From: Hatch (hatch_at_in.com.au)
Date: 2004-11-06 11:41:56
> Most consumer level 3d accelerators at the moment have a filled > triangle as their best supported primitive - you feed the hardware the > three corners in screen-space, and it fills the contents. Older > hardware only supported flat shaded (filling with a single colour), > then as you go down through the years you add gouraud shading (just > interpolating three colours at the three corners), texture mapping, and > eventually modern GPUs let you run a little program per pixel for the > triangle being filled. > > So yes, everything on the screen is made up out of triangles, at least > where hardware is involved. (software renderers can be very different > beasts..) So the coder would pass three sets of coordinates to the graphics card and would get a filled triangle on the screen? This sound like it would be very fast for building up an image. So if I used this method my hardware would have 320 X coordinates and 200 Y coordinates. What if one of the points is outside of this area, say -20,50? Is the hardware expected to be able to deal with points outside off 200 x 320 and fill the part of the triangle that is in the visible area or does the coder somehow take care of this? Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.