Optically speaking, when a designer optimises a zoom lens, he has to make (choose to make) a comprimise. This is because the lens designer has to settle for a an optimisation for the whole range. When he makes a prime lens, He doesn't have to make this compromise, because his lens oly needs to work at one focal length, not a range of focal lenghts
If we look at a single element of a lens (and assume it is "thin") and that element needs to be +6 dioptres the lens designer has the following choices
for example:
+12D front curve, -6D rear curve
+ 6 front curve, plano rear curve
+3 front curve, + 3 rear curve
plano front curve, + 6 rear curve
etc. etc..
Within the design, the "form" of the lens - which really means the combo of front and rear power will effect things like peripheral performance and most of the other abberations
IN addition, the designer can choose to use spherical or aspherical curves, can choose the index of the materials, can choose to use doublets or triplets of a range of different materials
All of these design questions apply to every element in the lens, and also to the whole set of elements
The maths involved in dealing with just two lenses, considering all of the above, whilst optimising for tangential and longitudinal shell error, chromatic aberrations, barrel / pincushion distortion is mind blowing, and really only in recent years have computers been up to speed enough to really help
When you consider the difference in optics, and number of elements between a Prime and a zoom, it is easy to see why the prime has much better optics
The zoom designer also needs to make compromises. The choice of curve for any specific element will be best at one (overall) focal length. That's why you often see zooms go soft, or exhibit a lot of distortion at the long end
That's why primes will be sharper and show less aberrations
- they are easier to design
- the designer makes less compromises
