[REL] SH4 v1.3 Real Navigation Mod

[Frederf]

I've given celestial navigation here a try in game. Oddly with this mod my sun was still about 1° or larger which confused me. Perhaps I shouldn't have tried to JGSME it? I was very excited to try and spent a couple days setting it up. I read more than my brain has taken in since graduation as well.

I did not try any LOP or sight reductions due to my unfamilliarity and my lack of reference material. I thought I'd learn to crawl before I could beat Jessie Owens. The following is my impressions and creative spark on the subject.

Latitude by Sun or Polaris:

Once I got an actual value for when to take the sun shot at noon (the time to the minute) the better I did. I knew my position and looked to see if the expected cel. bodies were where they were supposed to be when they were supposed to be there. My verdict on the sextant... brilliant idea, just plain not accurate enough. I had ambiguity of 1-2° with the big sun which I might be able to get down to 0.5° with a lot of practice, 30nm. Not the best. I also found it very hard to take most of my shots whci were very high overhead 60°+

During the day I was lucky to get partial cloud cover. This reduced the sun from a huge blinding glare ball to an actual circle to get a measurement. I don't know how good any of these shots were because I was running 1280x1024 (a very common but unsupported resolution by this mod) and 1280x960 to try and weed out the possibility of error.

Polaris is hard to find for me. Video game stars (Silent Hunter, OFP/ArmA) are harder than real life for some reason. Knowing it was directly north and between 1° and 5° above the horizon helped me find it (latitude 3°). Freezing the game with the sextant bottom on the horizon helped quite a bit as did finding a game station that was view stabilized like the flak or deck gun.

When I was 1° S lat. I had pretty much no idea how to take a measurement. I heard there's some method of using a Orion belt star at it's maximum altitude as a 90-L measure but didn't attempt it.

Longitude by SR/SS:

The SS times to the upper limb of the sun via this program I got were brilliantly right on. Plug in my known position (sent a patrol report), the date, and the SS time was right on. The SR time however was 2 minutes too late consistently. By the time the calculated SR time came around the sun was always exactly 1/2 way out of the water.

If I could find a source with SECONDS I could probably get very acurate longitudes, ironically much better than my sun latitudes. The chronometer is by far more accurate than the sextant in this regard. Of course I compared calculated times based on a known position (down to the minute) and not the usual finding position based on times. I figure I'd only be a minute of time off, so 4nm which is acceptable for me.

Conclusions:

1. Sun glare is killer. To create reasonable sun measurements a filter needs to happen or just to turn off the glare altogether.

2. Create a stable, dedicated sextant platform. A special clickable zone of the conning tower that had no-rocking-with-the-sea, any FOV you want (45°?) to have better readings.

3. Verify sun and moon size. This is probably just my failure to read install instructions or read angular size but I swore that the sun was not the proper 0.5°.

4. Look into a "deckgun" sextant. Try pointing the deckgun at a point above the horizon, then hit tab or zoom. Notice how the range number is based upon elevation and is very sensitive. If someone could do the same kind of coupled number-elevation for a "sextant gun" than sextant readings would go from +/-0.5° to +/-0.1° or better.

I'm still interested in this idea, just wary of how practical it is as the mod and game stand now.

[Munchausen]

Polaris is hard to find for me.

The easiest way to find Polaris is to use the Big Dipper (Ursa Major). The following PDF file illustrates:

http://www.fvastro.org/beginners/beginner-part1.pdf

Notice how the front edge of the "dipper" points directly to Polaris. If you'd like a complete sky chart, try this one:

http://observe.phy.sfasu.edu/SFAStar...rChartsAll.pdf

The SR time however was 2 minutes too late consistently. By the time the calculated SR time came around the sun was always exactly 1/2 way out of the water.

The halfway point is suspect ... interior lighting always switches from night (red or blue) to day when the sun reaches its halfway point ... and switches back at the same halfway point during sunset.

I assume you're using the sun tables that came with the mod. If so, what specific corrections are you making to adjust for your current position?

[Frederf]

Using the big dipper is no good as I was at 3°N and 1°S latitude respectively, the big dipper was through a couple billion tons of rock and slgihtly hard to see. Again this is a small matter that simple practice and star charts can easily remedy, not worried at all.

Actually my SR/SS times came from Almanac 1.0 by Stephen R. Schmitt, 2002 but I checked them across to the tables that came with the mod and seem very close. I'll do a direct comparison (since I inputted a weird lat/long that's not on the table).

20N Mar21 1943
Table: SR 0604, SS 1810
Almanac1.0: SR 0604, SS 1811

30N Jun4 1939
Table: SR 0431, SS 1924
Almanac1.0: SR 0432, SS 1924

So it looks like the table is from the same data (perhaps the same program, pre 1950 being hard to get) and the table simply averages over 1939 to 1945 while the program does not. Another thing I noticed is that setting the time zone to anything other than +0 Z in the settings gives confusing results. It seems to say "The local time in time zone X will be _____ when the sun is rising in time zone Z." which may be right but is entirely unhelpful. The local time that the sun rises in each time zone on any given day of the year should not be more than 10 seconds or so different.

I've also found that the calculated sunrise, the upper limb rise, the center body rise, and the red light switch happen at 4 seperate times. This notion of "The red light goes off when the sun is half way out of the water" doesn't hold true in my experience. Instead the red light tends to go off a few minutes after half-sun.

More testing ist reqwuiret.

[Munchausen]

Originally Posted by Frederf

Using the big dipper is no good as I was at 3°N and 1°S latitude respectively, the big dipper was through a couple billion tons of rock and slgihtly hard to see.

Polaris is usually used when you want to determine your latitude (assuming you're in the northern hemisphere) and, consequently, you need not wait until sunrise or sunset to take a sighting. Depending on the time of year, Ursa Major will eventually rotate around to where you can use to it find Polaris. Granted, star shots were usually made at twilight so, if the Dipper isn't around when you need it, try using Cassiopeia (shaped like a quirky letter "W" with the center spike pointing toward Polaris).

Another thing I noticed is that setting the time zone to anything other than +0 Z in the settings gives confusing results. It seems to say "The local time in time zone X will be _____ when the sun is rising in time zone Z." which may be right but is entirely unhelpful. The local time that the sun rises in each time zone on any given day of the year should not be more than 10 seconds or so different.

This is really the problem. "Base Time" is not really defined. If the watch was set to GMT, it would probably confuse most players (having the sun rise at midnight would draw far too many complaints). As it stands, it's almost impossible to determine if the error is in latitude or GMT.

For example, for a D.R. of 12*15' S 115*00' E, on 26 Jun 42, I calculated sunset to be somewhere around 1745 Z. If I were to then use Fremantle as my "Base" time, correction for longitude would move sunset to 1744 "Z" ... not exactly GMT. If I did use GMT, at 4 minutes of time per 1 degree of longitude, that would be about a 7 hour 40 minute "correction."

I've also found that the calculated sunrise, the upper limb rise, the center body rise, and the red light switch happen at 4 seperate times. This notion of "The red light goes off when the sun is half way out of the water" doesn't hold true in my experience. Instead the red light tends to go off a few minutes after half-sun.

Probably because you've changed the size of your sun (or not). I left my sun at default and the switchover is consistent ... midway between one limb and the other. Or, timewise, 10 mintues ... 20 minutes to fully rise or set. The moon, on the other hand, only takes 14 minutes (thereabouts ... I haven't had a full moon yet to make certain).

Btw, compared to the above D.R. calculation, my actual present position at "sunset" was 12*35' S 114*55'E ... and the sun actually set at 1900 ... 111 minutes late.

[gutted]

extremely interested in all of this.. but it seems like there may be too many problems with it.

hope someone can sort it out and give a good tutorial for the navigation beginner.

[Frederf]

I kinda have to complain that this "real nav" mod removed all the fine lines on the map for lat/long. This makes plotting your position, once calculated a real pain! I can still zoom in the same amount but now there's just blank ocean-blue paper there.

Polaris is fine, in face the W constellation is easier for me to find more often than Ursa major/minor. The base time should be defined as the local time in the center of the time zone that your base is in. Basically if your base is in San Francisco your base time is GMT-8. Other huge SS/SR time differences during some of my tests are probably due to other factors besides this. I find base time a no-brainer.

Further tests show that the red light goes on/off at mid-sun.


Can I ask what people have been doing for accurate latitude shots that exceed the rather rough judgement of the altitude of the sun using the sextant? Also using sun for latitude is not a simple 90-A = L arrangement but involves what day of the rear it is. Can anyone post details about that correction?

[Munchausen]

Originally Posted by Frederf

I kinda have to complain that this "real nav" mod removed all the fine lines on the map for lat/long. This makes plotting your position, once calculated a real pain! I can still zoom in the same amount but now there's just blank ocean-blue paper there.

I didn't install the mod ... I just try to use sun and moon tables to determine when to surface and when to remain submerged. So I didn't realize this was part of the mod. In fact, the navigation chart should have more, not fewer, lines. I suspect the lines were removed to keep the player from "cheating" ... but it's hard to plot a D.R. or fix without latitude, longitude and tic (incremental marks along lat/long) lines.

The base time should be defined as the local time in the center of the time zone that your base is in. Basically if your base is in San Francisco your base time is GMT-8.

My base (point of departure) was Fremantle. Although sunset is later than I'd expect using the Tables, it comes nowhere near what I'd get if I used GMT +/- Fremantle's time zone.

Also using sun for latitude is not a simple 90-A = L arrangement but involves what day of the rear it is. Can anyone post details about that correction?

Thanks to the tilt of the earth, the sun doesn't "top out" at the same height above the horizon every day of the year. The sun reaches its highest point ("topping out") at what's called Transit ... and you can build a table for transit times in much the same way as you build a table for rise and set times.

Example:

It's the 4th of July and you want to find your latitude by taking a "noon" sun shot. Using the sextant, you track the sun and find that it tops out around 1204 local time, at 82 degrees ... to the north.

Transit tables show that, at latitude 30* north, the transit altitude of the sun is 83 degrees on the 4th ... but the sun is south of the observer. So ... you've got the wrong table.

At latitude 20* north, transit time is 87N (the sun is now north of the observer ... so, somewhere in between 30* north and 20* north, the sun was directly overhead ... if you were to sail quickly from 30* to 20* you'd watch the sun pass from south to overhead to north of you).

At latitude 10* north, transit time is 77N.

Interpolating, transit time for 82N would be latitude 15* north. This would be your present postion.

[Frederf]

It should be a simple fix to slog into the SH4 nav mod and ... ya know.. make one that makes sense in terms of how detailed the chart is in lat/long lines. They do state something about cheating prevention as reason for some map changes. I thought the intial zoom level change was sufficient. Invisible forest, beautiful trees.

I was trying to plot a fix when I realized that a degree wasn't 60nm wide... not even close. I tried interpolating like "My longitude is 55.487% between 150°E and 160°E so do the math. It's 67.8nm between so it's 33nm east of.. blah blah math.

My base (point of departure) was Fremantle. Although sunset is later than I'd expect using the Tables, it comes nowhere near what I'd get if I used GMT +/- Fremantle's time zone.

I did this too. Once for Midway, once for somewhere else. In Midway the SS/SR times were really close, in the other place they were way off. I don't know if it's the Base Time / GMT difference necessarily that is making these times wrong. The Base Time might work as it should but there might be some other aspect of the game that is causing the times to be off.

On the other hand, I am still suspicious that Base Time might be a weird time zone like GMT+9.74345 because it's 15*9.74345° west of Grenwich for example.

The sun reaches its highest point ("topping out") at what's called Transit ... and you can build a table for transit times in much the same way as you build a table for rise and set times.

So a table is required? I could build such a table but no such table came with the mod in the hodgepoge of .pdfs and .docs. Also while I have your attention, there is this definition of transit or altitude that gets thrown around quite often and even through a few hundred pages of reading I can't get anyone to really explain it before talking about oblate spheroid this and equation of time that.

A transit of the sun is:
1. Not neccesarily at 12:00 local +/- 20 min or more.
2. Happens when the Sun is at the highest elevation
3. Happens when the sun crosses directly 000/180 true (also known as crossing the meridian)

Correct?

And how accurately are you measuring altitudes with the sextant? To the 0.3°?