[FIXED] JF-17 radar gunsight solution incorrect

[some1]

It is not possible to hit a target using the radar sight, even a non evading one. All bullets fly behind the target.

 

I set up a mission with AI set to a constant, level, 3g turn. I followed the target with a stable tracking solution, zeroed closure rate, textbook conditions, yet despite that all bullets miss.

 

 

17.trk

Edited December 3, 2021 by uboats

[J20Stronk]

Scuttlebutt is Deka may rewrite their entire Gun Sight code from scratch since the solution provided by ED isn't working for them, as you can see.

 

Apparently the real JF-17 has been upgraded with EEGS similar to the F-16, so we may get that as well if DIS fixes its code and gets documentation.

Edited April 9, 2021 by J20Stronk

[some1]

In DCS F16 and F18 you can snipe the target from 1km, so ED managed to get it right at least in some of their aircraft. Well, that's after a few patches, initially they were also bad. So maybe ED haven't shared all the latest tricks with 3rd parties yet. 

Edited April 9, 2021 by some1

[PetRock]

On 4/9/2021 at 3:17 PM, some1 said:

In DCS F16 and F18 you can snipe the target from 1km, so ED managed to get it right at least in some of their aircraft. Well, that's after a few patches, initially they were also bad. So maybe ED haven't shared all the latest tricks with 3rd parties yet. 

 

 

Didn't you also post that there is a problem with the F-5 gunsight lead calculation as well? Is there a bug that has been introduced in 2.7?

[some1]

All the testing was done in 2.5.6. 

[zitron]

Any update on this? I have to go full WWII eyeball mode to get any hits with the gun, the sight seems completely useless?

[Coole28]

It’s been mentioned before elsewhere, but the radar gunsight implemented in the Jf-17 (at least the earlier versions of the plane, like we have in DCS) doesn’t have the fancy all-aspect lead computing gunsight like the F-15/16/18, but rather has the simple distance/drop lead computation similar to older planes like the F-5.  
 

What this means for the Jf-17 is that the gun solution provided by the radar does NOT account for the direction and speed the target is moving, but rather only accounts for how far away the target is and where it currently is in relation to you.  You must manually account for the direction and speed the target is moving and lead accordingly.  As far as I know, this is modeled correctly for the Jf-17 we have.

 

The general procedure for longer range (more than 0.1-0.2nm) gunnery is:

1.) Set gunsight to SSLC (this is important, the other gunsights are for close range only)

2.) Line up the historic hit point line with the plane of motion the target is traveling (i.e. the gun sight line should match the movement of the target)

3.) Place the gun piper in front of the target based on the enemies distance, relative speed, and turn rate.  This is essentially a judgement call, and takes practice to get right.

4.) Squeeze trigger, realize you forgot to feed the cannon and eject due to frustration.

 

The gun piper by itself (LCOS sight) is only accurate at extremely close ranges, where the flight time of the shells is short enough that enemy movement doesn’t matter.  You really shouldn’t be using the cannon anyway unless you are too close to use your IR missiles, but at those ranges the piper is generally accurate.  Making longer range gun shots are a last resort for if you are out of missiles and breaking contact is not an option.

[some1]

Nothing that you wrote above matters in the scenario described in the first post. The target is stationary relative to the shooter, the only thing the gunsight needs is the target distance from the radar to calculate the bullet time of flight. 

Even a WWII gyro gunsight will work perfectly in this scenario, and in fact they do if you try the same in DCS P-51 or P-47. 

[Coole28]

2 hours ago, some1 said:

Nothing that you wrote above matters in the scenario described in the first post. The target is stationary relative to the shooter, the only thing the gunsight needs is the target distance from the radar to calculate the bullet time of flight. 

Even a WWII gyro gunsight will work perfectly in this scenario, and in fact they do if you try the same in DCS P-51 or P-47. 

The actual situation posted:
Jf-17 in 3G turn, Su-27 in 3G turn. Same relative motion.

The calculated gunsight solution:

Jf-17 in 3G turn, Su-27 stable in 1G.  Different relative motion.

This is how the Jf-17 calculates a gun solution.  It’s rudimentary, in fact it’s less advanced than even the F-86 radar sight.  But you aren’t supposed to be using the cannon in aerial combat except as a last ditch Hail Mary tactic.  As such the engineers didn’t waste time on a lead computation system. 

[some1]

48 minutes ago, Coole28 said:

Jf-17 in 3G turn, Su-27 stable in 1G.  Different relative motion.

This is how the Jf-17 calculates a gun solution.  It’s rudimentary, in fact it’s less advanced than even the F-86 radar sight.  

Thats highly improbable, because a computing gunsight working in such scenario is much more complex than the one which assumes in-plane motion. So your explanation using JF17 simplicity is invalid.

If they really made something simple, it would work in the scenario I described. That's why WWII gunsights work, no computers there. And the Sabre sight works on the same principle, it just uses radar to solve the distance problem, instead of stadiametric rangefinding.

Edited November 9, 2021 by some1

[uboats]

this will be reviewed in Dec

Edited November 10, 2021 by uboats

[uboats]

fixed

https://www.bilibili.com/video/BV1Xh411s7Ah/

Edited December 3, 2021 by uboats

[J20Stronk]

14 minutes ago, uboats said:

fixed

[zitron]

Great news! Thank you!

[Sajarov]

Great!!!

Now we will see if ED can coordinate properly with third parties the release of the next patch so no fixes are left behind ( like the last one).

Sent from my M2102J20SG using Tapatalk

[DragonJockey]

uboat, that's good news. It would have been great if Deka could include EEGS gunsight. PAF Thunders have been using EEGS in all blocks including Bravo. Any plan to update with EEGS? The gunsight is holding off Jeff in getting popular in Gunz-only BFM events. This update will be great service to Jeff pilots. Thanks,

[some1]

On 12/3/2021 at 5:12 AM, uboats said:

fixed

I know it's a bit late, but thanks. And congratulations. Of all the developers that had this issue reported in their aircraft at the same time (ED, Heatblur, Razbam, Leatherneck), to this day you're the only team that fixed it.