(05-15-2018, 11:16 PM)Espermaschine Wrote: Here is an image i made.
Circle on the left has a strokewidth of 1px.
Circle on the right has a strokewidth of 0,1px.
Circle in the middle has a strokewidth of 0,5px.
How is this even possible ?
- Assume you have a 10x10 square filled with white. Let's say its average luminosity is 1.
- Stroke over it a 5px-wide black line. Half the pixels will be black, and the average luminosity will be .5
- Now do the same with a 1px square (one pixel). If it is stroked over by a .5px line, its luminosity should be .5, too, so you can fill it it with a gray with a luminosity of .5
- In practice you can evaluate the area of a pixel covered by the stroke, and thus figure out the equivalent gray. This of course also holds for the edge/anti-aliasing pixels.
In reality things are a bit different, its more a matter of partial opacity than of luminosity, and the gamma correction can make the result non-linear.
