Mun lighting is wrong near sunrise/sunset.

[tater]

The sun is well up in ALL the images that follow. The Mun has no atmosphere, therefor there is no reduction in incident light except the brief time period when the sun is actually cut by the horizon, limiting the light by whatever percentage of the sun is obscured. It's gives an atmospheric look, certainly, but the important word there is "atmospheric."

Not noon, kinda late afternoon:

A little later

Later still.

Later.

Sun position for the last image above.

Edited November 10, 2014 by tater

[Decent Weasel]

Isn't the relative lack of intensity of light during twilight due to the surface being at an angle to the normal of the sun's rays, so 1m^2 worth of energy falls across a larger actual surface of the body? I mean, I know lunar dust reflects light in all directions, and so tends to appear the same intensity regardless of time of day, but the darkening of a surface at an angle is just radiant intensity. The moon wouldn't look like this, but it would be a question of modeling the soil correctly, not correcting for atmospheric effects.

Twilight was, in my understanding, the period of time AFTER sunset when the light emitted by the sky continued to illuminate the surface. On the Moon, radiant energy intensity per surface area of the moon would decrease as the sun moved lower into the sky, just like Earth. The difference, at least due to atmosphere, would be that once the sun was down, it would become absolutely dead midnight black immediately.

[tater]

No, it's not. The incident light should be the same for a surface normal to it. Note that the landers themselves are dimmer, they should actually be brighter if anything, as their vertical surfaces are now closer to normal to the incident light. In addition, any ground scattering would be towards the landers as well from the angle pictured.

This is flatly wrong.

The are reducing the incident light at low sun angles.

There can be effects due to the actual surface, and scattering, but there is no preferred direction to particles of regolith, so the scattering should be comparable regardless of time of day. I misused "twilight," I mean the period around sunrise/sunset in general, there should be no twilight at all, obviously.

<EDIT> note that the rover lander (tall one that had a rover slung below) is about the same brightness in all 3 images (slightly dimmer). If scattering from soil is isotropic, then it should be not as dim. If it scattered preferentially as a reflection (equal to angle of incidence), then the shot looking towards the sun should in fact have the ground brighter, not darker. Perhaps they coded illumination based upon Kerbin? So yeah, the light from a given cross section (normal) is spread out more on the surface, but it seems grossly exaggerated.

Another thing is the relative effect. The sun should be so intense, that even is measurably dimmer (grazing angle on surface) it would still be very bright. Images on the lunar surface are only not blown out entirely due to high f stop or a lot of filtering. So you'd get less light hitting a given munar surface area than a X hours before, but the quantity of light would still be greater than light during most of the day on Kerbin since none is reflected/absorbed by atmosphere. Seems overly dark with the sun still high.

Edited November 10, 2014 by tater

[Decent Weasel]

Good point about the landers. However, by that same token, darkening of the ground shouldn't be unexpected in general.

The incident light should be the same, the angle makes no difference.

If our surface is 60' off the normal, the illumination area (in keeping with cos(60) = 1/2) will be doubled. However, the radiant energy from the source will remain the same, so each unit area will receive half the light that it would at noon.

[tater]

Sun is almost half obscured in this image. The landers are good as they are illuminated. The surface was dark regardless of angle. I think it is looking at watts/m^2 or some equivalent, and basing the brightness on that. Still, using sun/moon as references, the moon is some 500,000 times less bright at full than the sun, and my yard (which is dirt/granite, I live in NM) is grossly brighter than the shots above under full moon light, and the sun here should still be a couple hundred thousand times brighter than that.

[tater]

Good point about the landers. However, by that same token, darkening of the ground shouldn't be unexpected in general.

If our surface is 60' off the normal, the illumination area (in keeping with cos(60) = 1/2) will be doubled. However, the radiant energy from the source will remain the same, so each unit area will receive half the light that it would at noon.

Yeah, true, I was thinking incident at a normal angle. But with no attenuation, that incident light is grossly in excess of what you'd expect at the same hour on earth/kerbin. So you've got cos60 per unit area of surface at that angle, but the incident light is higher to begin with (it's like 1300 W/m^2 at LEO, and under 400 at the surface (sun at zenith), right?, so ~3X dimmer). So 1/2 brightness per unit area with 3X the illumination and it should still be brighter.

Edited November 10, 2014 by tater

[Alshain]

I'm not sure there is anything they can do about it. I am not certain of this, but generally you tell the game engine where the light source is and it does the visual rendering. It assumes it's in an atmosphere. If I'm not mistaken, Unity is doing all this work and Squad is just telling them where the light source should be.

[tater]

Shadow angles are not dissimilar here:

Another with a low sun angle:

Obviously there is also the issue of photographic light capture vs the eye. Still, full moon is bright enough to easily navigate by on foot, even cutting through scrub with cactus around my house, and dusk on the Mun starts getting far more difficult to work with, even though the sun is entirely above the horizon.

In an image above someplace you can see my munbase 1… on a hill. I landed thinking that since I was well to the sunny side of the terminator, I could pick a spot well… it was surprisingly dark, and I set her down a few hundred yards from where things were actually flat due to poor visibility

Maybe it's just the brightness altogether. If the sun at the Mun was substantially brighter at all times, then dusk would be the same relative difference, but would be less noticeable until closer to actual sunset.

Edited November 10, 2014 by tater

[Decent Weasel]

Try color-matching white and black in the lunar images with white and black in the KSP screenshots? I know the moon's surface is crazy dark. (edit: I tried this and the results are middling.) Either way, yes, it should be brighter due to the omnidirectional reflective qualities.

I'm from New Mexico too, btw! What was that saying? "You might be from New Mexico if you regard half a ton of crushed rock as the ingredients for a nice yard." (Hey, I do.)

Edited November 10, 2014 by Decent Weasel

[sal_vager]

Considering that the sun used to shine right through planets, I am okay with a little inaccuracy when it comes to surface lighting at dawn/dusk