by Ian Burley
DxO Analyzer tests
Technical testing
At Four Thirds User we use DxO Analyzer 3.0 for computer-measured technical evaluation of lens performance. The system does not rely on error-prone human evaluation of technical test images. Naturally, we also evaluate real world images, but DxO Analyzer is an excellent solution for producing accurate test data.
Chromatic aberration
Extremely low, even at full aperture and where found, measured at well under one pixel. This is a very good result indeed.
Vignetting
A more than acceptable performance at full aperture sees just a 14% reduction in brightness at the extreme corners of the frame. This reduces to just 4% by f/4, where it stays for the rest of the aperture range. Again, this is very good.
Geometric distortion
Even magnified 5x the Sigma 105 Macro exhibits very low geometric distortion
At only -0.06%, this level of pincushion distortion is, basically, insignificant.
Actual 1:1 closest focus shot, un-cropped, using the Sigma 105mm Macro
DxO Analyzer's distortion measurements are made at non-macro distances, but I'm pleased to report that even at closest focus, there is very little visible geometric distortion. The example shot above is un-cropped and taken at closest focus, approximately four inches (100mm) from the end of the lens. The notches in the checkerboard pattern are actually imperfections in the ink-jet print of the test target. Each square is roughly 4mm wide. This is an excellent result.
Sharpness
At present we evaluate lens sharpness using DxO Analyzer's BxU (Blur eXperience Unit) measurement; a proprietary unit of measurement. The smaller the BxU unit, the less measured blur there is, so the sharper the image. We are in the process of adding a MTF (Modulation Transfer Frequency) chart to our DxO Analyzer test suite and this will produce more widely recognised test data.
In the mean time, using BxU results, at full aperture, with the Sigma 105mm Macro there is a noticeable, if modest difference between the centre and corners of the frame, but this disappears by f/4. Sharpness is at its best at f/5.6 and this is maintained well up to and including f/11, from which point there is a fairly steep, if not unusual, drop towards f/22. The aperture range of the Four Thirds version of the Sigma 105 Macro is limited to f/22. Some other versions offer an f/32 minimum aperture, though by this point a great deal of sharpness will have been lost through the effects of diffraction.
We decided to compare data from the Olympus Digital Zuiko 50mm f/2.0 Macro, tested on an Olympus E-1. The Olympus optic manages a tighter centre to corner match throughout its range and the sharpness fall-off is less steep. It's a more impressive technical performance, but the DZ 50mm f/2.0 macro is not only more expensive, but doesn't deliver a 1:1 closest focus reproduction ratio, only a 1:2 ratio.
Just for academic comparison, we also show the 42mm BxU data for the new 14-42mm kit lens, tested at its longer end of 42mm. Tested on the same E-510 body, the result indicates that the kit zoom is slightly sharper than the Sigma 105. The cost is in inferior chromatic aberration and geometric distortion, which is not bad for a kit lens, but the Sigma 105 Macro is significantly better in this respect.
So the Sigma 105 Macro is optically very good indeed, but not quite a class leader in the sharpness stakes.