Originally Posted by Sammi79
Using a 'Sextant' image overlay is a troublesome and innacurate workaround for simulating a real one. The scale is derived from angular values projected onto a flat screen - which as they are represented on a flat screen are stretched the further you go from the centre of the screen. The problem is exacerbated by larger fields of view (compressing more degrees into the same space) It is hard to explain without diagrams, so to get you to see the effect for yourself :-
1. start a nightime mission in the northern hemisphere.
2. Find Polaris (last star in the tail of constellation 'Ursa Minor'
3. move your view until you can see Polaris in the centre top of your screen and the horizon at the bottom.
4. Note the actual distance between Polaris and the horizon (or Polaris and the top of your screen)
5. Pan your view left or right, keeping the horizon at the same level, until Polaris is close to either side of the screen.
6. Note the new distance between Polaris and the horizon (or Polaris and the top of your screen)
You can also see the effect by panning your view until either the sun or moon are in one corner of your display. The circular sun (or moon) will appear elliptical. It is a fundamental part of displaying 3D images on a 2D screen, and the sextant image overlay marks are calculated for 1 position only. When I use TDWs navigator sextant order button, the image pops up dead centre, top to bottom (it will do this no matter of the resolution of the sextant image, it will be scaled to fit top to bottom) which is exactly where the marks are calculated for. If you wanted a moveable sextant image, this would have to scale itself as you moved it around the screen to adjust to the differences in angular degrees over the display.
The trick for making a good sextant image, is to know exactly how many angular degrees there are displayed on your screen in the exact centre from top to bottom (a distance in pixels), then using trigonometry, calculating the position of every single mark. I sent TDW a sextant image that works on 65.84 degrees as opposed to the original 60 degrees and it is better but still not spot on.
If anyone can tell me exactly what the number should be I can make a sextant image as good/accurate as it can possibly be. The trouble is in cameras.cam the angular values for the exterior views are set at 75 degrees, which no matter how you look at it, is larger than the displayed screen (top to bottom). To make the image I need to know what the angular value of the displayed screen (top to bottom dead centre) is - I suspect between 65 and 67 degrees, Unfortunately with no means of working it out from the cameras.cam 75 degree value, It means lots of trial and error using very well known celestial objects.
Then we get down to how realistic are the celestial objects in SH5? Is polaris in game the actual celestial pole or is it shifted slightly as in real life? Height of eye calculations are important if you are standing on a sphere, however if the sea in game is rendered flat, HE is irrelevant. There is much to work out before being able to even test a sextant overlay properly. This is why if possible it would be better to implement a sort of upside down stadimeter, with a large field of view, and an angle readout. (keeping the horizon in view you move the image with polaris or other celestial body down until it is in line with the horizon, then read the angle.)
If you want to use the sextant overlay for angular measurement, It must stay dead centre top to bottom. You will find if you measure an object this way, then move the sextant image left or right and try to measure the same object, you will have a different result.
|