Note 1:
If you want me to upload any full images or detailed sections of an image, ask. I've tried to post images, so my points could be followed and verified. With the reduction of image detail density required for web posting, some discussion points may not be easily seen. I used enlargements in PS and LR for analysis.
Note 2:
My near infinity test image test analysis and image mosaic construction is not complete due to my arm injury. I will complete that missing detailed "Results" section above and the conclusion paragraph below after I am healed. From initial analysis, it should not change my opinion stated below. If I find contradictions to that, I will post a "red letter" edit and an additional posting.
Note 3:
In normal use of my Takumar, I normally use about -1/2 to -1/3 of a stop of exposure. I did not do so in these tests.
Note 4:
I still have plans for teleconverter vs. extension tubes vs. “vanilla” comparisons. I will also run a set of tests on my Vivitar 90-180 flat field lens.
Conclusion:
My initial working hypothesis was that all 100mm -ish macro lenses have good IQ. Differences as to what constitutes a better lens must be made based on other attributes. I feel that I was mostly right.
Using any of these lenses on my K3 without comparing the IQ it to another lens leaves me happy with the results. I would not wish for a different lens. Therefore any ranking of these lenses must be based additionally on other non IQ parameters. However, I did compare the IQ on these lenses. ;)
The detailed relative resolution comparison of these lenses showed they were all about the same, very, very good. Extreme enlargement of the closest focus point images, there may be a very, very small difference that changed between the .5X and the 1X reproduction ratio images. One time the Takumar was the best by an exceedingly small margin, the other time it was the worst by that same very, very small margin. This was probably due to a failing in my manual focusing abilities. This could only been seen under extreme enlargement, to the point of pixelation. Looking at something one quarter of a millimeter wide and enlarging it to fill 23" monitor is beyond expected macro lenses use. Macro lenses are not designed for that degree of enlargement. A more reasonable approach would be to use bellows for that degree of enlargement. Winner: TIE
Vignetting testing also produced exceedingly close results. The Takumar and the D-FA were almost identical and almost perfect. The Phoenix was 1/16th of a stop darker in the corners than the Pentax built lenses when wide open. That does not merit any negative points. Winner: TIE
Color testing produced a result unexplainable by me. All three lenses produced almost identical measured color values on red and green objects and varied by up to six percent on the blue object. All three lenses behaved the same. Winner: TIE
Purple fringing testing produced almost equal and perfect results with attempts to provoke a defect. Winner: TIE
Distortion and decentering testing produced equal and perfect results amongst all three lenses. Winner: TIE
Test image with .5X magnification produced images almost equally sharp images with very slight differences probably due to my focusing. Winner: Tie
Test image .25X magnification also produced images almost equally sharp images with only very slight differences probably due to my focusing. I was most comfortable with my focusing in this test. With only a moderate amount of magnification small differences were found. In order of sharpness: Takumar, D-FA, Phoenix. Because the differences are small and I cannot rule out focusing mistakes, I declare a tie.
Near infinity test image test analysis is not complete. I cannot finish it until my arm heals. My initial impression was that all were very close is sharpness. Bokeh looked very similar too. I will return to this section when my arm is better.
The Takumar was the easiest to focus manually in sufficient light by virtue of its large focus ring. Winner: Takumar The D-FA by virtue of its wider aperture was easier to focus in less than perfect light. Winner: D-FA
The D-FA has both auto-focus and full time manual focus, the Phoenix auto-focus only, and the Takumar has no auto-focus. Winner: D-FA
With respect to lens automation, the Takumar has only auto-aperture, useless with modern K cameras, and both the Phoenix and the D-FA have fully automated aperture control. Winner: D-FA and Phoenix
The readability of the scales and easy manual focusing, the Takumar is the obvious choice if exacting measurement of reproduction ratios are needed. I have not needed that for a project for over thirty years. In the digital age, end result sizing can be controlled with photo manipulation programs negating the need for precise reproduction ratio measurement on the lens.
Overall these three lenses are all excellent as judged by their results. They are all well corrected full frame lenses used on an APSc camera. The Takumar was designed as a professional lens when the Spotmatic was the professional SLR of first choice (the only popular one having TTL metering). The D-FA is a lens of modern design planned as a high end optic. It is not surprising that these lenses are excellent performers. What is surprising is that the Phoenix can be compared to them.
All the Takumar IQ parameters were almost the same as the D-FA. I know of no advantage the Takumar holds with the exceptions of the scales readability, aperture ring, and the large focus ring. What the D-FA additionally has that the Takumar does not, is the aperture controlled by the camera, weather resistance, and auto-focus with full time manual override. The D-FA also extends to 1X magnification compared to the Takumar's .5X magnification. The Takumar will get relegated to the back of the shelf, only to be used on my bellows. What I will miss is that large focus ring. I usually do not use auto-focus for high magnification shots. The D-FA can also be used as a short auto-focus telephoto lens. This allows me to move my beautiful 105mm f 2.8 Super-Multi-Coated Takumar to retirement alongside my newly retired Takumar macro.
The Phoenix, though cheap and clunky in appearance, competes in IQ at a level on par or just below the two Pentax lenses. How did they do that? Only on the one test at .25X magnification of my wife's miniature collection did the Phoenix show an observable lesser IQ than the Pentax lenses. That deficit is minor. Where this lens fails is in comparison, is in usability. It is a bit harder to focus manually than the Pentax made lenses. It is also hard to read the scales on the lens, the D-FA and the Takumar are better. The aperture ring has only full f stop clicks. I'll be using the D-FA, not the Phoenix, because of the slight IQ difference that can be occasionally seen, weather resistance, and the full time manual focus override. This lens will go for sale or moved to the back of the bottom shelf.
The D-FA lens, IQ wise just about on par with the Takumar and occasionally slightly ahead of the Phoenix, suffers from no aperture ring, a small focus ring, front lens element not deeply (only 5mm) recessed, and a funky lens shade arrangement. With the D-FA, I'll just need to take an extra lens hood with me when I plan on using magnifications greater than .5X. However, it is my choice for the best of the three lenses. It has great IQ, weather resistance, better manual focus ability than the Phoenix, and faster aperture than the other two lenses. It can extend to a reproduction ratio of 1X, which I have never used. Given its speed and ease of use, it also will make a great general purpose short telephoto lens. Plus it is cool looking, very similar to my DA Limited lenses!