06-04-2011, 09:43 PM
about the dithering issue, i was refering to maneko's edit.
but using the dithering on that also carries theoriginal's problem to an extent wich is shading that body with a gradient. you see, the main problem of using a gradient is that you dont reflect said body's surfaces being hit by the lightsource(giving it defined highlights, rendering it completely flat.
take a cube for example. a cube would have one of its faces being directly in front of the lightsoyrce applied to it, wich would make it brighter than the rest of the faces. at the same time, the face wich is in the exact oposite of the light source would look darker because the bcube 's shape would cast a shadow over said face. finally, there would be two remaining faces that would be partially hit by the light and therefore would cast a color that would "theorically" be between the brightest color and the darkest color.
on more complex objects it works in the same way, although considering that unlike a cube, round surfaces would require a more subtle, less rought transition between highlights and shadows.
when it comes to sprites, you dont need to really worry about having every surface properly afected by a lightsource, but yet you do need to apply said shadows on key places so it would be easier for the viewer to notice the volume and shape of your character/object. specially if you cant aford to use a basic linework to determine an object's shape because its either too small or because it would look bad to do so.
solid blocks of color would always, ALWAYS have a higher contrast and provide much mroe readbility rather than working your way with dithering patterns.
and its most likely that if you consider a dithering pattern as a need in your shading process then you probably suck at dithering and would ultimately make a huge mess that would look like shit so for the 273891273981723983127th in my life, do not dither.
but using the dithering on that also carries theoriginal's problem to an extent wich is shading that body with a gradient. you see, the main problem of using a gradient is that you dont reflect said body's surfaces being hit by the lightsource(giving it defined highlights, rendering it completely flat.
take a cube for example. a cube would have one of its faces being directly in front of the lightsoyrce applied to it, wich would make it brighter than the rest of the faces. at the same time, the face wich is in the exact oposite of the light source would look darker because the bcube 's shape would cast a shadow over said face. finally, there would be two remaining faces that would be partially hit by the light and therefore would cast a color that would "theorically" be between the brightest color and the darkest color.
on more complex objects it works in the same way, although considering that unlike a cube, round surfaces would require a more subtle, less rought transition between highlights and shadows.
when it comes to sprites, you dont need to really worry about having every surface properly afected by a lightsource, but yet you do need to apply said shadows on key places so it would be easier for the viewer to notice the volume and shape of your character/object. specially if you cant aford to use a basic linework to determine an object's shape because its either too small or because it would look bad to do so.
solid blocks of color would always, ALWAYS have a higher contrast and provide much mroe readbility rather than working your way with dithering patterns.
and its most likely that if you consider a dithering pattern as a need in your shading process then you probably suck at dithering and would ultimately make a huge mess that would look like shit so for the 273891273981723983127th in my life, do not dither.