Comparing six lenses at 135mm

[CarbonR]

Here are 6 lenses tested at 135mm, one zoom in KAF2 mount, one prime in K, and 4 different Takumars in M42. All pics were taken on tripod, RAW > jpeg conversion only, crop100%, the lighting (sun) was almost constant during the test, a little bit more shining for the forst two lenses, I volontary corrected nothing.

It's here (2000x2000, 1800ko) : http://pix.nofrag.com/d/f/c/78d884bf27c0f5782f75c165b964f.html

[Spotmatic]

Thanks for the great comparison work! Nice to see the differences.

While the old pre-set Takumar (I have that one also, also in Asahiflex mount and in chrome) seems a little less sharp it still strikes me that this old lens has almost no red fringing around highlights in the lens of your subject, while the newer ones do. Do you have an explanation?

I saw the same in older, non-MC lenses. For instance the Jupiter-11 and Tair-11-2 are old and not multicoated, but there's also no fringing to be seen. Strange, but I do not have much optical knowledge. What do the experts on this forum say?

[no-X]

Many forms of purple fringe are caused by axial (longitudinal) chromatic aberration.

[Spotmatic]

[no-X]

The most common form of CA is transverse CA = contrasty lines near image borders are surrounded by colored lines (red/cyan). This form of CA can be easily corrected in post-processing. This form of CA is visible mostly in in-focus areas. Changing aperture doesn't fix it. It doesn't influence bokeh.

But axial CA is almost invisible in IF areas, but it displays in (even slightly) out of DOF areas. Because this form of CA doesn't affect in-focus objects, it doesn't affect lens resolution - and the resolution was (and is) the main priority for many manufacturers. Why to care for axial CA, when ISO 12233 chart, kodak tables etc. must be in focus while picturing them, so this defect won't show?

Priority is resolution and correction of transverse CA - these can be easily measured, they are visible on testing charts and most of reviews judge lenses according to these aspects. Axial CA can't be easily measured, it affects bokeh (bokeh can't be measured too), so it is no wonder, that correction of axial CA is not the main priority for manufacturers.

I'm not expert, but I think that axial CA is a fallout of other corrections, especially of those, which increase sharpness. Majority of old (40+) lenses don't show this issue.

[indianadinos]

Hello,

Thanks for the great explanation ...

Best regards

[aoleg]

[no-X]

In this case, there would be no difference in purple fringe intensity between different lenses and existance of purple fringe wouldn't be dependant on DOF. I wouldn't be dependant on aperture settings, too. But in reality, it depends on particular lens, aperture value and DOF. Digital sensors can amplify these defect (because they can be more sensitive to specific wave-length), but they aren't the cause.

Sensor bloom and purple fringe are two different phenomenons. It would be quite difficult to explain it in English for me, so here is scientific explanation...

[Spotmatic]

On a side note: at home I have cable TV with a HD Receiver and a Samsung HD Plasma screen. Some of the channels I like are National Geographic HD, Discovery HD and History Channel HD. I cannot remember things like CA from the past, but in several documentaries (apparently shot with digital HD cameras) I see green and purple fringing all over the place, most notably in wide angle shots. Is this a sign of not-so-good lenses being used?

[ChrisLilley]