LuftWolf and Amizaur's Weapons and Sensors Realism Mod

[mike_espo]

Hey Luft! Thanks for the response. Hopefully next patch will correct Layer effects. 33% sounds about right..

Anyway, another thing: What freq should I use to change ohio to make more quiet? Im not sure what the values in DB editor mean...

thanks

[Amizaur]

Ehm, I think 50% torpedos was attracted by CMs and from those 100% detonated on them. So it's rather 50% * 100%

100% detonation of 50% successful torpedo attractions, so 1/2 of the times a CM was used against a torpedo it destroyed

in theory, because I did not statistic research on this phenomena , but I didn't ever see a torpedo that hit a CM and not detonated

[LuftWolf]

Amizaur,

Ah, I had a question about the CM torpedo detonation information. However, I had taken it as a relatively useful bootstrap, as I have seen plenty of torpedos sail past CM's at different depths, so just because a torp goes through the bearing of a CM doesn't mean a detonation, especially for subrocs, airdrops torps, and close shots.

In any case, I think the basic principle is the same, just disregard the exact percentages.

Thanks for the correction.

[LuftWolf]

Mike,

The DB value effects how loud or quiet an object is at given speeds is the Passive Sonar/Sound Level, or Passive SL.

It is listed 2/3 of the way down the list of object parameters on the right side of the screen.

[Amizaur]

Originally Posted by LuftWolf

Amizaur,

Ah, I had a question about the CM torpedo detonation information. However, I had taken it as a relatively useful bootstrap, as I have seen plenty of torpedos sail past CM's at different depths, so just because a torp goes through the bearing of a CM doesn't mean a detonation, especially for subrocs, airdrops torps, and close shots.

and I guess they would not detonate on submarine target if they run past it at different depth so I don't count such event as hit personally

[LuftWolf]

I don't either... I guess the 50% detonation value that has been reported is a misinterpretation of the 50% spoofing chance. :hmm:

[LuftWolf]

*Sounds of a chair hitting the floor*

ARE YOU KIDDING???

There are two types of AI sub passive sonar in the game HF passive and LF passive.

A sub will only have one of these two types.

Ok, are you ready? Unless I am reading this wrong, each array only has a 150 degree detection cone, mean ALL AI SUBS ARE BLIND for 210 degrees!?

In addition to that, both sensors are essentially useless anyway.

This is very problematic, even for me.

I think this counts as a high priority fix.

Can someone please tell me I'm wrong about this?

Edit: In looking at this more closely, it seems each array has the same geometry as the Sphere arrays on human controlled subs, that is to say, with a detection arc of 210 degrees and a dead zone of 150 degrees.

I think the best fix would be to give AI subs with towed arrays both a forward and rear facing sensor and ones without just a forward facing sensor, but I'm not sure if the AI could handle two sonars, plus the AI couldn't be made to make course changes required to actively employ the arrays. The most straightforward fix is to give AI subs with TA's a 360 degree sensor with more or less TA like properties, slightly reduced, assuming that this takes into account proper AI array behavior, and to give AI subs without towed arrays a sphere-like array that works for something as opposed to now where they are very very deaf.

This last fix has the advantage from a modeling standpoint of being minimally invasive, I would just make the HF passive sonar more effective by increasing its range, so it's more or less like the Kilo array (as these are typically on older subs anyway) and to change the LF passive sonar to more or less Pelamda TA properties, minus the restriction on cone. This may give the AI subs an advantage, but then again, we want that.

[LuftWolf]

Yes, I am feeling very good about a quick fix for this. It involves changing simple values for those of you who want to fix it now, and not wait for the full modelling treatment that I am now determined to give it for our next version.

I am going to change the "good" AI sonar to have geometry identical to the FFG towed array (the only other passive array modelled without a deadzone, I am using this model to avoid the potential bugs that could come from doing it another way) with sensitivity similiar to Pelamda and max speed near the TB-23 (I have to play with this to see what seems right, since this also effects ships for the time being).

As for the "bad" sonar, for subs without a TA, the geometry will stay the same, but I will make it essentially equal in sensitivity to the Kilo cylindrical array.

I will post the exact values and a instructions on how to make the changes via Ludgar's editor soon.

[LuftWolf]

I think I should mention that in testing the mod, we have found out that for some unknown reason, the ADCAPs run at 63kts.

If you look in the DB, the max speed is set to 60 as is the max on the fire control panel There is no good reason why this should be happening from a database or modding standpoint. The possible explanation available to us is that the physics engine is miscalcuating a thrust value due to the particular parameters (weight, drag, etc) of the ADCAP.

I don't mind, as its said to be able to go even faster for short distances, but I thought I'd let you guys know, as it's now a "known issue" with v1.03.

[Amizaur]

I think that AI subs should have second generic hull sonar added, they should use it without problems, just like AI Akulas or Kilos. Also some subs (like Trafalgar) could get generic towed array. I only don't know if AI subs actually deploy and use towed arrays :hmm:

For 63kts ADCAP - the issue is older than mod, because all faster than 55kts torpedos do some strange things, for example Type-40 torpedo reaches 72-73kts sometimes... their database speed is 65kts.
Have to see what will Japanese Type-89 do, have to provoke a Harusio to fire at me

[LuftWolf]

***Actually ignore this, if you've come late. Amizaur did a thorough speed test and the engine is bugged for all torps from 58-159kts.***

Ok guys, from the strange but true file.


The game engine is hardcoded to treat the max speed of the ADCAP at 55kts for calculation purposes. Raising the max speed of the weapon to something above 55kts, in effect, tells the engine to compute the physics at 55kts, and then add whatever the value the max speed is above 55kts to *any calculation*, up to about 3-4kts, based on early tests. So, when you tell the weapon to run at a particular speed, the game calculates that speed, and then adds the value above max speed to the torpedo calculations.

I have tried to run at max speed 60kts, and for any setting, we get about 4 kts above, with max 63kts. When I changed it to 58kts, I got a max speed of 61kts, and an over of about 3kts. If the max speed is set to 56kts, then the over will be 1 kts and the max speed will be 56kts, where the two effects would intersect on a graph.

This is unfortunate and unexpected. I think Ive had an interesting day of discoveries. I smell DARPA's hand in this.

In all seriousness, why would SCS hardcode the algorithms to 55kts for the ADCAP? since that is the only explanation for this regular but odd behavior...

Edit: I just read Amizaur's post, it could be a bug in the physics engine, but this is very particular behavior.

[Amizaur]

I think it's not ADCAP-only issue, it's common for all torpedos. Make another AI used torpedo 60kts and see how fast it will go...

Just verified that some AI subs ARE COMPLETLY DEAF to the front !!!
That would explain some stories about ramming a Han without detection.

For example Han, Harusio and Ohio are deaf in front 60 degrees zone.
The deaf zone it's not 210 degrees, I forgot that 150 degrees cone of AI sensor means 150 each side, so efectively 300 degrees. But that still leaves 60deg deaf zone in front of many AI subs...

[LuftWolf]

Does anyone have any idea what max and min "Beam width" do in the sensor profile?

That is the one attribute that makes me very hestitant to change cone geometries in any arbitrary way. :hmm:

[Amizaur]

Well, I don't know unfortunately... tried to change (increase) it to simulate worse angular resultion of towed array in front-rear axis, expected wider contact trace on broadband, but the result was target track scattered for whole sonar display :-(

SCS says it support modders, then some kind of modding documentation would be nice... all we know about database entries is what Ludger and jsteed have guessed.

[LuftWolf]

Oops, Amizaur corrected me. The cones are actually double the value listed in the DB, around the axis of symetry.

So the AI platforms with only one sensor are only 60 degrees blind. That's not so bad, but still... the sensors are extremely weak, as it is, and should be corrected.