Flight Simulator Tutorial #6 – macro review, scene derivation and integration, mapping of the 2D u-v data into a chartable 1D array, rejecting image artifacts

This section finishes the macro analysis and continues with the conversion of the u-v 2D formula array into a chartable 1D array. It also adds two columns to the chartable array, a masking condition for each triangle and masked u-coordinate which will throw out of the visible portion of the chart any shape which has a minimum of one vertex behind the virtual pilot. The section also explains the phenomenon of time derivation of the scene due to the action of the control devices (throttle and joystick) and how we can model movement by recursive integration of the scene (using a looped copy-paste operation performed by a macro).

[sociallocker][/sociallocker]

1. george says:

I almost did that few days ago when I put the sun in, I realized that putting labels there it won’t be good since they would be upright all the time. I decided to synthesize the cardinal points (including zenith and nadir later with Excel chart curve so that when you roll the plane they faithfully follow in rotation. The change in color of the landscape (green upright, maybe red upside down) was in the store from the beginning but I still have to implement the altitude function first (I don’t have that yet so I cannot do that). If you check the coaster I have something similar (three sets of points with three colors). They will all come together but I need to stick with the tutorials first.

2. george says:

That wold be a huge gain in speed. For now I am not worried about markers, but I will implement the large lines (anchored at the ends only later. I will soon add relief (uneven) and that is possible only with tiling. However later I will make a fast training version for 2007-2010 with no-vertex squares.

3. Peter says:

I also think traking the zenith (your sun image) and cardinal points of the compass relative to the aircraft axes is worthwhile. The increment dy should be omitted for these updates. You will need this for the compass (VOR beacons could be good) etc.

Mind you if you track zenith, North and East you will have the large rotation matrix linking the ground axes to aircraft axes. You could bypass the incremental update of the ground vertices and obtain them directly from the fixed set.

4. Peter says:

Hi George. You mentioned that the ground mesh updates faster if you omit the pretty markers. How about accepting the more rapid rendering that is achieved with lines only but then add another data series with judiciously placed markers; three or four lit highways or a set of runway approach lights say.