Based upon the other interest shown in the Samyang 14 thread, I thought those interested in this economy ultra wide angle lens would like to see how it compares to the legendary Zeiss 21. The objective was two fold:
Firstly the issue was to see if PT Lens visually diminished the sharpness of the Samyang (Rokinon branded) 14 f2.8 image captured on a 5DII. This distortion correction software for the Rokinon is not embeded into the Breezebrowser RAW conversion, as it is for the Zeiss. Rather it is a tool added as a Photoshop filter in post. The same test was performed with the Zeiss Distagon 21 f2.8, but PT Lens action was done as part of the RAW conversion. By switching back and forth one can readily see the changes in barrel (mustache) distortion correction being applied by PT Lens. While the Zeiss is being used as a control for the Rokinon application of PT Lens, one can contrast the overall sharpness and FOV of the Rokinon Vs the benchmark of wide angle lenses, the Zeiss 21.
These are presented in a larger size than I normally post so that one can see both the full frame effects of distortion inherent in these lenses as well as any changes in the sharpness of detail. Other than final, default sharpening of Topaz InFocus applied uniformly to all 4 images, there is no editing except downsampling to JPEG presentation size; I even left in the sensor dust bunny at the top. The Palace of Fine Arts in San Francisco was convenient, well lit target, and has lots of parallel straight lines for this test.
First the Rokinon 14 WITHOUT PT Lens, shot at f8
the identical Rokinon 14 TIFF file processed the WITH PT Lens
Open both images and tab them back and forth. Notice how the edges of the frame come forward and the center of the frame gets further away? This has the overall illusion of making the subject seem a bit smaller (distant). It looks as if there is some vignetting compensation too, but there is no way for the program to know what f stop is being used. I studied this at 100% looking for some obvious decrease in sharpness, and decided if any, it was pretty minimal. Some pixels may be added where the image has to expand a bit at the edges, but it seems as though most of the distortion correction is in taking out extra pixels in the center. The fine detail subjectively has the comparable resolution in both images.
Next we compare the a similar style mustache distortion correction by PT Lens on the output of the Zeiss Distigon 21 f2.8. Shot at f 5.6 (optimum sharpness for this lens) WITHOUT PT lens.
and the same RAW file converted WITH PT Lens as an automated Breezebrowser filter.
the distortion is not as accentuated in the center of the frame, but the action carries more towards the edges (see the dust bunny move). My assessment of 100% pixel peeping was that there was no change in image sharpness.
Finally compare the Rokinon Vs the Zeiss: image 1 Vs 3 and image 2 Vs 4. Obviously the 14mm image has a wider fov and smaller details than the 21mm image, don't expect them to be identical. Downsampling of the Zeiss shot could be done with cropping of the Rokinon, but thats another software action affecting IQ. Suffice it to say that the Rokinon does pretty well in comparison, as it is less than $400 at Amazon and the Zeiss more than $1800.
I am in no way affiliated with PT Lens, but its a reasonable $25 program.