Just thought I'd post this image of a pod interior scaling check. I've been working on the pod since the mid 80's, and "grew" a computer program over the years that started out exclusively for my illustration business. It was to build perspective layouts from 3D data that clients could look at, and request changes to perspective and so on, before the actual finished illustration was started. But a cool side benefit became the ability to lay a background image down and move/rotate a 3d vector image over, to match up reverse-engineered work to the real thing. In this way I could pin down using a system, whether errors in my measurements existed, or whether lens distortion was the culprit. The program handles various lens distortion issues. It also can "grid" out an environment so that true plan views can be generated manually from the data. I find that although the program is time-consuming to use, it offers far more flexibility and accuracy than say something like Photomodeler, which assumes you know what lens was used before hand. For something like the pod, you can never really know what lens was used, what its particular parameters were, nor whether the available images have been skewed, had their aspect ration altered to fit the format for a particular publication, or often how the image was cropped (which can make it difficult to nail down the lens center of view). But using the proper system, you can tease out photogrammetry errors and even find out "exactly" where the camera was located relative to the shot.
This image is a preliminary scaling check of my interior work. Still much work to do. Incidentally, it is looking very promising that, indeed, the pod set
really does fit within the pod volume quite snuggily.
The little squares that represent pushbuttons are NOT the top of the buttons, but the PROJECTION of the buttons down to the panel surfaces. So where it looks
like the squares are sitting lower than the buttons, they're actually in the correct positions. Keep in mind that any innacuracies are due to not yet
having dialed in the orientation of the wireframe to the image. There is one green line off on the far left that is heading off into timbuktu; that is a
display bug where an end of a vector line is crossing the projection picture plane and freaks out.
I paid for the disc after all.
