12-15-2014, 03:44 AM
It's the asphere. This is a ray aberration plot containing only spherical aberration with severe enough spherical to not care about other aberrations.
Spherical is created due to the variation in the angle of incidence across the surface of a curved piece of glass. With an asphere you can control it. The asphere equation (the one appearing here, not a Qbfs asphere or other type) has terms of different orders. The high order terms rule absolutely away from the center of the lens and the lower order terms typically 'supress' the high order terms near the center of the lens. If you optimize like a monkey and just let design software (i.e Zemax or Code V) vary all of them it may walk the design into a space where the spherical is "nuked" at the edge of the pupil/element but not so much towards the center, or it is overcorrected in the center.
[sharedmedia=gallery:images:1315]
An aspherical element which has a maximum or minimum away from the vertex (think schmidt corrector plate) is very difficult to manufacture. We tend to avoid designing these, though they can be made and with increasing ease through either molded aspheres or MRF polishing. The fourth element of the 50/1.4 asph from samyang (see here) cannot be cheaply made through normal manufacturing methods, the only economical way to make it is through molding.
It also has "set edges" to allow it to be easily braced within the mounting stuff in the lens barrel.
Samyang's method to hit their price points is to 'go easy' on QC, and to aggressively shrink their optics to use as little material as possible.
When you design a lens in cad (e.g Code V) you specify an f number and a sensor size (or field, as we like to say in the industry). This does not mean the lens does not work at a faster f number or a wider field, it means we don't care what happens outside spec.
If rays solve outside f/1.4 (or whatever this lens was designed for, marketed and on-the-barrel F number does not mean it is that fast or that wide/telephoto) they will contribute to the image unless they are vignetted through e.g a mechanical aperture (iris, flocking) or an optical aperture (cutting the lens off).
The edges of several of the elements in this lens are very ugly (i.e abrupt and sharp edges), I would suspect rays are passing through them that weren't planned for, and those are manifesting in the out of focus highlights.
These things are usually not tested for / known about in design, since there is no metric for evaluating them.
I get paid to design and build very abormal optics, feel free to ask me any questions related to stuff like this.
Spherical is created due to the variation in the angle of incidence across the surface of a curved piece of glass. With an asphere you can control it. The asphere equation (the one appearing here, not a Qbfs asphere or other type) has terms of different orders. The high order terms rule absolutely away from the center of the lens and the lower order terms typically 'supress' the high order terms near the center of the lens. If you optimize like a monkey and just let design software (i.e Zemax or Code V) vary all of them it may walk the design into a space where the spherical is "nuked" at the edge of the pupil/element but not so much towards the center, or it is overcorrected in the center.
[sharedmedia=gallery:images:1315]
An aspherical element which has a maximum or minimum away from the vertex (think schmidt corrector plate) is very difficult to manufacture. We tend to avoid designing these, though they can be made and with increasing ease through either molded aspheres or MRF polishing. The fourth element of the 50/1.4 asph from samyang (see here) cannot be cheaply made through normal manufacturing methods, the only economical way to make it is through molding.
It also has "set edges" to allow it to be easily braced within the mounting stuff in the lens barrel.
Samyang's method to hit their price points is to 'go easy' on QC, and to aggressively shrink their optics to use as little material as possible.
When you design a lens in cad (e.g Code V) you specify an f number and a sensor size (or field, as we like to say in the industry). This does not mean the lens does not work at a faster f number or a wider field, it means we don't care what happens outside spec.
If rays solve outside f/1.4 (or whatever this lens was designed for, marketed and on-the-barrel F number does not mean it is that fast or that wide/telephoto) they will contribute to the image unless they are vignetted through e.g a mechanical aperture (iris, flocking) or an optical aperture (cutting the lens off).
The edges of several of the elements in this lens are very ugly (i.e abrupt and sharp edges), I would suspect rays are passing through them that weren't planned for, and those are manifesting in the out of focus highlights.
These things are usually not tested for / known about in design, since there is no metric for evaluating them.
I get paid to design and build very abormal optics, feel free to ask me any questions related to stuff like this.