Special Aircraft Service

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: A discovery regarding the cumulus cloud construction  (Read 284 times)

0 Members and 1 Guest are viewing this topic.

WxTech

  • Modder
  • member
  • Offline Offline
  • Posts: 6040
A discovery regarding the cumulus cloud construction
« on: February 20, 2024, 05:41:09 PM »

When the field of view is wide, about 90 degrees and wider, cumulus clouds located near the screen edge take on a strange appearance, whereby the 'outside' edge is squished rather flat and hard-edged.

In the 3-panel view below, I use a cloud texture I made up years ago to help see how the clouds are put together. In each view, the part of the texture having a letter "A" inside a box is more or less centred. The rightmost view has that 'cloud' located at the centre of my screen. In the middle view I have panned the camera so as to place that object fairly near the upper left screen corner. In the leftmost panel that object has been moved right up to the upper left screen corner.

Note how the left edge of the "A" gets quite compressed. This is the result of the object actually being a multi-faceted polygon, kind of like a crude hemisphere bulging toward the camera. This is how the shading is applied depending on the direction of the light source. That is, each of the several textures making up an individual cloud is not simply flat; it has a 3D shape.

What I'm trying to understand is the reason for effectively rotating the object depending on where it lies within the field of view. When central, the object directly faces the camera, and is fully symmetrical. But as the view pans so as to move the object to the left, the object in effect rotates so as to push the left (outside) edge away from view. Which causes the compression of that side due to foreshortening.

This is, at first blush, nonsensical. For instance, other objects like buildings do not exhibit such 'rotation' simply by panning the view; that would be really odd. I wonder if this is a necessary consequence of the fact that a cloud is virtually always made up of at least a few individual such objects. Even so, it still seems to me that it should be feasible to maintain the unchanged orientation of the 'bulge' toward the camera at all times.

The ultimate downside to this effective rotation occurs when the camera is very near a cloud. As the view pans a cloud element can go from fully covering the view to completely moving outside the view. This is very unnatural. Too bad the generation of these clouds seems to occur in a .dll...

Logged
Great minds discuss ideas. Average minds discuss events. Small minds discuss people. - Hyman Rickover (but probably predating his use.)

WxTech

  • Modder
  • member
  • Offline Offline
  • Posts: 6040
Re: A discovery regarding the cumulus cloud construction
« Reply #1 on: February 21, 2024, 03:24:17 PM »

To better see what multi-faceted polygon is created, I added a grid to one of the texture parts (the letter 'A' in my 4x4 panel). This is a 105 degree wide FOV, and is not cropped (as was done in the 3-panel mosaic previously). Now it's clear that the square texture is treated as a pyramid, which I've traced out on one object with darker blue lines. The peak of the pyramid in projection always moves away from the middle of the square along a radial aligned with the screen center.

Logged
Great minds discuss ideas. Average minds discuss events. Small minds discuss people. - Hyman Rickover (but probably predating his use.)
Pages: [1]   Go Up
 

Page created in 0.033 seconds with 26 queries.