I think the main issues with Kirby's shading is that it comes off as a close spotlight and appears linear- which adds to the banding effect.
Real quick, to get a graphical intuition of why banding happens in this context:
In a simplified sense, the amount of brightness an area has, due to a directional light (the sun), is related to the dot product between the surface normal(N) and the light direction(L): |N||L|cos@ = brightness, where @ is the angle between the vectors. Note the "cos@", that's the key. Given that the angle between the light and spherical surface changes linearly, there is a one-to-one relation between the cos@ curve:
What's the point? The output of cos@ isn't linear, therefore, shading isn't linear... The space between shades shouldn't be evenly spaced out on a spherical object. No matter how many colors you do or don't have, you'd still issues. (Though, more shades makes it easier to accidentally create banding)
This has the obvious effect of appearing visually inconsistent with the volume.
...I don't draw enough to give you a non-mathematical answer =/...
Real quick, to get a graphical intuition of why banding happens in this context:
In a simplified sense, the amount of brightness an area has, due to a directional light (the sun), is related to the dot product between the surface normal(N) and the light direction(L): |N||L|cos@ = brightness, where @ is the angle between the vectors. Note the "cos@", that's the key. Given that the angle between the light and spherical surface changes linearly, there is a one-to-one relation between the cos@ curve:
What's the point? The output of cos@ isn't linear, therefore, shading isn't linear... The space between shades shouldn't be evenly spaced out on a spherical object. No matter how many colors you do or don't have, you'd still issues. (Though, more shades makes it easier to accidentally create banding)
This has the obvious effect of appearing visually inconsistent with the volume.
...I don't draw enough to give you a non-mathematical answer =/...