Ship vs ship... oh thou blind gunner!

[+Stary]

I was testing the beautiful new ships released recently (cheers WhiteBoySamurai!) and after setting simple test mission in ME, I'm shocked that three SumnerFRAM2 2 to 3 miles apart from Tarawa, CG-52 can't hit a thing. All rounds short, some 1/4th off the targets. I remember there was bug related to guns offset from ships CG, is it the case? I'm asking because the recent new naval addons, aswell as Tomcat on the horizont, are tempting see more sea action in the missions and campaigns by our fellow action gurus.

With ships unable to succesfully engage each other from such short (too short maybe?) distance it would be a bit hard to do some nice user made missions.

[fallenphoenix1986]

Funny you should mention this, I was thinking about the possibilities of player ship to ship engagements last night. Pretty sure I remember reading something about a group that had made a player vs computer tank engagement scenario, much the same idea I guess.

 

Also If the CG offset to the barrel is still an issue would it be possible to have the shell "fire" from the CG point but still have the muzzle flash effect emitted from the barrel?

 

Craig

[+WhiteBoySamurai]

Yeah, from what I understand, regardless of where the guns are on a ship, the AI "aims" from the center of the model. And unfortunately wherever we put the muzzle flash is where the rounds come out.

 

I thought at first that "ViewportPosition" for each gunner would make a difference, so I've tried placing them near the center of the ship, and I've also tried placing them right near the gun barrel, but I haven't noticed any measurable difference in gun performance either way.

 

Also, I hate to admit it, but I don't know exactly what the "RangeFinder=" and "BallisticComputer=" values do or what units they are. All I know is lowering them makes the guns miss even worse. >_>

 

Anyway, I can see why that bug would affect general accuracy, but I don't know why the rounds always seem to fall so short. Recently I tried a "shore bombardment" mission where I lined up a flotilla of USN ships and had them sail along the desert coast shooting at stationary targets. They couldn't hit jack squat at any reasonable ranges. Even the Iowa had all its rounds fall short until I moved it to just a couple miles from the targets.

 

If anyone has any ideas I'd love to hear them.

Edited June 25, 2011 by WhiteBoySamurai

[+CrazyhorseB34]

Same here! Put USS Iowa vs. two Udaloys a couple of miles apart and broadside.......results where a pathetic gunnery display. But, it looked cool flying over it!

 

 

[Fubar512]

A long time ago, I realized that errors in ballistic calculations cannot be completely solely attributable to where (on the model) those calculations are made from. How does that explain gross errors in ranging? And what about errors on weapons that were located adjacent to the model's centerpoint?

 

Some obsevations that may help.

 

1) Ranging has to do with the lifetime of the round. Right, Stary?

2) This series uses accurate (real world) ballistics tables....in a virtual world that's 66% scale! Of course you're going to have accuracy issues at range.

3) Gunfire is determined from the pivot points of the nodes (yaw & pitch) that one assigns to it...not from where the effect lies.

4) I suspect that the game engine aims visually groundobject to groundobject....that's why there's no over the horizon gunfire.

5)ASMs seem to work fine, as I proved many years ago, having set up ships that killed other ship targets that were well over the horizon. In that case, guidance type was the key, weapons that employed anti-radiation guidance were the only ones that worked for me (note that the target must be equipped with an active emitter).

[+JonathanRL]

 

They cant hit anything when there is more targets then sea either ;-)

[+Stary]

1) Ranging has to do with the lifetime of the round. Right, Stary?

2) This series uses accurate (real world) ballistics tables....in a virtual world that's 66% scale! Of course you're going to have accuracy issues at range.

 

 

These two points are similar to what I thought. But the rounds seem to be still alife in the simulation hitting the water. And setting them to say 0.1 of Mach speed would kill performance beyond playable point in most cases. Side note -Is it me, or in general bulletsdrop and deceleration ratios are too fast? I was able to outfly own bullets in a dive several times when testing Raptor, also with Lightning.

The ranging, hmmm I have a problem with that too, but thinking about it, IF all the units are real world, except for the map, that wouldn't be that of an issue. If it's the other way and all units are 66% then whole simulation thing is bit murky

 

 

@JonathanRL: same effect as with my tests from WW2 install

[+Stary]

anyone knows what armingtime= from bulletobject.ini does? Defaults to 5.0, seconds?

 

Setting MinBulletMach=0.2, will test

[Fubar512]

Side note -Is it me, or in general bulletsdrop and deceleration ratios are too fast? I was able to outfly own bullets in a dive several times when testing Raptor, also with Lightning.

 

 

LOL, that's happened more than once in real life:

 

"The F-11 Tiger is noted for being the first jet aircraft to shoot itself down. On 21 September 1956, during a test firing of its 20 mm (.79 in) cannons, pilot Tom Attridge fired two bursts mid-way through a shallow dive. As the velocity and trajectory of the cannon rounds decayed, they ultimately crossed paths with the Tiger as it continued its descent, disabling it and forcing Attridge to crash land the aircraft; he survived."

 

Source: http://en.wikipedia.org/wiki/Grumman_F-11_Tiger

[+Stary]

Yeah famous story. Makes Hot Shots planes (F-11) jokes dumb

 

anyway, still same results with 0.2 M

[+JonathanRL]

The Aircraft used in Hot Shots are Folland Gnats, are they not?

[+KnightWolf45]

Or Bae Hawks

[+Stary]

ah, my bad I think they're Gnats

[Slartibartfast]

Yes they are Gnats...

[Fubar512]

Stary,

 

OT, but you were right about the load that lowering the bullet mach imposes on the system. I lost over 30 FPS going from the default to Mach 0.2!

 

Of course, having 18 30mm gatling guns shooting at you at once doesn't help, either

 

Ballistics.wmv

[+Stary]

it's an old trick I started using in Baltika's BoB install -stock Mach value is too high for lower muzzle velocity guns, so hitting bombers from > 0.6 mile was somewhat impossible due to bullets being psyhically removed before reaching the target

 

but yes, unfortunately, the impact of framerate is showstopper

[+Spillone104]

If I remember right the diameter of the bullet influences the "resistance" and so the deceleration.

 

As for the Iowa battleship, I've tested it for over a year and tryied to calibrate the muzzle velocity of the 16" guns to permit the ships to actually engage targets with a reasonable accuracy to at least 15 Km away. but the 5" really shoots short.

[+WhiteBoySamurai]

I have a feeling Fubar is right about ground objects requiring line-of-sight to other ground objects. If that's true, though, it's a real shame for simulated artillery as well as big gun ships.

 

I wonder if there's a workaround, though, like placing the object's ViewportPosition up really high so it can see further over the horizon...?

[+Stary]

wouldn't that screw alot of other things? Also, maps in SF/SF2 are generally flat, no Earth curvature model exists so theorically two ships close to "global zero" height shouldn't have problems with each other's visibility / line of sight

Edited June 27, 2011 by Stary

[Fubar512]

VisualRange and maxvisible distance play a role here. Both are "virtual", in the sense that they will not effect the range at which a given 3d model is physically seen by the player, hence they don't effect frame rates. But they determine the range that the AI detects and takes action against, a given target.

[+Stary]

but having ships see/have knowledge of presence of each other at 10 miles still doesn't fix the pathetic gunnery skills at point blank range

[Fubar512]

but having ships see/have knowledge of presence of each other at 10 miles still doesn't fix the pathetic gunnery skills at point blank range

 

You're from Poland, a seafaring nation. Therefore, there's little excuse for you not knowing that sober sailors squids aren't worth a damn....we need to get a few drinks in them...that'll improve their aim

[+Stary]

d'oh... I live in mountains...

[+gerwin]

Old topic, but I want to add some new observations:

I am changing all naval guns to these parameters.

Caliber=10.000000
MuzzleVel=1000
Accuracy=50.000000

It may not be perfect or realistic, but at least with these values the ship vs ship engagements are interesting again: At close range the guns overshoot quite regularly, and at longer range the rounds fall short regularly. But in either situation there are regular hits as well.

Ships seem unable to sink another ship, but they can take out sensors and weapons.

One other important thing: For collision meshes to work the Max/MinExtentPosition of the "Hull" and "Superstructure" must envelop these actual parts. Parts of the ship that fall outside the the Max/MinExtentPosition cannot ever get hit. LodViewer allows one to check this nicely, but only when the Data.ini system-names for these parts match the node names of the 3D model.

Edited July 13, 2019 by gerwin

[+gerwin]

Here are some drop-in naval guns for the OBJECTS\GUNS\ folder, to try. Should cover most of SF2NA.

3IN_MK22, 5IN_MK12, 5IN_MK18, 5IN_MK19, 6IN_MK16, 76MM_AK726, 76MM_MK75, 100MM_AK100, 130MM_AK130

For the guns with an original caliber of around 3" or 76mm I lowered one value: MuzzleVel=950

 

SF2NA_Tweaked-Naval-Guns.zip