Globe improvements

[Yankes]

my build is available on https://github.com/Yankes/OpenXcom `test` branch (`master` is obsolete)

[ps]
I made some test and i get something like this. Volutar this version is good enough or you will prefer true dither?

[Volutar]

Yankes, neighbor stripes should be "smoothed" but not overlapped through each other. Unfortunately this depends on zoom level. While this picture is quite acceptable, zoomed out picture will look more "noised" at twilight zone.
I suppose you've set rand()%8 value? I prefer %4, though %8 is much better than %20 which was at first. Flaw of this "smoothing" approach is indistinguishable "full night" and "full day" line.


I've attached vs2003 built version. Which is, I suppose, quite old (in terms of git branches).

[Volutar]

By the way, quality of picture greatly depends on display you're using. I have S-IPS display at home and IPS+TN at work. And I have noticed that ocean blue seems to bee more "smoothed" (despite on noise) because of falling into dark on TN more, than on IPS.

[Volutar]

Another try.
I've slightly modified palettes.dat file to make globe terrain shade more solid and realistic (ocean part kept intact).
This modification doesn't affect on other parts (they are used only for earth textures).

It might become too dark, but day parts become more distinguishable from night.

[Yankes]

Its look better that standard palette.

[Volutar]

New version of shading makes polar iceberg zone "normalized", comparing to original.
Ocean color is lighter than texture of mixed polar/water.
On a screenshot contours of "non processed" area above and at the right side. So you can see how "smoother" and nicer new lighting looks.

https://www.fourmilab.ch/cgi-bin/Earth?imgsize=320&opt=-l&lat=22.875&ns=North&lon=65&ew=West&alt=1000000&img=learth.evif

On the second image is Earth view from initial point (openxcom) with timestamp above. I think it should be used as fine example for terminator line position and size at particular moment.
So there are some mistakes in inclination calculation.

[Volutar]

Okay.
I was experimenting with making twilight zone more realistic.
There's some results.
Outsize "main" terminator line is a twilight zones - wide area where there's no full night (terrain is slightly illuminated by atmosphere) and not so wide area of sun near horizon (it's not full day).
Some noise pattern also present, it is more noticable on ocean surface, and less on land.

[Volutar]

https://www.youtube.com/watch?v=sqmpNWExKmo
Video demonstration. Though it looks weird, smoothed slightly + encoding artifacts.

[SupSuper]

Just FYI, Yankes' patch is now officially in the codebase.

[kkmic]

That's good to hear.

GG Yankes

[Yankes]

today globe, tomorrow rest of OpenXcom
SupSuper how about adding skin color to battlescape? dont change much but increase immersion

[luke83]

i second his skin colour suggestion

[SupSuper]

today globe, tomorrow rest of OpenXcom
SupSuper how about adding skin color to battlescape? dont change much but increase immersion

Sure, just throw me a pull request, they're easier to manage than poking around your various repositories.

[Conti-k]

Is there any chance to tweak new geoscape lighting, and make the noise more "smooth", or completely remove it? It reminds me of low shadows samples when using area lights, or area shadows/skylight (scanline renderer)

[kkmic]

Is there any chance to tweak new geoscape lighting, and make the noise more "smooth", or completely remove it? It reminds me of low shadows samples when using area lights, or area shadows/skylight (scanline renderer)

That was something I asked about some time ago in this very thread. That and the next two posts should explain it.