Apparent diameter of Moon from Earth incorrect

[granthutchison]

Could it not also be magnified by the atmosphere in real life,but not in Celestia,Jestr

Jestr, there is no "atmospheric magnification", despite the fact people always seem to talk about it in this context. If there were, astronomers would long since have noticed it.

Grant

[HankR]

The "magnification" occurs, not in the atmosphere or in the eye, but in the brain.

- Hank

[maxim]

The 'magnification' is an interesting psychological effect. I've made some testing recently and found quite some causes that make the moon look bigger, or not so big:

First it seems bigger against horizont, because the eye can compare it against something. A mid distant silouette of a big town or a rocky landscape makes it look bigger that a flat or far distant silouette. Near or very near silouette shapes make it look smaller than mid distant ones.

Near or mid distant horizonts with some 'smaller' object in front of the moon (a tree or a person) make him look really big. Watching the moon through trees makes him look bigger - this effect is more intense if you are not standing directly unter the tree, if there are no leaves, an if there are lots of small branches.

A more yellow or redish color seems to make him look bigger. I've found him looking bigger during the last eclipse, althought he was straight up in the sky. Somehow he seems to be smaller the more stars you can watch around him.

Finally he appears bigger to me on clear nights, and not so big if it is hazy or blurry cloudy. Opposite he seems to be bigger against sharp 'wellformed' clouds if he is full and lighting them in a mountain-illusion.

Ok, I've to stop here, or I gonna get moonstruck
maxim

[Matt McIrvin]

Moon illusion aside, the main reason that it looks unnaturally small for many users of Celestia is that the field-of-view default is considerably larger than the angular size of most people's monitors. I noticed that version 1.3.1 tries to do better with this, but it still runs up against a couple of fundamental, unavoidable problems. First, the computer simply doesn't know how physically large your monitor is and how far away you are sitting (this is also a big problem in Web design, incidentally, especially with size units in CSS). Second, even if it were possible to overcome that, a realistic default would make Celestia harder to use, since it's difficult to get oriented when you're looking at the virtual sky through a little virtual window.

As it happens, it's a full Moon tonight and the Moon is rising right outside my window. So I compared it with the simulated Moon to calibrate Celestia's field of view, and found that the application window I was using at the time (almost full-screen on my 17-inch CRT) was between 11 and 12 degrees across, from top to bottom (which seems about right). But that's a difficult size to work with; most constellations are bigger than that in the sky, and if you go to a planet it will fill the window completely. It's easier to use Celestia with a window that is unnaturally wide-angle, say 30 degrees or more. But that will mean that the Sun and Moon will necessarily look unnaturally small unless you have a gigantic monitor, or put your face right up to the screen.

[alvarezp]

Indirectly on-topic: To help avoid arguments of mis-sized objects, couldn't Celestia draw more flexible grids? I mean, every (float) "n" degrees or so, so (at least in Celestia) we could measure in degrees instead of "in fingers"?