RA mod ADCAP problem

[Ghost Dog]

Hi all, I'm noticing a problem with ADCAPS while using the RA mod.

my install was this:

Dangerous Waters
1.04 update
RA mod
RA DWX Update


I notice that Active homing is almost useless with mk 48s, my fish never seem to track a submerged target. they seem to kill surface ships just fine. is this a known bug or am I doing something wrong with these torps?

[LoBlo]

Hm... its been awhile since I tried the RA mod, but I think that I remember this problem. I never figured it out and so just swapped to the old doctrines.

Anyone else had that problem?

[Castout]

Try launchig your ADCAP at lower speed like 40 knots to 45 knots max.

The rotating head of the ADCAP may miss acquiring a target if it travels too fast. Not to worry about launching the ADCAP at lower speeds because when the ADCAP is locking into something it will go to top speed.

[Ghost Dog]

I did some experimenting with preset speeds on the ADCAP. They will now home in on a submerged target BUT they are extremely 'short sighted'. ie: they will only home in when within roughly 500 m. sometimes they wont even 'see' the active countermeasures. unless your TMA is DEAD ON, you have no real chance of guiding these fish onto target without cheating.

passive homing seems to work great, although I always thought that active homing was the prefered method against enemy subs.


do the RA guys intend to fix the ADCAPS? (taking a close look at the database files I notice the Spearfish uses the same sensors and might have this problem as well).

[Castout]

It's not because they are short sighted on active it's just that the rotating head needs time to scan all the areas and it may acquire target rather late if your preset speed is too high.

I too complained about this to GrayOwl and he explained it to me and once I reduced the ADCAP launch speed my ADCAP is deadly as ever.
The problem is in RA the ADCAP plays by a different rule than in stock or LWAMI. in RA the ADCAP does have a rotating head just like in RL. So you need to chnage your mindset instead of blaming the mod.

If you still don't like the slower launching speed you may want to adjust the rotating speed of the ADCAP. I fiddled with it though at the end I realized the current setting to be the most realistic because th head needs to scan(rotate), ping out then wait for a ping return before rotating to new direction and the amount of time that it waits will determine the range at which the ADCAP able to acquire target. You can find underwater sound speed table in the internet.

It's just that imo the ADCAP in RA should be made smarter once they stop acquiring. Such that they start to scan at the most immediate area where the target was lost. so that they will re-acquire target much faster.

[Bellman]

There seems to be a problem in ASW mode which is depth related. If the target sub clears the torps 'active' beam in the horizontal plane, the torp slows and changes depth immediately. The 'lock' is broken quite easily and the resulting depth disparity causes serious reattack problems.

In gameplay terms this is a killer for manual TMA players and even testing with 'Truth' on leads to repeated reattacks.

[Castout]

Originally Posted by Bellman

There seems to be a problem in ASW mode which is depth related. If the target sub clears the torps 'active' beam in the horizontal plane, the torp slows and changes depth immediately. The 'lock' is broken quite easily and the resulting depth disparity causes serious reattack problems.

In gameplay terms this is a killer for manual TMA players and even testing with 'Truth' on leads to repeated reattacks.

You may want to tell that to GrayOwl via PM or to any of RA modder.

[LoBlo]

Originally Posted by Castout

It's not because they are short sighted on active it's just that the rotating head needs time to scan all the areas and it may acquire target rather late if your preset speed is too high.

There's still something fishy going on. I had a passive and active torp simultaneously acquire a target, then both simultaneously loose the target and then both needed the target at 45% AoB before requiring simultaneously...

... all that points to some doctrine level bugs in the behaviour. Problem is that the doctrine variables are so complex now its hard to decipher. Maybe GrayOwl would be willing to rename his variable to make them better understood, or provide some annonation within the doctrine file itself.

[Castout]

Originally Posted by LoBlo

There's still something fishy going on. I had a passive and active torp simultaneously acquire a target, then both simultaneously loose the target and then both needed the target at 45% AoB before requiring simultaneously...

... all that points to some doctrine level bugs in the behaviour. Problem is that the doctrine variables are so complex now its hard to decipher. Maybe GrayOwl would be willing to rename his variable to make them better understood, or provide some annonation within the doctrine file itself.

Hmm I myself haven't heard from him for quite some time.

You could try PMing him. He's a nice person.

[Theta Sigma]

Is he still active in the RA community? I don't see his name attached to any of the recent posts on Sukhoi or Red Rodgers.

[LoBlo]

On the other forums I think he goes by CrazyIvan. I'm sure he's around, but I don't think he actively post on this forum anymore because someone attempted to get him in trouble for modding the .dll files.

Anyhow, more play testing as in order anyway. I tried speeding up the sensor rotation and seems better now. And while we are on the subject, seems like the passive sensor are too good on the ADCAP now 6000 meter detection range at 45 knots on loud targets. Seems like the flow noise would be a little more prohibitive. Maybe 4500 meters at 35-40 knots seems more feasible, but I'm pulling those numbers out of thin air. Does anyone have any good open source info?

[LoBlo]

Originally Posted by Black_Dingo

passive homing seems to work great, although I always thought that active homing was the prefered method against enemy subs.

I've been play testing. The short detection range against subs is a matter of a downgraded NRD in the sensor dialog (roughly estimated as a sensor's sensitivy to noise). Basically the stock torp seekers had a NRD of about -14 where the RA ADCAP active sonar is about -8, so its just a matter of the sensor not as esily detecting the subs now. If you want to adjust it, change the NRD in the sensor dialog. A -12 gives roughly a 1nm homing range on a Kilo.

On a broader note on the ADCAP behavior, I've been looking hard into the RA's team's ADCAP doctrine. I have to say its a marvelous piece of work and its obvious it took hours of testing. Bravo to the RA team. However... there seems to be some fundamental flaws in how the doctrine files handles its contacts/sensors. Not by any fault of the RA team, but simply how DW handles its variables.

First off the way the doctrine is constructed the Mk48 "scans" with a sensor head back and forth (its really blinking conical sensors on and off in a sequential fashion). When a target is detected to attack it changes its status to "Lock", stops scannings, turns on its attack sonar, and homes in. If the torpedo looses a contact ("LostTrack"), it changes back to "Not Locked", turns off the attack sensor, and continues scanning (blinking its sensors on and off).

The problem arises in that the torp's "Lost Track" applies to *all* contacts held by the torp (if you look at a debugview after an attack you will see that torpedos routinely holds multiple contacts at the same time, but only attacks one). And if it looses *any* contact, not just the one being homed on, then its status is changed to "Not Locked", and the sensor blinking starts again.

All this adds up to the bizarre behavior I was noticing in testing where the torpedo would be homing and then all of a sudden loose the contact and quit homing, never to reattack. The torp was essentially changing to "not locked" via the lost contact bug (when it looses another contact it was holding) and starts scanning again. Sometimes it would re-attack, but a lot of times it won't (I think its because the attacked ship is no longer a "NewTrack" and won't initialize the homing triggers, but can't prove it).

How to fix this?... problem is I'm not sure. You have to be able to tell the torp to go to "not locked" and start scans whenever the track being attacked is losttrack, but not go to "not locked" if tracks not being attacked are losttrack. I tried changing the losttrack and not locked triggers to the torphoming doctrine, but it gets just as buggy.

The ultimate solution I'm not sure of. Anyone got any ideas?

[Bellman]

Good work LoBlo.

As the AI sub defeats the attacking torp by a simple hard horizontal turn and the torp immediately changes depth, some delay before this change and rescanning ability is required. Have seen a more recent Beta doctrine but the weakness persists.

This behaviour is obviously only a secondary indicator - its solution is more difficult. However the RA team are to be congratulated on many other subtle changes to torp search and cm performance.

You may wish to PM GrayOwl, a nice guy, heavily involved in the team. I am sure you can contribute !

PS. UGST perform similarly (or did) but Spearfish not tested......
Having too much fun with Silent Hunter currently - 100% reality is the target.(See Dls.)

[Bellman]

CrazyIvan gives the RA rationale for planned changes to torpedo doctrine:-

http://www.redrodgers.com/forums/sho...?t=4912&page=6

[LoBlo]

I believe I've found a potential solution to the "sensor blinking off" inappropriately bug, but still am having trouble with the reattack logic (the torps just don't reattack after passing a CM well.

Since the RA Mod team is hard at work perfecting the logic anyway, I think I suspect that they are already aware of the behavior, so I'll leave it to them.