might be happening.

I can say the two explosion thing seems to happen on loaded servers or when there's heavy desync as you can frequently see this when SAMs self-destruct or hit a target. The double explosion might also be a fragmentation script.

I see this a lot and I'm pretty sure it's not intended behavior. A friend and I did an attack with 114Ls on a group of vehicles out in the open and we had similar results, where the intended targets weren't being hit. What seemed to be happening was the vehicles on the outer edge of the perimeter were taking the first hits, with the missiles seemingly grabbing the first thing they saw. For us, the order didn't matter as we were RTB and trying to expend unused ordnance, so it was just an improvised test to see what would happen.

1 hour ago, Floyd1212 said:

Another video, also with poor results.

Fired 8 missiles and shacked 4(?) targets at under 5km. I checked INU alignment several times during the engagement and it showed .011, .012, .012, and .003.

 

I've seen this a few times, too, where the missile is having to climb up and can't really look down as easily. I'm uncertain if it's intended behavior or not, I only know that 114Ls struggle in climb-up situations at this point.

After some fiddling around myself, my theory at this point is that there is an offset happening within the INU that's causing wrong info to get passed. My reasoning is if you check the self coordinates (with the TSD PP function) and cross reference those numbers with the info bar, there is frequently an offset of ~50 meters in the northing axis and ~150-250 in the easting axis. This only seems to happen in multiplayer; if you check singleplayer, the self coordinates are dead-on. The offset in multiplayer doesn't seem to disappear, even after a reset of the INUs. I think it's related to how if you use the STORE function with the TADS you don't get exact coordinate results when you use that point as an acquisition source. The test I'd like to try is storing a target point on a target, slaving to that point, and lasing it to see if the 114L goes for that point or actually goes for the target.

1 hour ago, skypickle said:

@NeedzWD40 please clarify this:

"LOBL is the longest range mode due to the nature of moving targets, and will be utilized when targets are moving or attempted when the range is less than 2.5km. Even though LOAL might be indicated <2.5km,"

So LOBL is used for targets <2.5 km away. So why would you say 'even tho LOAL might be indicated'

Because as soon as the range is <2.5km, the missile will instantly go terminal. So even if it indicates "LOAL" it's immediately going LOBL after launch.

[VKing]

6 hours ago, NeedzWD40 said:

Because as soon as the range is <2.5km, the missile will instantly go terminal. So even if it indicates "LOAL" it's immediately going LOBL after launch.

I think the terms are getting a bit muddied here. The missile can't be LOBL if it acquires a target after launch.

[Poptart]

 

Quote

"LOBL is the longest range mode due to the nature of moving targets, and will be utilized when targets are moving or attempted when the range is less than 2.5km. Even though LOAL might be indicated <2.5km,"

I think the stress is on the word attempted here. At greater than 2.5k on a stationary target LOBL will not be attempted. Even if you have LOAL indications in the cockpit, the missile will attempt LOBL if the data indicates a moving target or range<2.5k. This is why you often see LOAL symbology briefly before the missile establishes a LOBL track. 

[NeedzWD40]

10 hours ago, VKing said:

I think the terms are getting a bit muddied here. The missile can't be LOBL if it acquires a target after launch.

Just how it was explained to me many moons ago. <2.5km is always implied LOBL regardless of indication. For us, it basically just means that we'll not get a DBS trajectory and if we don't get an explicit MSL TRACK before launch, there's a chance it will grab the wrong target.

[QuiGon]

On 5/30/2023 at 5:24 AM, NeedzWD40 said:

I see this a lot and I'm pretty sure it's not intended behavior. A friend and I did an attack with 114Ls on a group of vehicles out in the open and we had similar results, where the intended targets weren't being hit. What seemed to be happening was the vehicles on the outer edge of the perimeter were taking the first hits, with the missiles seemingly grabbing the first thing they saw. For us, the order didn't matter as we were RTB and trying to expend unused ordnance, so it was just an improvised test to see what would happen.

I've seen this a few times, too, where the missile is having to climb up and can't really look down as easily. I'm uncertain if it's intended behavior or not, I only know that 114Ls struggle in climb-up situations at this point.

After some fiddling around myself, my theory at this point is that there is an offset happening within the INU that's causing wrong info to get passed. My reasoning is if you check the self coordinates (with the TSD PP function) and cross reference those numbers with the info bar, there is frequently an offset of ~50 meters in the northing axis and ~150-250 in the easting axis. This only seems to happen in multiplayer; if you check singleplayer, the self coordinates are dead-on. The offset in multiplayer doesn't seem to disappear, even after a reset of the INUs. I think it's related to how if you use the STORE function with the TADS you don't get exact coordinate results when you use that point as an acquisition source. The test I'd like to try is storing a target point on a target, slaving to that point, and lasing it to see if the 114L goes for that point or actually goes for the target.

Good info, thanks!
Also thanks @Floyd1212 for the videos!

[skypickle]

On 5/30/2023 at 12:11 PM, Poptart said:

 

I think the stress is on the word attempted here. At greater than 2.5k on a stationary target LOBL will not be attempted. Even if you have LOAL indications in the cockpit, the missile will attempt LOBL if the data indicates a moving target or range<2.5k. This is why you often see LOAL symbology briefly before the missile establishes a LOBL track. 

this clause is what is confusing : "Even if you have LOAL indications in the cockpit,"

Why would you have LOAL indications in the cockpit if <2.5k or a moving target?

[Nerdwing]

Im glad Im not the only one with this issue!  It being an INU discrepency in MP definitely also explains the significant displacement I see in Stored target locations I slave to as Gunner.

Is this officially reported as a bug?  Ive outright stopped using Longbows in MP due to their unreliability.

 

EDIT: Also the double explosion is probably the fragmentation script if its run on your server.

Edited June 3, 2023 by Nerdwing

[Floyd1212]

Some more testing from earlier. 

This time I performed a lase and store, then slaved to the stored point.  (Sorry for me fumbling around trying to figure out what happened as the slaved point was way off to the right of the target and I didn't realize where I was looking.)  Then I fired a Lima at the stored point on the ground, and it missed even further to the right.

Then I stored another point off to the left, and fired at that, then I tried hitting 3 vehicles and the missiles all went somewhere else.

I'm happy to make more videos if they are helpful, but otherwise I will stick to laser-guided missiles to avoid the frustrations.  Looking forward to seeing them work properly in a future patch...

[GrEaSeLiTeNiN]

1 hour ago, Floyd1212 said:

Some more testing from earlier. 

This time I performed a lase and store, then slaved to the stored point.  (Sorry for me fumbling around trying to figure out what happened as the slaved point was way off to the right of the target and I didn't realize where I was looking.)  Then I fired a Lima at the stored point on the ground, and it missed even further to the right.

Then I stored another point off to the left, and fired at that, then I tried hitting 3 vehicles and the missiles all went somewhere else

I'm happy to make more videos if they are helpful, but otherwise I will stick to laser-guided missiles to avoid the frustrations.  Looking forward to seeing them work properly in a future patch...

It’s the trees. Try targets in treeless areas and the hit rate should improve. Don't know about MP but that’s been my experience in SP.

Edited June 2, 2023 by GrEaSeLiTeNiN

[Moxica]

On 6/3/2023 at 12:52 AM, GrEaSeLiTeNiN said:

It’s the trees. Try targets in treeless areas and the hit rate should improve. Don't know about MP but that’s been my experience in SP.

 

I've had this happening to me on several missions with no trees between me and targets.
Why would it go that far to the right because of trees?

On 6/3/2023 at 12:33 AM, Floyd1212 said:

Some more testing from earlier. 

This time I performed a lase and store, then slaved to the stored point.  (Sorry for me fumbling around trying to figure out what happened as the slaved point was way off to the right of the target and I didn't realize where I was looking.)  Then I fired a Lima at the stored point on the ground, and it missed even further to the right.

Then I stored another point off to the left, and fired at that, then I tried hitting 3 vehicles and the missiles all went somewhere else.

I'm happy to make more videos if they are helpful, but otherwise I will stick to laser-guided missiles to avoid the frustrations.  Looking forward to seeing them work properly in a future patch...

Phew! What a relief to find I'm not alone experiencing this!
-In single player that is.

Edited June 6, 2023 by Moxica

[Nerdwing]

Didnt see anything in patch about this.  Has it been officially reported?

[NeedzWD40]

On 6/8/2023 at 2:32 PM, Nerdwing said:

Didnt see anything in patch about this.  Has it been officially reported?

According to the changelog, these are the changes to AGM-114L:

• Added for AGM-114L  new icon for ME
• Fixed: AGM-114L seem to fail to hit targets fired while aircraft is moving
• Fixed: Player-CPG cannot launch AGM-114L after AGM-114K
• Fixed: Yaw limit after you don't launch AGM-114L at once
• Fixed: George launched AGM-114L from 9km when in weapons free mode    
• Fixed: AGM-114 typo in weapon label

I haven't had a chance to test any changes since the patch and my last MP mission prior had no discernible variances from previous experiences.

[FalcoGer]

Didn't test the new patch yet, but the issues described don't really match with our experience with the missile hitting the same offset spot over and over. If there happens to be a target somewhere near that offset spot, it goes for that, otherwise it just hits the ground, not even a unit. I think this is the closest one to our issues:

INU display didn't show misalignment, but on a hunch I had reset it anyway and that had helped

Also that double explosion never happened. Might be a script running on the server. Most of the misses happened on hoggit gaw.

From what I can tell, the missiles are using the INU, which they shouldn't. And even if they did, all the offsets would cancel out anyway, which they don't. And the INU drifts way too much, considering that it's getting GPS updates constantly, using a modern laser ring gyro and using ground doppler radar for keeping track of the position accurately. Modern airliners with the same laser ring gyro + GPS tech get maybe a few tens of meters of drift over 8h flights. And they don't even have the doppler radar feature. The apache getting a hundred meters drift after 20 minutes of flying seems ridiculous.

Edited June 12, 2023 by FalcoGer

[FalcoGer]

On 5/30/2023 at 5:24 AM, NeedzWD40 said:

going LOBL after launch

Going lock on before launch after launch? That is silly talk!

[backspace340]

I wonder if this isn't an INS issue but a desync issue - there's a bug report for LMC desync (where pilot and cpg have different views of where the TADs is pointing after the CPG uses LMC) - if the target coords takes the pilot's view of where the TADs is looking (instead of the CPG's), that might explain why the target points or hellfires are hundreds of metres off target. 

https://forum.dcs.world/topic/314716-los-reticle-desync-between-pilot-and-cpg-when-using-lmc/#comment-5236896

Edited June 13, 2023 by backspace340

[Floyd1212]

I see where you are going with that idea, but in all of my videos I posted I am flying solo in a MP server and switching seats. Is there such a thing as desync between the PLT and CPG seats when not in a multi-crew setting? 

[Nerdwing]

Flying MP tonight, I'm getting big swings in INU accuracy from 0.009 up to 0.034.  Is this enough to induce the kind of weirdness we are seeing in MP with the Lima and its target selection?  It seems like its definitely enough to make storing target coordinates less useful.

[McGraw]

Surely the L's current best feature is being able to be fired and forgot? Pop up, fire, run away

[HoneyFox]

I have a question about the trajectory of AGM-114L in LOAL mode. It makes sense to fly a little bit sideway to make the active-radar sensor "feel more comfortable" detecting ground targets. But the missile still keeps this off-boresight angle when it's reaching target pretty close, which often causes difficulty at the terminal phase of the trajectory. (shown as the first case in the picture below)

My question is: since the missile should be able to lock on target when it's close enough (just like how LOBL mode works), why doesn't the missile stop flying side-way and go straight towards the target so that it will have a more straight terminal trajectory? (shown as the second case in the picture below)

Edited June 30, 2023 by HoneyFox

[NeedzWD40]

8 hours ago, HoneyFox said:

why doesn't the missile stop flying side-way and go straight towards the target so that it will have a more straight terminal trajectory?

My understanding is that it should be this way and that the current mechanism is either an oversight, WIP, or the developers have access better data than my recollections and research. One side effect of the current behavior I recently noticed was that it is extremely easy to knock out targets in revetments, as the missile has a tendency to circle around to the rear of the vehicle rather than the front.

[FalcoGer]

On 6/13/2023 at 3:04 AM, backspace340 said:

I wonder if this isn't an INS issue but a desync issue - there's a bug report for LMC desync (where pilot and cpg have different views of where the TADs is pointing after the CPG uses LMC) - if the target coords takes the pilot's view of where the TADs is looking (instead of the CPG's), that might explain why the target points or hellfires are hundreds of metres off target. 

https://forum.dcs.world/topic/314716-los-reticle-desync-between-pilot-and-cpg-when-using-lmc/#comment-5236896

 

from what I can tell, the missile goes after what the CPG shoots at, even if there is a desync in the TADS display. I've had shots hit with desynced TADS and shots miss with everything being proper.

Also I don't think the INS confidence has anything to do with the actual drift, I believe it's just random values that simply change once alignment is done, but otherwise don't mean anything. INS reset and then waiting always seems to help, no matter what the displayed confidence is.

INS should not drift as much with a modern laser ring gyro set and constant GPS corrections. It's constantly correcting and keeping track with the doppler ground radar.

Also INS shouldn't be apart in any of this, but it somehow is. The missile should get the direction from the TADS angle and laser range, build it's DBS picture of the area, acquire and then track the point. And even if it were INS related, it would have to have it's own independent INS, not being fed any information from the helicopter about it's current position, then the helicopter would tell it to shoot at some coordinates (which are wrong, due to drife), but the missile is properly aligned while the helicopter is not. Then the missile would go correctly (according to it's own, perfect INS) to the wrong coordinates (from the bad helicopter sent coordinates) and miss. That's the only scenario i can think of where that could possibly happen.

But I don't think the missile has a full INS on board, an accelerometer at best, to know the difference in locations after being launched. Could be wrong though.

Also with INS misaligned, when the TADS is slaved to a point, the cued line of sight indicator (the dashed cross) is not centered in the display. The TADS is pointing offset according to the bad INS data, but the indicator somehow knows the exact position of the point. And that's when the hellfires miss. I don't feel like making a bug report with the next update being around the corner, which might fix this or break it in a different way, but this is a complicated mess.

Edited October 13, 2023 by FalcoGer