That's dead sexy! haha.... Very nice.
Lab snacks!!!!!!!
The video that I made months back and the posts about correcting diode beams was made in conjunction with an evaluation of the (then) new Oclaro diode. The argument is that spherical correction optics as opposed to aspherical optics, are going to introduce some spherical aberration. There are four surfaces in a lens pair and each will contribute to the aberration. Even a plane parallel window in a converging beam will introduce spherical aberration. This might seem strange, but because the outer rays in a converging beam are angled inward more steeply they will be stepped outward further, during the passage through the bulk of the glass after the entrance to the higher index glass (vs air). Upon leaving the other side of the window these rays will once again return to their former converging angle, but will take a little longer to come to a focus due to the offset within the glass. Obviously, the thicker the glass the more pronounced the offset and the more spherical aberration that will have been introduced.
This "thick lens" effect is combined with the spherical aberration that the curved surfaces introduce as well. It would be daunting to try and calculate the total amount of spherical aberration from a given cylinder pair and it is probably not necessary. There is undoubtedly some residual spherical aberration in a diode beam after the collimator because even though that optic is aspherical the source is finite and the asphere will avoid introducing spherical aberration only from a point source. With a variety of cylinders on hand you can pick and choose as well as flip the lenses to introduce the optimal amount of spherical aberration (this can be negative as well as positive) to counter what comes out of the collimator. This is empirical, but practical.