[Mraah]

Originally Posted by Redwine

Tested with a value of 10 instead 100....

Ships, small and bigs, stiull detected by early SD.

Planes, issues appears, in example, some of them seems to be not detected.

The Probability of this early SD is 85% , may be this has some influence, but 3 sucesive waves of G4M Betty, H6K, and AM6 Zeros, was not detected and pop-up at sighth range.

I will back to 100 and test a new time.

Aparently, this value should be a minimun surface to detect, if you put there 100000 nothing must to appears into the radar, but nothing works as predictable in this files...

Aparently it must be useful into SJ to prevent them to detect samll contacts as planes or life boats, but not sure.

Here into SD we need a Max surface limit....

Stll testing.

Originally Posted by ref

Hi Red, I think the problem with the radar not picking aircraft could be much simpler then that, I haven't played SH4 much, but a very important value everyone seems to be ignoring is surface (minimum square meters surface the radar will pick up), most planes don't have a surface of 100 m2., have you tryed reducing this value?, the ranges couldn't be that wrong if the original was picking ships at the published values, and the only thing that changed from 1.1 to 1.2 is the minheight.

Ref

Hiya RedWine and Ref,

I wondered what the actual surface area of A6M2 Zero was and I looked on the web and discovered this (to my amazement) ;
http://rwebs.net/avhistory/history/Zeke32.htm

EDIT NOTE (thanks Redwine!) .... divide my sqft by 10.75 below .... argh

1. Wing Surface Area = 232 sqft (70.71 m2) (21.58 m2)
2. Horizonal Tail Surface Area = 51.68 sqft (15.75 m2) (4.80 m2)
3. Fuselage Surface Area = 116 sqft ( 35.35 m2) (10.79 m2)... I couldn't find the surface area for the fuselage so I calculated it using the drawings and rounded down .. about half the wing area.
Total = 399.68 sqft (121.82 m2) (37.17 m2)... exposure of zeke directly above observer.

If you add in the frontal area of 58 sqft (17.67) (5.39 m2) ... again, my calculation based on drawing ... your total is 139.49 m2 (42.56 m2)

Using this calculation based on my TEST #1 below :

Zeke @ 500m altitude
1. Bottom Surface exposed = 2.88 m2
2. Front Surface exposed = 17.66 m2
Total = 20.54 m2 (1.91 m2)

Zeke @ 1000m altitude
1. Bottom Surface exposed = 5.76 m2
2. Front Surface exposed = 17.65 m2
Total = 23.41 m2 (2.17 m2)

EDIT NOTE (thanks Redwine!) .... divide my sqft by 10.75 above .... argh

Note how in the Aircraft CFG file for the A6M2 they have a Length and Width value (which is the same for G4M) of L=11.5 W=13.8 ... total 158.7 ... really close to what I calculated (139.49). I tried changing this to like 100x100 but appears it didn't change the detection result using a surface of 100. I might try this test again. I believe that it's actually using a polygon count or something instead of a basic box dimension "height x length" or "width x length" like from a flat source like the ship's dimensions or aircraft L x W. What you see is what ya got . I'm still lost ....

.... But ....

I've been playing with the Surface area value and I discovered it's not as simple as thinking a value of 100 means that it's won't detect anything below 100 m2. I think it's a "constant" that changes with range. If you increase range using a constant surface of 100 you also "push" the 100 trigger further out. Basically, you're pumping out more power (gain) and expecting a larger return of energy. Unfortunately, SH4 default values might not equal the result we want, ie set maxrange to real world range expecting real world detection.

I did some tests back in v1.1 with the SD with these results (the size of the A6M2 worked beautifully as a "base 100") ;

TEST #1
MaxRange = 55560
Surface = 100
Resultant Detection Range :
1. H8K / H6K @ 33,700 m +/- 500
2. G4M @ 26,500 m +/- 2000
3. Zero @ 21,100 m +/- 500

TEST # 2
MaxRange = 55560
Surface = 200
Resultant Detection Range :
1. H8k /H6K same as test #1
2. G4M same as test #1
3. Zero @ 14,600 +/- 300

TEST # 3
MaxRange = 111120
Surface = 100
Resultant Detection Range :
1. H8k /H6K same as test #1
2. G4M same as test #1
3. Zero @ 31,900 +/- 500

TEST # 4
MaxRange = 40000
Surface = 100
Resultant Detection Range :
1. H8k /H6K same as test #1
2. G4M same as test #1
3. Zero @ 14,600 +/- 300

As you can see, TEST #2 = TEST #4 , although max range was reduced about 15k. The larger aircraft aren't affect at all with a value of 100. I used a crazy value of 4500 in one test and only detected the H8K and H6K at about 9000 m.

Using the SD that Jace11 created along with Ducimus' modifications I believe we should increase the surface area to eliminate ship contacts since we've already have the coverage area pointing away from the horizon it's only exposing (my theory) the ship's superstructure and masts. However, increasing the surface will also decrease the range at which you detect the aircraft. It needs to be juggled (like Ducimus did in v1.1 ) but if you juggle too many values at one sitting you'll end up chasing your tail.

Before I embarked on my SD testing in v1.1 I asked myself these questions :

1. Is the SH4 sensor world flat? .... yes, I think so, even though visually it appears ships "grow" as they get closer (curvature of earth).

2. Are the sensor's dynamic? ... meaning, searching in a moving 3D world. Yes, Jace11 proved that with his "stormy sea" posting.

3. Are the sensors seeing through the water? ... thus including the part of the ship below the ocean when it calculates surface area .... I'm still debating this one!

4. And the BIG question ... What is the detection routine order for radar? ... Again, still debating this but Ducimus posted that at PreciseRange it's automatic. If an object meets the Surface requirement then it's detected, otherwise within PreciseRange.

Anyway, sorry to run off into a lecture ... . Please take no offense . I just got off a double mid-shift at work so I'm still wired from all the coffee!!! :rotfl:

Someone please correct me if I'm wrong!!! Or tell me to put the calculator away!!!

Picture example of my thought process ....
:hmm: