There are two rows of rectangles using these PNG images as textures; the images and rectangles are all 200x200. In the upper row the transparency is set to 0.0F and in the lower row it is set to 0.3F. For columns 1 and 3 (without alpha) the lower image renders differently because the Material2D transparency is used; which is not the case for columns 2 and 4 where the image alpha overrides this. Hence in columns 2 and 4 the upper and lower rendered output is the same. The rendered image color for each pixel is obtained from a gray scale texture by multiplying the emissiveColor by each pixel's intensity; see VRML annotated reference manual "section 4.6.11 Textures" for more information (MPEG-4 being based on VRML for much of it basic function).
To show the rendered output is really derived from emissiveColor a ColorInterpolator is started at t=5s and changes the color of the right hand set of image shapes repeatedly over a 10s period.
Snapshot of rendered output.
![]() |
Snapshot taken during interpolation of colors.
![]() |