|Image 1: This is an example for a depth resolution
that is too low (240): "resolution 320 240 240". You see the frayed edges
of the blue parts of the object. There the object is so thin that it is
not hit by a calculation with z-resolution of 240.
No antialiasing is done: "antialiasing 1"
|Image 2: This image has an increased z-resolution
(2048): "resolution 320 240 2048". The edges are no longer frayed out.
Nevertheless you can see disturbing patterns of pixels caused by the so-called Moire-effect (on the orange parts of the image e.g.). They could be reduced by antialiasing, which is not done in this example: "antialiasing 1"
|Image 3: Here is the image for reference.
It has a z-resolution that is high enough (2048): "resolution 320 240 2048" like the image on the left.
Additional it has an antialiasing value of 4: "antialiasing 4", which means that every pixel is the average of 4x4 sub-pixels.
Look at the example image above. Sit in front of the screen at the distance
you usually work at, and then hold your thumb between screen and your eyes
(nearer to the eyes than to the screen). You should sit straight in front
of the screen, not incline your head and not look from beside at the screen.
Now look at the thumb as usual. Pay attention to the images on the screen
(without looking at them directly, always look at the thumb!)
If you see four images now, then the thumb is too near to the eyes. Move it a bit further away from the eyes and look at it again as usual, while paying attention to the background. Change the distance thumb-eyes until you only see three images. Now the one in the middle is seen three dimensional, but perhaps not sharp. To get a sharp image is a matter of concentration, with increasing experience it gets easier. Watch the non-sharp image for a while (but don't look at it directly). There'll be a point at which mind begins to recognize, that the data that it gets, make really sense and it tries to make it sharp. For some people this process is a bit difficult (as it was for me), others have no problem with it.
By the way, this way to view is not an exotic feature of Quat, in the Internet such stereo images are widely used. This way it is possible to display a photograph of a landscape three dimensionally, for example
To start a calculation in the text only versions, a so-called "initialization
file" is needed, in which the object, the view, the colors and intersection
planes are defined. This file (and all that are "included", see below)
are simple text files, which can be written in any common text-editor,
(for example, "edit" with DOS, "notepad" with Windows, "vi" with Unix).
The syntax used is very simple, in every line there is a keyword with a
list of parameters following. (e.g., "viewpoint 0 0 1" defines the viewpoint)
In this initialization file you can include other files automatically (by using the keyword "include"). This makes it possible to define color ranges seperate from fractal data or intersection planes (in files with a different suffix, e.g.). If such a file is included at a special location in the initialization file, it is the same as if its contents would be typed in at this location. The initialization file is only required to start the calculation. Once a PNG file has been created, they aren't needed any more. The fractal data is stored in the PNG file itself an can be reconstructed from it.
Suggestion of how to separate the data in different files:
INI this file suffix type is given by Quat. Is is the initialization file. All the information needed to generate an image can be saved in it. It is also possible to "include" other files (as suggested here).
OBJ Fractal parameters determining shape and viewpoint, from which the object is shown and some raytracing parameters, which specify the illumination, are defined in files with this suffix.
COL stores color or color-range of the object
CUT These files serve for the declaration of a intersection plane configuration. (in most cases probably only a single intersection plane).
This concept may appear a little bit complicated on the first view, but it offers high flexibility. For example, it allows to give the fractal other colors very quickly, or to add some intersection planes without changing something on the object-definition itself: Only one line in the INI-file has to be changed. But for those who still think this to be too complicated: You can use all keywords in the initialization file, too.