Light Test: -samples vs -super

Below is an old test conducted in 2012 to figure out the finer differences of super and samples. Around this time, there was a lot of debate around the results of the different lightmap blurring algorithms and level designers were divided between their preferences.
With the computers at the time, it was a tedious task to conduct tests like this on completed maps with compile times taking upwards of 8 hours for the final version.

I Enjoyably liked the results of -samples 8 > -super 3, the compile time was roughly the same for these switches but the results were different.
To make sure I couldn't spike the results I only used this skybox shader. Where I used the q3map_sun flag without sunExt due to testing samples done from the compiler alone and not from the sky shader.

textures/del_test/sky2
{

q3map_sun 1 0.788 0.347 100 135 50
q3map_noFog
surfaceparm sky
surfaceparm noimpact
surfaceparm nolightmap
surfaceparm nomarks
surfaceparm nodlight

skyparms env/box365/box365 - -

nopicmip
nomipmaps

}

Here are my results;





In conclusion, It's quite clear that -samples 8 produces a smoother result while the compile times were the exact same as -super 3, however the -super 3 produced an uneven shadow which seemed darker. I know for sure now why I was told to use -samples 8, the shadows are much smoother with the exact compile time. Another thing to note is that -samples has no cap which means you can go much higher and use -samples 12 or 16 or even 99. It really depends on the effect you are going for though.