28mm: CZ, Minolta and Minolta, corner

The lens compared here are the:
Sony "Carl Zeiss" 16-35 f/2.8
Minolta 24-105 f/3.5-4.5
Minolta 28mm f/2.0 prime.



Here the corner pictures:

f-stop	Sony CZ 16-35	Minolta 24-105	Minolta 28mm f/2.0
f/2.0
f/2.8
f/4
f/5.6
f/8
f/11

Quite interesting, isn't it?

The CZ confirms its superb quality. The lack of corner sharpness that we noticed at 16, 20 and even 24 mm is less of a problem, it decreases when zooming in. This is typical of wide-angle zooms (and in this respect, the Sigma 12-24 is atypical).

The Minolta 24-105 is surprisingly good, even if the full aperture (f/4) is not perfect, f/5.6 is quite usable and f/8 just as good as the CZ.

The 28mm prime is the worse of the three: corner shading is enormous and corners are never really sharp. Other tests showed that the reason is field curvature. That makes this lens totally unsuited to landscape photography for example (quite the contrary of the 20mm prime). Does this mean it is a bad lens? No, it only makes it an imperfect lens, but is has other advantages when used as it was intended, i.e. for available light photography. See next post.

28mm: CZ, Minolta and Minolta, center.

At 28mm, the lens compared will be:
Sony "Carl Zeiss" 16-35 f/2.8
Minolta 24-105 f/3.5-4.5
Minolta 28mm f/2.0 prime.

The Minolta 28mm f/2.0 prime is a rather old lens, which is not available new any more. Sony only kept the 28mm f/2.8 on offer. It is a very special lens, designed for low light and giving a very particular look in available light photography. Some people refer to this look as "Minolta colors", but there is more to this story. The lens warrants a more detailed review than the simple comparison which is posted here. Here it simply looks less sharp than the CZ...

Here again, you need to click the pictures to see the full resolution. It should open in a new window (or new tab if you use them). This make it easier to open a few pictures and flip between them for comparison. Personally, I use tabs so I just need to click on them to flip back and for between the pictures I want to compare. The name of the pictures gives lens, focal length and aperture.


Here the center pictures:

f-stop	Sony CZ 16-35	Minolta 24-105	Minolta 28mm f/2.0
f/2.0
f/2.8
f/4
f/5.6
f/8
f/11

The pictures from the prime were not taken on the same day and are a bit darker, but even when taken in exactly the same conditions, the prime gives warmer colors than the CZ zoom. The CZ is dead sharp full open, the 24-105 and the prime need to be stopped down a bit. The corners tell a very different story...

Bonus: 12mm (Sigma 12-24)

The Sigma 12-24 f/4.5-5.6 is a unique lens, no other lens goes to 12mm on full frame. That makes it impossible for me to post comparison pictures at 12mm.

There have been many reviews of this lens over the net. I found the ones of Kurt Munger and Juza Ea to be amongst the most interesting.

When reading various reviews about this lens, one might wonder if everybody tested the same lens, considering how different the results can be. There seems to be a lot of sample variation amongst copies, but the lens is also particularly difficult to test. Shooting resolution targets, for example, requires extreme care: because of the angle of view, you need to be so close that any difference in flatness of the test picture will cause problems. I am literally talking fractions of millimeters here, and insuring that a test picture (typically about 1 meter base) is flat with a precision better than one millimeter on its complete surface is not easy. You can't do that by just hanging a newspaper on a wall…

The lens may also respond differently depending on the camera, for example for corner shading (vignetting). Just as Kurt Munger, I found the full aperture to be almost unusable because the corners are too dark. We both tested on a Sony A900. Juza Ea, on the other hand, says that full aperture is perfectly usable and publishes pictures with far less corner shading. He used a Canon EOS 1D MIII. It is quite possible that the two cameras respond quite differently, at 12mm the peripheral rays will reach the sensor at an angle which causes problems with digital sensors. The sensor technology is different, so could be the response.

I did not test the lens at full aperture or f/5.6, which I find barely usable. Below are the 12mm pictures at f/8 and f/11.

Center: f/8 and f/11

Corner f/8 and f/11:

(remember to click the pictures to see a 100% crop in a new window)

Conclusion? In my opinion, 12mm is quite usable at these apertures. From the other tests, we have seen that the lens is also close to the CZ 16-35 at 16mm, but that there are much better lenses at 20mm or 24mm (Sigma FTM confirms poor corner sharpness at 24mm), so it is probably best to mainly use the lens between 12mm and 16mm. Luckily, this is the most interesting part of the lens. And to put things into perspective: when Nikon made a 13mm f/5.6 about 40 years ago, it was such a difficult lens to make that they only made a few hundreds at a the price of a medium-sized car. You can read about it at the Kenn Rockwell site.

24mm: CZ, Minolta and Sigma, corner.

Second series of the 24mm comparison, the lens compared are:
Sony "Carl Zeiss" 16-35 f/2.8
Minolta 24-105 f/3.5-4.5
Sigma 12-24mm f/4.5-5.6 EX.


24mm corner crops:

f-stop	Sony CZ 16-35	Minolta 24-105	Sigma 12-24
f/2.8
f/4
f/5.6
f/8
f/11

Here the CZ and the Minolta give identical results at f/11, but the Sigma is definitely below the other two, although we can probably only tell the difference because the magnification is so high. f/8 is still usable, but anything below that and the performance is visibly degraded for the Minolta and Sigma. At f/4, the Minolta performance is so poor that it actually shows on the small pictures above.

The CZ is still ok at f/5.6, but f/4 and f/2.8 are quite poor. In truth, however, this is rarely important: only the extreme-extreme corners are affected (the top right of the picture is noticeably sharper than the bottom left) and this is not likely to show in actual use, where the corners will be dark (for night scenes) or outside of the depth of field.

Lateral chromatic aberration ("purple fringing") is visible in the 3 lenses (look at the tree trunk). This sort of aberration does not go away when one stops down and is very difficult to suppress for ultrawide. The CZ is not better than the other two in this respect.

The Minolta 24-105 confirms its use as an ideal travel lens, provided you stop it at f/8 or f/11.

Next: 28mm. CZ, Minolta 24-105 and Minolta 28mm f/2.0 prime.

24mm: CZ, Minolta and Sigma, center.

At 24mm, the lens compared will be:
Sony "Carl Zeiss" 16-35 f/2.8
Minolta 24-105 f/3.5-4.5
Sigma 12-24mm f/4.5-5.6 EX.

Presenting the results starts to be a bit more difficult, so I decided to write a table with reduced-size pictures. You need to click the pictures to see the full resolution. It should open in a new window (or new tab if you use them). This make it easier to open a few pictures and flip between them for comparison. Personally, I use tabs so I just need to click on them to flip back and for between the pictures I want to compare. The name of the pictures gives lens, focal length and aperture.


Here the center pictures:

f-stop	Sony CZ 16-35	Minolta 24-105	Sigma 12-24
f/2.8
f/4
f/5.6
f/8
f/11

At f/11, all 3 lenses give identical results. At f/8, we can probably only tell the difference because the magnification is so high.

The CZ already attains maximum sharpness at f/4. The contrast is higher than the other lenses.

Still, when taking pictures in daylight, there is nothing wrong with f/8 which makes all three lenses perfectly usable for landscape pictures. The Minolta 24-105, in particular is small and light, which makes it an ideal travel lens. How is it in the corners and at other focal lengths?

Next: corner performance.