Yep, another sunrise/sunset topic.

[Von Due]

After searching the forums and getting the moon/sun table I am still wondering how this works in SH3.

For one, the game world is flat and rectangular but still I am going with the degree marks on the nav map trying to figure out how to come to agreement with the rest of my crew when the sun sets and rises. I am using the interior light switch as indicator.

Testing, I sailed off to E 0* 0' N 60* 0', got there 3. August 1939. Submerged in time to position myself within no more than 5 metres from the actual spot.

3. Aug. Sunset 2117 GMT
4. Aug. Sunset 2108 GMT

7 minutes difference.

I understand how the sunset hour changes as the days go by but 7 minutes from one day to the next seems a bit much. Is the switching to night lighting a good enough indicator or should I have watched the sun set myself?

Digression:
The next sunrise, Immediately after the lights went to day lighting, I got the true bearing to where the sun rose. It was close to 54*
At 1200 GMT I noted the sun's true bearing at about 166*
In a whim, I looked at these angles:
90* - 54* = 36*
180* - 166 = 24*

36* + 24* = 60* which happened to be my latitude. Question that needs answering is whether this was a coincidence or not. The thought is that the true bearing will be offset from due east the further away from the Equatoral line I get. I want to see how much that offset is at various latitudes. The offset from due south at noon is more puzzling. I haven't got a clear idea what's causing that.

[Sailor Steve]

I can't answer for bearings, but a quick look at an almanac for any year and any place reveals that sunset time changes by about one minute in the month of August, so you're right to question the in-game sunset times.

[Pisces]

The light in the game flips when the sun is halfway on the horizon. Yet RL sunset/-rise calculations are based on the full disk being obscured, or the upper edge being on the horizon. So you have to take that time difference into account.

[Von Due]

Thanks to both of you. It seems to me the game world is a bizarre mix of spherical geometry applied to a flat world. It's a challenge making head and tail of the observations. I remember getting headaches years ago when I last played SH3 trying to figure out the timetable. I feel that headache is returning :)

Basically, what I have been assuming so far is that, in the game, at a given latitude, the sunrise/sunset would happen at the same hour local time every 15 degrees longitude. I'm more than ready to challenge that assumption but I will carry on testing, getting the sunset hour at N 50* and N 60* at W 0* and W 15* (which would be centre of that time zone I assume).

The light switch before actual sunset is precicely the sort of info I needed. Now I know the switch master can't be trusted and I will watch the sun myself, if possible. I could need an Always Clear Skies mod for that :har:

[Pisces]

Also, iirc there was something funny with where the game considers the timezones to end. Like GMT (or UT) is valid within 7.5 degrees on either side of the Greenwich Meridian line. Whereas in RL the timezone exists only 15 degrees on the west side of the Greenwich Meridian.

And then iirc there was also mention in one of the old Celestial navigation threads of local time or base time not being correctly displayed for the local timezone after a savegame reload. I never noticed this myself. But something you might want to keep an eye on and do some investigating.

[Von Due]

Greatly appreciated info there, Pisces.

EDIT: Just a short report

8. Aug 1939

W 15* 0' N 75* 0'

Tracked the sun at 0000 local time, again at 0100 and last at 0300.

At 0000 the sun was at true bearing 346.5
Lowest height above horizon at just before 0100 due north.
True bearing at 0300 was 30*

It adds up neatly that in 3 hours, the sun moves 14.5* pr hour, given errors in measurements and eyeballing.

The sun at the lowest at 0100 local time (ship time 0200) on the other hand gives me more to ponder.

Right now I'm considering Earth to be cylinder shaped but who knows.

[Gargamel]

As Pisces said, the game bases day/night from where you loaded. So when load in port, the time is accurate to the exact local, not your timezone. As you move through the game world, the clock is not adjusted, nor does the sun alter it's path (based on whatever the game engine works off of, as you are finding out).

When you load a game from a save mid patrol, the clock is saved and loaded without change, but the game adjusts sunrise/sunset as if the ships clock is now the local time. So during operations off the America's, or on a trip to the pacific, you could save in middle of the night, but load during the day, without the clock moving.

Not sure if this helps in your calculations, but it is something to consider.

[Von Due]

Thanks Gargamel. Every test run I've done so far I have started as a new career, setting sails at 1. Aug and not saving, just aborting and restarting for each test. Much quicker than sailing 3600 km hither and dither :) beside, the measurements have to be done at the same day each test run but it sure adds to the already mindnumbing sun cycle that the game can't keep track of itself. If the sun cycle can be sorted then I'm sure the other bits can be introduced to the calculations later.

[danzig70]

The sun doesnt seem to follow local time for me. If I start a patrol in the pacific it just follows the main "game" time, not the local time.

[Sailor Steve]

How do you start a patrol in the Pacific in SH3?

[zygoma]

FWIW, I lived in Alaska for 30 years. Most of those years were at 60d 30m N, 151d 04W. As the year approached the summer and winter solstices, the amount of change of possible daylight were quite dramatic (the weather TV guys announced the change every day) . At solstice, the change was around 7 minutes a day, so we lost (or gained) about an hour a week at the extremes. As we changed towards the equinoxes, the amount of change decreased, until there was zero change right at equinox. Mind you, this was where I spent most of my time, in what we called the "banana belt", because it seldom got colder than about -30F.
When I lived in Fairbanks (known to locals variously as Squarebanks or Bareflanks) the change was even more dramatic, as we were then just 100 miles south of the Arctic Circle. At 67d 11'N, 149.10'W, the change was about 9 minutes. At August we (in Soldotna, on the Kenai Peninsula) saw about 4-5 minutes' change; same in February.
When I was a firefighter/paramedic in Fairbanks, every Summer Soldtics a bunch of us and the local constabulary would head up to Eagle Summit, just barely north of the Arctic Circle, for an all-night party, with parachutes set up for tents and lawful (and sometimes unlawful intoxicants) and watch the sun barely dip to the top of the horizon, then continue on its inexorable climb back up. At that time of year, the sun's path described a circle in the sky over 24 hours. But since we were only barely north of the Arctic Circle, the effect only lasted one or two days. We did *not* observe the Winter Solstice, as the average temp that time of year was around -60d F. The people up on the "north slope", along the coast, were the poor folks who got to enjoy (or endure) the couple of months of complete darkness in winter, but complete sun (plus or minus rain) in summer.
In the sim, I've been frustrated to find the sun still up north of Scapa Flow at 9pm in summer, until I remembered my own experience with extreme sunlight changed that far north.
But it sounds that the "old timers" here in SH3 have been able to note the celestial mechanics as modeled in the sim. To you folks, I say "Good on you, Mates!" for taking the time to analyze the interaction between the real world and the sim's behavior.
You folds continue to astonish me with the depth to which you're able to observe and analyze this kind of thing. I'm in nearly constant awe at the attention to detail. I like to think that if I hadn't gotten my "owie de la cabeza" a couple of years I'd be in there, too, but alas, the vagaries of dangerous professions have taken their toll on me :(
I am impressed beyond words. :salute:
-- Zygoma -- the perpetual n00b

[Von Due]

I did a test on the timer, saved mid patrol in a different TZ than Wilhelmshaven and reloaded. Both the ship clock and local time were showing what they showed when I saved so I'm not sure what the issue with local time is.

Other than that, I decided to compare the game world against this http://www.weatherimages.org/latlonsun.html It's not very immersive, simwise (unless I could find one in German of course, like Kapitän Lazybutt ordering his nav officer to hand him the hours) but it would be interesting to see how well the game world compares to the real one at higher latitudes.

Luckily, SH3 doesn't mind me pressing the Win key in-game, so it's easily accessable.

One thing I wonder is, in real life, how did they keep track of the sunrise/-set hours when in the middle of the big, wet nowhere?

[RConch]

As bad as SH II was, one decent item was in the Captain's cubby was a wall calender that gave you sunrise and sunset and the phases of the moon if I remember right.:hmmm:

[Von Due]

Alas, Herr Rconch, that is no more. I remember and it was one of the things about SH2 that made me forget the super destroyers. It is perhaps the feature of SH2 that I miss.

Note: Interesting, the sun calculator in that link gave me the sunrise hour at 0 east 50 north 6. aug 1939 at 0535. Luckily the weather was perfect, clear blue sky and no wind and I got front seat tickets for the show. At 0535 the sun was just touching the upper edge of the horizon mist, disc halfway above the horizon, at the same time the switch master flicked the switch.

Perhaps this needs a complete rethink. I might have to go along with sunrise is when the disc is halfway up, same I will assume for sunset. The big question though is what the game counts as sunrise and how much it matters when it comes to game mechanics.

EDIT: 9. Aug E 0 N 60
Again, at the announced 0557, the sun was halfway up the horizon (tip above mist this time) and the lightswitch guru did his deed. Indeed it seems that the Ubisoft sunrise is a bit out of phase.

Now if jumping out of the sub and into the browser wasn't so sonar bleepin non immersive but so far the best chance I have to get the hours right for some nice planning and scheming seems to be just that.

EDIT 2
Went to W 15 N 60
Now the sunrise was less on time. Announced it was at 0607, but happened about 5 minutes earlier. Not a lot off but still. Now for some really strange stuff

All sunrises at E 0* were pretty spot on to what was calculated. However, that was only when I used the ship time, and had the TZ box set at 1 hour east, regardless of where I was. Local time had nothing to do with the calculated times it seems. Only if I set the TZ box to places I've never been did local time match the calculated time. For the record: I have not saved or loaded the patrol. It was a new career, starting 1. Aug. Local time and ship time match at 0 longitude but still, I had to leave the 1 hour east in the calculations to get the correct ship time. I am not going into the moon and that cycle. That thing I have already dismissed as nice to look at, doesn't work at all.

[reaper7]

Have you seen this thread regarding the way SH3 handles the clocks:
http://www.subsim.com/radioroom/showthread.php?t=193974
http://www.subsim.com/radioroom/showpost.php?p=1869621&postcount=21

[Gargamel]

Betcha the engine handles the sun as a point particle and then draws a radius from that.

@SS: The south east Asian campaign in gwx?

[Von Due]

reaper: Thanks and I will digest it the best I can. I did find that the longitudes where time would jump between zones weren't where I entered the zone but it was halfway through that zone. Glad to read others agree!

...and then it went haywire. I traveled 22.5 degrees west and up to 60 north and both the ship time and local time missed the mark by around 20 minutes iirc, nothing time zones could explain.

[Von Due]

Ok, now I feel stupid. Let me see if I can get this right.

At say, 1200 GMT, local times at 15* west and 15* east will then be 1100 and 1300 respectively. At 7* 30' local time would be (west) 1130 and (east) 1230 but the displayed local time is only valid for each multiple of 15*. So: To get the actual sun time for the longitude I am at I will need to know how far off I am from the mean median of that time zone and subtract/add from/to the displayed local time 4 minutes per degree or 30 seconds per km. I hope I am correct so far or else I'll trade this VIIB for a moped.

The further north and off the Greenwich median I go, the worse the discrepancy between calculated sunrise hour and actual in game sunrise hour becomes. The parallel longitude lines are the prime suspect as the warping to fit the cylinder world would get worse the farther from Equator I get. It would be good for a laugh to see what the game would suggest to be local time at E 60* N 89*

What I could need it seems, is the maths to convert in game long/lat positions on planet Cylinder Prime to actual long/lat positions of Planet Earth that would give me the correct hours. I am a bit at loss how to go about that other than spending a few weeks real time mapping the GMT time when the sun is at its highest position. The visual warping in the optics does not help either.

[danzig70]

Here's a good map with lat/long and time zones. Note that the map projection may not be the same as the game uses. http://www.lib.utexas.edu/maps/world_maps/timezones_2001.jpg


PS: I am in the pacific because I write my own campaign files and do not follow a historical scenario.

[Von Due]

Thanks! I did find a suitable projection which not only look nice and coloury and a lot like the SH3 worl map but also makes the longitudes I'm reading at high latitudes make sense. It's an equirectangular projection

http://www.nevron.com/gallery/FullGalleries/diagram/mapprojection/images/MapEquirectangularProjection.png

Here's a comparison between SH3 at E 60 N 80 (measuring the distances 80x120 km and 60x120 km from Equator and the GM median, also verified with the degree marks on the map) and Google map at the same coordinates

http://i189.photobucket.com/albums/z193/WhoCaresWho/SH3E60N80.jpg

http://i189.photobucket.com/albums/z193/WhoCaresWho/GoogleE60N80.jpg

The lat/long positions I am reading are correct. I am now assuming it is the sun cycle itself that isn't correct for all positions. That would be a blow to my attempts of finding the sunrise/-set hours and perhaps the reason for all I know Ubi never mentioned these hours in SH3 but hid the almanac in that locker I can't open.

[Von Due]

On one of the last test runs I will do, it seems, the in game sunrise hour happened at calculated sunrise hour for a position 258-230 km west of my position, ingame sunset happened at calculated hour for a position 274-302 km east of my position. This makes no sense at all. No time zone or mismatching ship/local/GMT time can explain that as far as I'm aware.

I will do one last attempt, going to various positions at a few constant latitudes to see if the offsets show any pattern, then at a few constant longitudes. Unless I can find that pattern, I will say the game runs a sun cycle of its own creation.