F-14 radar gunsight solution incorrect

[some1]

It is not possible to hit a steadily turning target using the radar sight. All bullets fly past the target.

 

I set up a mission with AI set to a constant, level, 3g turn. I followed the target with a tracking solution, zeroed closure rate, yet despite that all bullets miss. No such issues with F18, F16 or M2000. 

 

Note that the flashing BATR diamond appears in a different place than the pipper. In a steady tracking solution, they should overlap. 

Begin firing:

BATR flashing behind the target (roughly where the bullets actually went)

 

I recorded a track, but, as usual, it's unusable. Though you can still take control of the aircraft and try for yourself. 

14.trk

Edited April 8 by some1

[captain_dalan]

That is a very weird gun track

[Kula66]

I’ve noticed this too ... in a hard turning shot you have to lead the target despite a radar solution.

[Snappy]

Same,

I noticed this as well, its more obvious with the target at high aspect . With the pipper directly on aircraft

centermass / middle of enemy plane and in the lower part of the gun range, the bullets tend to pass aft of the target.

 

Of course you can aim in front of the aircraft or just on the tip of its nose to make sure you get hits, but would be nice

if heatblur could take a  look at it.

KInd regards,

Snappy.

 

 

 

Edited Monday at 08:29 AM by Snappy

[Victory205]

That's awesome. HB has the level of detail realistic on the gunsight. The radar lock would wander across the skin of the target, causing ranging errors, affecting the solution. The bullet at TOF cue would also behave exactly as you described. 

 

You are a half a mile away, get closer if you want to hit anything.

[Snappy]

2 hours ago, Victory205 said:

That's awesome. HB has the level of detail realistic on the gunsight. The radar lock would wander across the skin of the target, causing ranging errors, affecting the solution. The bullet at TOF cue would also behave exactly as you described. 

 

You are a half a mile away, get closer if you want to hit anything.

 

Yes in the tacview screenshot he is half a mile away , but in his HUD screenshots he is already very close, he should get the Breakaway X if he gets any closer?

Or did that X have more of an advisory character in real life and people flew in closer to make sure the bullets hit the target? Serious question, not meant sarcastically .

 

Regards,

Snappy

 

 

 

 

Edited Wednesday at 05:14 AM by Snappy

[some1]

4 hours ago, Victory205 said:

That's awesome. HB has the level of detail realistic on the gunsight. The radar lock would wander across the skin of the target, causing ranging errors, affecting the solution. The bullet at TOF cue would also behave exactly as you described. 

 It doesn't wander. It just shows you incorrect place in the sky, 100% of the time (not enough lead). And BATR cue always shows lagging behind the pipper. Like the computer can predict actual bullets path, but it shows you pipper in a different place nonetheless, and switches to correct solution only after you fire the bullets.

 

4 hours ago, Victory205 said:

You are a half a mile away, get closer if you want to hit anything.

Anything closer than 0.4 nm (2400ft) and you get a huge flashing breakaway X on the HUD. I agree that's quite far for a gun shot, but hey, I just follow the limits shown on the screens.

 

Anyway, if you get closer and aim for the canopy, you'll hit his engines at best.

Edited Wednesday at 06:44 AM by some1

[Skysurfer]

I think what is Victoryis saying is that the real thing isn't pin-point accurate either and has some errors in it. I would expect that when turning and loading up with G's. Personally I never had an issue with it, just take some lead and walk the bullets across the plane, tip to butt. 

[some1]

I'm all for inaccuracy if that's what the real thing is doing, but currently what we have in DCS F-14 looks to be "accurately inaccurate".

 

[Naquaii]

We have had a little bit of a discussion about the breakaway and concluded that we probably set it to 0.4NM due to video footage as we couldn't find a better number. After rummaging through my documents we concluded that 1000ft is probably a better number and co-incides with the inner cue.

 

We'll have a look at the gunsight accuracy, it's not something we've noticed before.

[Victory205]

This is going to take a bit of explanation on the gunsight operation and modes. Might be better to put in in the normal section of the forum when I get a chance to write it up?

 

Short answer is that the BATR worked exactly as you see in the sim. It was often in a different place than where the pipper was when the trigger was pulled, depending upon a host of factors. The radar "scintillates" along the airframe when close to a large target, and the dynamics of the shooter aircraft have large effects on the bullet stream prediction error, especially at longer ranges given the design of the gun sight, the mode employed and the fact that the BATR wasn't the holy grail in terms of accuracy. It wasn't meant to be, it was a training aid for snapshots. 

 

One thing you all need to get a grasp on- no aircraft system is as precise as you seem to expect from growing up in a digital world. 

 

Things like the breakaway X were advisory, there is no minimum range for a 20mm projectile. When you stopped shooting was based on the threat of collision or damage from the target destruction. For missiles, we memorized LARS for the weapons and used them according to the latest data feedback, which wasn't programmed into the AWG9. When I was the Training Officer, we put the "LAR of the Day" in the flight schedule and someone from each flight was selected randomly to draw it up from memory accordingly, including threat LARS. 

 

It's difficult to get hits with a gun at range, especially with a 50 rd burst limiter on a maneuvering target. DCS seems too easy to me. Another aspect is that we didn't carry gun load outs with tracers. You could see the vapor trails or aberrations from the bullet stream at times, but while we exercised the gun systems routinely, I don't ever recall having tracers on board.

 

Fill your windscreen with your enemy as close as you dare and pull the trigger. Press the target.

[some1]

Thanks for the write up. I understand what you're saying, and although the DCS simulation does not currently behave like you describe, one can argue that the end result is the same, and that means you cannot consistently snipe targets from afar. 

[Victory205]

You can't in real life either. I would suspect, from what you are saying about other modules, that it is probably too easy in those, based on perception.

 

Are you aware that with an STT, that the F14 gunsight is "historical"? It shows the bullet impact, one time of flight ago. 

[some1]

I've been playing the same scenario with various DCS aircraft, some are accurate out of the box, the rest like F-14 consistently undershoot. In these aircraft it's just a matter of adding extra lead to what the gunsight is showing. That's why it seems suspicious, as it looks more like an error in bullets ballistic calculations, than anything else. 

 

Flying in circles with non-manoeuvring target, it shouldn't really matter if the gunsight is historical or not. Relative to the shooter, the target is in the same place as it was one time of flight ago, or will be one time of flight in the future. It's just a matter of taking the range from the radar, angular rates from the gyros (or INS), and calculating where the bullets would be after travelling that distance. 

[Golo]

Might be computational delay. If you take measured data needed you have to process them and then draw solution on HUD. it might take some time (microseconds?), but by that time target will move little bit, so you are shooting short. Unless that delay is taken in account already or is insignificant.    

[some1]

This data does not change over time in a scenario I described here. Distance is constant. Angular rates are constant. Speed, altitude, everything is constant.

 

Of course taking a shot at an uncooperative, manoeuvring enemy fighter is more art than science, but that's a different story.

This one here is just a bit of math, as long as your sensors are accurate.

[Victory205]

Depends. You are barking up the right tree in that a stabilized tracking scenario is best for, well, any gunsight system, but it isn’t quite that simple. Any closure, or opening also induces error, as does atmospheric changes, like a decent or climbing shot, or any tiny perturbations in flight path for any reason.

 

Did you know that the impact point is different between the 6000 rpm and the 4000 rpm firing rates? Fun stuff.

[AH_Solid_Snake]

Not to shut down future discussion on the gunsight itself - its fascinating. But it sounds like from Victory205's various responses this could be marked as No-Bug? The underlying simulation might not precisely reflect reality but the results seem to tally with what the SME expects to see?

 

I guess when the pipper is on but your rounds are going behind the target a real pilot just shrugs and pulls a little lead - a sim pilot immediately questions the nature of his artificial reality 

[Snappy]

6 minutes ago, AH_Solid_Snake said:

Not to shut down future discussion on the gunsight itself - its fascinating. But it sounds like from Victory205's various responses this could be marked as No-Bug? The underlying simulation might not precisely reflect reality but the results seem to tally with what the SME expects to see?

 

I guess when the pipper is on but your rounds are going behind the target a real pilot just shrugs and pulls a little lead - a sim pilot immediately questions the nature of his artificial reality 

 

Best to leave that decision to heatblur.Naquaii already said they will look into it.
Then they will reach their own conclusions.

 

 

 

[Slant]

10 hours ago, Victory205 said:

Depends. You are barking up the right tree in that a stabilized tracking scenario is best for, well, any gunsight system, but it isn’t quite that simple. Any closure, or opening also induces error, as does atmospheric changes, like a decent or climbing shot, or any tiny perturbations in flight path for any reason.

 

Did you know that the impact point is different between the 6000 rpm and the 4000 rpm firing rates? Fun stuff.

 

I am looking forward to a write up if you ever decide to do it. I find this discussion very interesting. One question I had for a while, why is the BATR flashing in the first place? Is it losing radar lock? Is it because of a stream of bullets giving all kinds of wrong radar returns? Or are they too tiny to register on the radar?

Edited 20 hours ago by Slant

[Snappy]

2 hours ago, Slant said:

 

I am looking forward to a write up if you ever decide to do it. I find this discussion very interesting. One question I had for a while, why is the BATR flashing in the first place? Is it losing radar lock? Is it because of a stream of bullets giving all kinds of wrong radar returns? Or are they too tiny to register on the radar?

 

I would guess the flashing is on purpose, to clearly differentiate it from the actual HUD target symbol, which is the same icon ,but non- flashing.

 

Even more so, because I think in reality the HUD target symbol is not always directly centered on the actual position of the enemy aircraft in your HUD FoV,  unlike it is usually in DCS.

Probably another thing less precise in reality than in digital simulation. If you look up various HUD footage tapes on Youtube, you can see the HUD projected target symbol is sometimes outside of the enemy aircraft or lagging somewhat behind.
 

Another reason could be, that it flashes as it re-calculates after every x no . of fired rounds during a longer burst, but that would seem pretty advanced.

Certainly Victory  knows much better if thats the reason or something else.

 

Regards,

Snappy

Edited 17 hours ago by Snappy

[Slant]

I understand the RL limitations. It's an amazing feat to project an image onto the hud and have that line up with a real world object at all, even if everything is straight and level non-maneuvering. Add ACM to that and it's a miracle the computer dealing with HUD symbology isn't just throwing its hands in the air and telling you good luck, you may or may not find the target if you look near the symbology cos that bastard keeps moving around and so do you, by the way...