Dear Garth,
I am a newbie using XenoDream 2.6 running under Windows 7 Pro and slowly working through the “Using XenoDream 2.6” manual.
I have a cursory understanding of how color works with raster images and with vector images. Raster images are made up of collection of color pixels whereas vector images are made up of math-defined “elements” (i.e., points, lines, and shapes) for which a color property is specified as either a color or a gradient.
I am trying to understand how color works in XenoDream. Pages 6-7 of the “Using XenoDream 2.6” manual say that images are created with “points” (which can be clearly visible as objects are progressively displayed in the Preview Window of the Main Window). What are these “points”? How do these “points” relate to raster or to vector images?
Thank you!
Doctor T
The points are random samples in the 3D object. For rendering we have a grid of pixels and additional information such as depth and shadows. For each point we find the nearest pixel, see if it's in front of the current depth at that pixel and if so, put the new color and depth into that pixel. (With color anti-aliasing, or transparency, there is some blurring of the colors at the pixel, depending on relative depths.) So a raster image results from accumulating these point samples. But the image only has the flat colors that depend on the position within each holon. So the lighting calculations are needed as a second step to see what the effects of shining lights on the object would be. The raster image you see in rendering is an intermediate view with raw colors.
(07-23-2017, 07:22 AM)Garth Thornton Wrote: [ -> ]The points are random samples in the 3D object. For rendering we have a grid of pixels and additional information such as depth and shadows. For each point we find the nearest pixel, see if it's in front of the current depth at that pixel and if so, put the new color and depth into that pixel. (With color anti-aliasing, or transparency, there is some blurring of the colors at the pixel, depending on relative depths.) So a raster image results from accumulating these point samples. But the image only has the flat colors that depend on the position within each holon. So the lighting calculations are needed as a second step to see what the effects of shining lights on the object would be. The raster image you see in rendering is an intermediate view with raw colors.
Dear Garth,
Thank you for your reply.
You said “The points are random samples in the 3D object.” So each point has an x,y, z position within the object? And, enough points create a shape?
Then you said there is a “grid of pixels.” Is it a 2-D grid? Or is it a 3-D grid (that is, a 2-D grid with layers)?
Doctor T
(07-23-2017, 07:56 PM)DoctorT Wrote: [ -> ]You said “The points are random samples in the 3D object.” So each point has an x,y, z position within the object? And, enough points create a shape?
Yes.
"Then you said there is a “grid of pixels.” Is it a 2-D grid? Or is it a 3-D grid (that is, a 2-D grid with layers)?"
It is a 2-D grid, not with layers but several types of data. For each pixel there is a color, and also a depth of the closest point that landed in that pixel. The depth corresponds to z position. (There can also be a region for each pixel, which allows holons to have different lighting effects, but that's a more advanced topic.)
Dear Garth,
Thank you for your reply.
I am trying to understand the difference between the two coloring methods: holon and metamorph.
Ignoring lighting, as points are mapped to the 2-D pixel grid, there are two pieces of information recorded: (1) the depth, which is determined by the z position of the closest point, and (2) the color, which I assume is determined by the information entered in the Color Page, the most important of which is the coloring method.
Although I read the description of the two methods on pages 66-67 of the “Using XenoDream 2.6” manual, perhaps you can -at your convenience- elaborate a bit more.
Thank you!
Doctor T
Yes, the Color page controls determine how the points in the object get their intrinsic color.
In the Metamorph method, each metamorph produces three patterns which vary with x,y,z position, and you choose one of these patterns or a mixture, and map it to a gradient to choose the colors in the pattern. The patterns generally follow the shape of the metamorph in some way, or they may just be stripes in some direction. This makes sense for coloring constructors. For iterators this will not follow the shape of the iterator holon, so it usually looks better to let it inherit patterns from other holons.
When there are no constructors, as you have seen, the iterators make a more complicated fractal shape. Stripes or other geometric patterns can be used, but it's usually nicer to have coloring that follows the shape. For this you need something like the Holon Sequence method. The patterns come from the unique sequence of iterations that define each point, giving each pixel a possibly unique color. Unfortunately this is hard to control in any nice way, so setting the colors tends to involve either randomizing or trial and error adjustments. Some details of this method are given in the Help and the user guide so I won't say more about it unless you have specific questions.