New Astronavigation Mod (ANM) – A Conception

[don1reed]

Both SH3 and SH4 have MapFiles. In SH3 it's located:

C:/Documents/SH3/data/cfg/Careers/skipper's name/0/_____.Map

In SH4 it's located:

C:/Documents/SH4/date/cfg/SaveGames/00000000/SaveData.Map

There isn't a "Map File" for SH5. (That I know of.)

An amateur mathematician, Celestial Navigation Enthusiast and practicianer; not to mention, an all-around nice guy, Michael Jones, created a script file using Python and Stellarium to take the data from the SH4 Map file and convert the numbers found therein into Geo Coords. Numbers such as:
-1198199.13, 6006413.50,0.00

by dividing these numbers by 120000.

i.e., -1198199.13 / 120000 = -9.9848993. The (-) sign indicates West Longitude and South Latitude. The quocient then becomes: 9°59.1'W.

6006413.50 / 120000 = 50°3.21'N.

Everytime a "save" is done on your computer while playing, it auto-updates your true position on the game chart, even if your boat's icon is hidden.

Alas, SH5 does not have a MAP file that I know of.

And, of course, the Python/Stellarium method of celnav was used because the celestial canopies in SH3 and SH4 are not accurate, whereas, Stellarium's canopy is true.

Once Python obtained your true coordinates, it pulls up Stellarium and presents the correct celestial canopy you would be viewing if you were in fact, at that location. All that was left to do was to take of round of virtual sights of altitude and azimuths of selected bodies and have Nories, or HO Pubs 211, or 229, or 249 and almanac at hand to do the sight-reduction...five times per day, as in real life.

Then again, the next problem was the matter of the Mercator Projection Chart. None are true in the SH games. They're charts meant to fit the resolution of your display. So...forget Dead Reckoning as well.

When Navigators purchase charts from authorized chandlers, the charts they buy are printed to very precise specification and are miniature ratios of the real world, i.e., 1:1,000,000 scale. This is an extremely precise ratio, which means it is accurate to the length of one Toyota Prius to 1M Prius, or 1 nm to 1M nm, or 1cm to 1M cm, etc. One could say it's etched in stone. The chart in these games...not so much.

Edit: I'm looking forward to see this come about and wish you great success.

Cheers,

[Nico71]

After so many attempts to get real nav to work in SH by the community - with variing success - I'm inclined to give up on it. I did it successfully in FSX, with sextant, weather ships and whatnot. That being said, I really hope that you get this to work! That'd be really awesome!

[Sartoris]

Sounds like there are some problems to overcome, but man, it would be so amazing if you could pull this off!

[silent marshal]

Hi folks!

I made some fine tuning with the calculation background (thank you for your critical comments!) and it seems that most of the problems are solved.

Two examples:

Exact Position SH5 (SP1): N 64,385° E 3,296°
Travelling at 135° 60km from T1 (first shot of the sun) to T2 (second shot).
Exact Position SH5 at T2 (SP2): N 64,032° E 3,664°
Gegißter Punkt (= estimated position G1): N 64,080° E 2,200°

By drawing the Exact Position at T2 I got (SP): N 64,025° E 03,65° (compare with SP2).

Second example:
SP1: N 33,247° W -17,253°
Travelling at 210° 40km
SP2: N 33,01886° W -17,465°
G1: N 33,75° W -16,75°
Graphical solution leads to SP: N 33,01° W -17,463°

There are still some things to check but the results seem to be ok. If you don't change your position from T1 to T2 (who doesn't?) the results are more accurate. Aberration is a result of drawing in plane geometry in spite of the curvature of the earth. The greater the distance between T1 and T2 is the greater the aberration would be (but it's not the only factor).

[silent marshal]

How to find your own position – Part 1.

With the ANM-Algoritmus you will have three output options to find your position with graphically methods as tools provide at the nav-map:

(a) Find your position as in real life by drawing two „Standlinien“. It is the most complex option for the user because you have to draw a lot of angles, distances etc. You also have to consider transformations from WGS to UTM_120 (the coordinate system of SH5). You also have to be careful when using the algebraic signs. You will see, this is an option for the sophisticated navigators among us .
(b) Let the „Navigationsrechner“ do all the algebraic stuff for you and find your position with horizontal and vertical distances (cartesian coordinates). This is the arcade version…
Method (c) has the same background as (b) but for those who like using the triangle, the position can be found with polar coordinates.


Now I will show you how it all works when you are at sea. I will not show any interim results which will go beyond the scope of that post.

You are somwhere in the atlantic sea, approximately 200km NW of Ireland, but you don’t know exactly. It is the 5th of October 1938, 10:12:00 UT1. You fix your assumed position (G1) at 56°N / 11°W. Now take your Sextant and shoot the sun!

Input values 1st measurement
Date: 05.10.1938
Time (T1): 10:12:00
G1: 56,0°N / 11,0°W
Angle of sun h_b: 22,898° (in this example, we use precise values).

From SH5 we read our exact position: 56,5°N / -10,0°W. Now we have all we need for the calculation. Start the Navigationsrechner!

Results of 1st measurement: (take a look at the three options mentioned above)
(a) Azimut A: 141°; Radialversatz u: -5,8km
(b) W: ---,-km; V: ---,-km
(c) Azimut epsilon: ---°; Distance D: ---,-km

What’s that? No values for (b) and (c)? These values are a result of the whole procedure so you have to make another measurement at afternoon.

But in the meantime we can draw our first „Standlinie“ with the results of (a). We start with drawing the Radialversatz u with the circle.



Now we draw the Azimut A:



What you can see here ist that A is always positiv from north heading clockwise. Attention: Parameter u is negative so you have to draw the Standlinie diagonally opposite of A! The Standlinie is tangential to A. Don’t be lax at this point. Your position will be terrible wrong! Here we go:



Well done!

For our first example, we loiter around at G1, not changing our position (phi…course and s…distance). Time has come to shoot the sun a second time:

Input values 2nd measurement
Time (T2): 14:25:00
h_b: 24,423°
s: 000,0km (!)
phi: 000,0°

Results of 2nd measurement:
(a) A: 211°; u: -79,64km; Horizontalversatz w: 57,85km (NEW!); Vertikalversatz v: 4,49km (NEW!)
(b) W: 118,9km; V: 63,6km
(c) epsilon: 63°; D: 133,5km

Ok, lets start with (a):

Draw u and A like done above. Because the triangle of SH5 only reaches from 0° to 180° we have to draw 360°-211°=149°:



Now we extend the two Standlinien so that they intersect:



The intersection point (“Markierung 1”) is your true position. But beware: this is your position at WGS relativ to G1. You have to transform (stretch) the intersection point to UTM_120. That’s what the Horizontal- (w) and Vertikalversatz (v) is good for.



Vola! Our true position ("Markierung 2") is 56,505°N, 10,01°W. The error is about 0,3km. But consider that all input values are exact ones. If you use your sextant manually and additionally the Nautical Yearbook, the error is expected to be greater.

That’s a lot of drawing we have to do and it’s getting more complex if you change your position from T1 to T2. With (b) and (c) you can minimize your work load:

Using (b) - cartesian coordinates W (horizontal) and V (vertical) in UTM_120:



Even better! The often you use the default drawing tools the greater the error will be. You can also use (b) and (c) to check if your drawings are ok.

Using (c), things are more elegant (in my opinion):



Now let‘s compare our results:
Exact position from SH5: 56,5°N / 10,0°W
Position calculated with the ANM-algorithm: 56,5049°N / 10,0089°W
Using (a): 56,505°N / 10,01°W. The results with (b) and (c) are nearly the same, the variance is in a range of 100-200m.

The display of results summarizes the output values:
Red…(a), blue…(b), green…(c). The yellow marking shows the calculated variance from the exact position.



The true distance from G1 to our exact position was about 85km. The greater this distance is, the greater the variance will be. This is caused by the curvature of the earth and the plane geometry of the ANM-algorithm.

The results of the graphically methods will depend on your screen resolution. For this example I have used 1920x1080.

That’s all for this time, the next time I will show you how to consider position changes during the two measurements at T1 and T2 („Versegeln der Standlinie“).


Hope you like the results!

[Sartoris]

Looks very interesting! When do you think you'll be able to release this mod? I hope I'll be able to start using it soon!

[silent marshal]

Originally Posted by Sartoris

Looks very interesting! When do you think you'll be able to release this mod? I hope I'll be able to start using it soon!

It's frustrating: the theoretically background is finished but it seems to be no modders for realization out there. Have to be patient...

I will post a second tutorial for dynamic navigation soon (part 2).

[CaliEs]

Originally Posted by silent marshal

How to find your own position – Part 1.


...
Results of 2nd measurement:
(a) A: 211°; u: -79,64km; Horizontalversatz w: 57,85km (NEW!); Vertikalversatz v: 4,49km (NEW!)
(b) W: 118,9km; V: 63,6km
(c) epsilon: 63°; D: 133,5km

Actual i doubt that w "Horizntlversaz" and v "Vertkalversaz" has been correcly calculated.
A rough estimation shows me: w = 52.4km and v = 3.9km.

[silent marshal]

Originally Posted by CaliEs

Actual i doubt that w "Horizntlversaz" and v "Vertkalversaz" has been correcly calculated.
A rough estimation shows me: w = 52.4km and v = 3.9km.

Hi CaliEs!

Rough estimation, hmmmm. The ANM-algorithm consists of 82 mathematical equations. It's not easy for me to counter your critique unless I know a bit more about your algorithm.

Maybe you can explain your theory a little bit more, then I can reply seriously and "defend" mine.

But thank you for your comment.