[FIXED] DMT targeting losing targets

[Brainfreeze]

Just this week, the DMT has suddenly become near unusable for no apparent reason 

 

Let say I designate a target for the DMT just by pressing DESG on the EHSD screen (no tdc slew or action) ... I fly around the target, not touching anything and when I face it again the DMT/TV is looking somewhere else like 500m away from the designated target.  I slew it back to the target, tdc action down once on it... fly around and when I face it again it is miles away again.   

 

Driving me crazy.   Anyone had that?    BTW TDC action / no action is not ticked in options and nothing binded to TDC action/no action toggle 

[myHelljumper]

Could you make a track showing the behavior you described please ?

[Brainfreeze]

Ok so attached 2 tracks.  

1 track (DMT bug): 

1- select DMT/TV  press DESG on EHSD -> go around.  DMT still pointing ok

2- Slew to a truck, TDC action down -> go around. DMT pointing 500m bottom right of truck. 

3- Slew back to truck, WP inc long -> go around. DMT pointing again 500m away from truck.

 

Did a second track with similar sequence because watching track 1 it seemed on the 3 pass the DMT was still actually pointing at the truck...which was definitely not the case when I recorded it. ???  

DMT bug2.trk DMT bug.trk

[Aernov]

Earlier in development this occured constantly and was attributed to INS drift during high-G manoeuvres messing up DMT target point location (even if IFA and GPS coupling was active). I haven't noticed this behaviour recently, with IFA and POS/GPS enabled, in correct time periods and coalitions for GPS to work.

[Brainfreeze]

2 hours ago, Aernov said:

Earlier in development this occured constantly and was attributed to INS drift during high-G manoeuvres messing up DMT target point location (even if IFA and GPS coupling was active). I haven't noticed this behaviour recently, with IFA and POS/GPS enabled, in correct time periods and coalitions for GPS to work.

We got back in time again    

[Aernov]

AV-8B wpn sel, DMT drift.trk

Yep, got this behaviour too. Also, this track has Mk. 82 AIR selection bug, when those are placed on outermost pylons, they can only be selected separately from the same bombs on other pylons, and top row of OSBs will be occupied by two instances of MK82s in two modes and "gun", so, if you have another kind of weapon onboard, it can only be selected with station buttons on armament control panel. Selection bug is reported in separate topic.

[Hornet81]

I haven't had any issue with that!

[myHelljumper]

Hi, thanks for the tracks.

 

I watched it and I was able to see the problem but as I'm really not knowledgeable on the Harrier so I asked our testers to take a look at it.

 

They think it might be a problem on your side with the latest TDC action/no action changes.

 

Do you have the TDC action/no action enabled in the special options for the Harrier ?

Edited May 7, 2021 by myHelljumper

[Aernov]

Tracks were made before TDC action/no action was enabled for DMT.

[Hornet81]

 

[Brainfreeze]

1 hour ago, myHelljumper said:

Hi, thanks for the tracks.

 

I watched it and I was able to see the problem but as I'm really not knowledgeable on the Harrier so I asked our testers to take a look at it.

 

They think it might be a problem on your side with the latest TDC action/no action changes.

 

Do you have the TDC action/no action enabled in the special options for the Harrier ?

 

Nope. The option is not ticked under the menu 'special'

[Hornet81]

@BrainfreezeI do not have any issue with DMT. Everything is perfectly ground stabalized.
 I was checking your tracks and seems the WPT you designated was not over any truck since the begining. So the DMT stayed in the same initial position.
In your mission editor try to place a WPT over a target  (zoom all IN), DESG again and make sure that becomes target point in your HUD ( Distance to target ). You will see that after some turns the DMT is still pointing the target either if you are in TV or INS mode.     

[Brainfreeze]

only the first pass is targeting the wpt (ground). Not that this should matter at all.  We are able to bomb the ground too.  

anyway, initially I deliberately did not slew to show that DMT was then stable.  Only once i slewed it would it wander around and lose the target while the plane is turning around.

[myHelljumper]

One of you tester might have found the source of the issue, can you guys test this after adding a big deadzone on the center of your TDC axis (he tested with 8%)?

 

It would look like the problem comes from a false movement on the TDC combined to the very high sensibility of the INS mode.

[Fri13]

The DMT shouldn't be able to be locked on a INS coordinates. It is a sensor responsible to lock on contrast on target and then deliver the changes in the rate that ARBS system reads and then generates the coordinates to the INS system for the target. 

 

DMT shouldn't be possible be slewed on the ground or be left on the ground. If you do not get a contrast lock, then the DMT will not stay on the ground but will move with the aircraft.

If you didn't get a valid lock or the TV jumped on something else (while a moving target tracking etc) then you needed quickly to switch to backup aiming modes (not implemented in Harrier). Once you come for a re-attack, you want to have the coordinates generated from the previous attack as you do not get the TV on the ground there but you use INS diamond. 

 

 

 

 

 

[Fri13]

4 hours ago, myHelljumper said:

One of you tester might have found the source of the issue, can you guys test this after adding a big deadzone on the center of your TDC axis (he tested with 8%)?

 

It would look like the problem comes from a false movement on the TDC combined to the very high sensibility of the INS mode.

 

No effect.

Still happening.

The TDC can be even set for buttons instead Axis and it happens.

 

00:00-00:20 Target acquired and locked to overfly

02:04 Steady turn back to target

02:13 TV back in gimbal limits, designation elsewhere by hundreds of meters (in totally different forest even)

02:43 overfly of the target. 

 

I made this twice and the exact same thing happened, TD shifted to same position even when the route was different. Difference is that other is with the DESG and other is without. 

 

 

 

There is no false movement on the TDC axis for the INS mode. As the INS mode shouldn't even be moving the TV or any other target information. The INS mode is usable only when the system is in INS mode and then target diamond can be moved only when it is inside the HUD. Once the TD is outside the HUD field of view it doesn't anymore slew around anything. The INS mode is only for the visual targeting using HUD and correcting the Target Diamond visually on the target while not flying straight at it, so that ARBS system can calculate the slant range from the corrections using aircraft round speed and barometric altitude. In the Harrier the INS mode is "master of all sensors" so you can slew any sensor and you can do so when the target designation is outside of your HUD (so even behind you or so). 

 

There is as well a bug report about Harrier TDC ignoring the axis deadzone values as it reads the input directly from the XInput. Can't find it now as valid bug reports gets moved around among resolved.  But it would need to be fixed first if it is still valid so that TDC follows the deadzone values. 

 

INS_lost_coordinates.miz.trk

Edited May 7, 2021 by Fri13

[Fri13]

Decided to check my videos today for how the Target Designation (TD) shifts/moves. And found these things. 

 

In the first video (Jump2) look carefully the target designation diamond in HSDI at 01:30-01:34 when level from left banking. 

The diamond jumps toward the aircraft. 

 

 

 

The exact same thing happens in the DMT Jump video but at 02:42-02:45 position.

 

 

Again the same behavior, everything is fine until left side turn is leveled and the TD at the 90 degree jumps toward the plane.

 

It is clearly not a INS drift as it would mean few things:

 

1) The aircraft position would be required to shift as much as the error is. This is impossible with the tightly coupled GPS/INS. 

2) The INS drift would happen gradually over long time (like an hour). Not by a smooth turn. 

3) The Target Designation has the coordinates for the target, either as a generated map coordinates or then as relative position to the aircraft by INS.

 

The INS system works wrong in the Harrier, hypothesis is INS is tied to the coordination system somehow relative to the Harrier attitude. So when the Harrier attitude axis moves across negative and positive angles, the target designations gets "dragged" or shifted with it. 

 

There is this similar problem with the TDC axis, that using a axis to move TDC across the axis cause stop of one axis. Again when moving from positive to negative side.

 

It might be related what happens with the TPOD when the aircraft axis moves across the TPOD gimbal position (that Elmo refuse to move in bug reports)

 

The DMT has as well other problems. Its movement is tied to the polar coordinates instead own gimbal speed. So when you have TV at long distance then it will slew faster, than if you have TV at close distance.

The TV should move at the two speeds moving with TDC DOWN or TDC. The simple TDC is slow one for "sweeting" or getting sensor accurately on position. And then TDC DOWN is the fast mode where sensor is moved quickly for long distances. They both should be with fixed degrees per second rates, regardless the distance. IIRC The TV should be slewed inside the HUD with TDC and have it stabilized only to the HUD and not to the ground, until TDC DOWN is released. The primary method is to have TV locked to the FPM and fly it on the target and then press and release TDC DOWN to try to acquire a contrast lock on TV. If it success then it will lock on the contrast. If it doesn't then it will stay locked to FPM. The secondary option is to use TDC to move the TV box in the HUD on the target and then press and release TDC Down to try to acquire a contrast lock. If the contrast lock is not achieved then the TV box will jump back to FPM.  

 

The Polar coordinate control system is such that further from the plane the TD is, then faster it moves. Closer you have it, then slower it moves. This makes example moving a MAP DESIGNATOR challenging as you can't move it in time at 5 nmi range to other position as it speed slows down considerably. And this doesn't work in the map mode when it zooms in/out as the speed is somewhat relative to the display size, so at low zoom level the TD movement is slow at long range as is at short range. So moving TD example from 50 nmi to 70 nmi is at same speed as moving it at 1 nmi to 2 nmi. 

 

Hypothesis is as there is many things wrong across various places and sensors and their behavior and so on, and all are somehow tied to others causing problems in each others. 

 

This is not about TDC deadzone or "jittery pot" or anything in the gaming device. It is directly in the programming of the target system and the logic how sensors and TD are moved. 

 

Edit: As additional point, is the TD suppose to move outside of the HSI compass ring? I have always wondered that why the TD is allowed to move outside of it to areas where are OSB labels and all other symbols and get "behind them". Same thing is with the aircraft symbol when target designation is centered. Like shouldn't the TD and aircraft symbols stay inside the compass ring and if they move further than the selected distance and AUTO scale is selected then scale would be switched to one bigger one to keep these symbols inside compass ring?

Now it is done when the symbol reaches the MFCD edge and you can lose the sight where they are as they get mixed somewhere. 

Edited May 8, 2021 by Fri13

[Brainfreeze]

Tried again today both in the TDC action / no action mode enabled and disabled.   I have the giro disabling ins option ticked on (no idea what this does) but tried with both ticked and unticked just in case... still getting this target jump which happens when the screen switches from INS back to TV

 

So I did the same dive bombing (30deg Z profile) using the TPOD this time... Same thing happened... When I roll to dive the tpod loses sight for a few seconds and when it gains sight again it is completely off the locked target.  

Edited May 8, 2021 by Brainfreeze

[myHelljumper]

Hi guys, the dev and testers worked on this issue and it look like they fixed at least part of the problem.

 

The problem we fixed was with the DMT camera "drifting" when gimbal locked.

 

We are still looking at this to find if there is another problem with this.

Edited May 9, 2021 by myHelljumper

[Brainfreeze]

6 minutes ago, myHelljumper said:

Hi guys, the dev and testers worked on this issue and it look like they fixed at least part of the problem.

 

The problem we fixed was with the DMT camera "drifting" when gimbal locked.

 

We are still looking at this to find if there is another problem with this.

 

[Schmidtfire]

If it helps for reference, there is live footage in this thread along with some screenshots:
 

 

[Fri13]

2 hours ago, Schmidtfire said:

If it helps for reference, there is live footage in this thread along with some screenshots:
 

 

That video shows well how the TV tracks a high contrast (the end of sandwall).

 

Then on the moment when the original tracked position is covered by the CBU bomblet smoke, the TV jumps to right to the brightest contrast and keeps a lock there.

 

And TV eventually loses the lock and it returns to a acquisition mode (wide open cross) and jumps to the FPM up in the sky.

 

 

[Fri13]

23 hours ago, Brainfreeze said:

Tried again today both in the TDC action / no action mode enabled and disabled.   I have the giro disabling ins option ticked on (no idea what this does) but tried with both ticked and unticked just in case... still getting this target jump which happens when the screen switches from INS back to TV

 

Look carefully your EHSD that what happens on the Target Designation diamond. 

Check my two videos above that does same thing happen to you. As the TD should jump position even before TV mode comes On.

 

(I am not even sure should the TV mode return back On in re-attack as it doesn't know where target is and where to try to get a contrast lock.) 

 

The GYRO mode in navigation panel IIRC  resets the INS alignment and so on aircraft position information. It is backup mode to get a basic attitude information going. And you can not go through it to other mode without performing that reset. If you do, then you need to perform alignment again.

 

23 hours ago, Brainfreeze said:

So I did the same dive bombing (30deg Z profile) using the TPOD this time... Same thing happened... When I roll to dive the tpod loses sight for a few seconds and when it gains sight again it is completely off the locked target.  

 

With a TPOD you need to be in the INR mode (SSS Aft long) whenever TPOD is to be masked. As Point and Area tracking are contrast based and they don't care about INS or coordinates. Their job is only to generate a target coordination by having a contrast lock on target. So they do not follow target slant range and vector like the Inertia mode does.

 

The targeting pod should have a masking map that will perform automatic laser shut off when laser is emitting and it reach 5 degrees from the mask border (this was at least in 2nd gen pod) and disallow laser firing when masked. Why you want to learn to set TPOD on INR mode everytime you are estimating that target gets masked (terrain, smoke, trees, clouds, airframe etc).

 

 

[myHelljumper]

Hey guys, we think we nailed the issue with the "DMT camera drift when gimbal locked" fix.  It should come out with the next update, if you guys could report back on this after the update is out that would be great :).

[Brainfreeze]

On 5/10/2021 at 2:27 PM, myHelljumper said:

Hey guys, we think we nailed the issue with the "DMT camera drift when gimbal locked" fix.  It should come out with the next update, if you guys could report back on this after the update is out that would be great :).

Not an extensive test series but have had a couple of tries and it seems to have been fixed indeed.