11-08-2011, 05:01 PM
[quote name='Studor13' timestamp='1320762957' post='12764']
OK, so for those out there who rely (solely) on Photozone's resolution numbers, here are my own non-scientific results at 35mm (large files 6MB) - about 2m from subject to lens (60x focal length):
18-35mm @f4.5
http://www.studor13.si/music/photos/PZ/1835-35-f45.JPG
35mm AIS @f4
http://www.studor13.si/music/photos/PZ/35AIS-f2-f4.JPG
18-55mm II @f4.5
http://www.studor13.si/music/photos/PZ/35Kit-f45.JPG
18-55mm VR @f5
http://www.studor13.si/music/photos/PZ/35VR-f5.JPG
The results at other focal lengths are in favor of the 18-35mm universally, except wide open at 18mm.
As I have said (twice), it is beyond believe that PZ can claim that the resolution in the corners of the 18-55 VR at 35mm wide open can be almost double that of the 18-35 also at 35mm and also wide open!
[/quote]
I would not give too much about lens test. I test all my lenses in reallife and give them back if I dont like what I am seeing. There is not a single formal test procedure which gives fully reliable results for every lens they test or fully mirrors real life performance when shootimg 3D subjects as opposed to 2D test targets. Mostl formal tests nevertheless provide some adequate guideline what to expect from a lens, though not always. That being said, I found photozone to be more reliable than others. (Even though I fundamentally disagree on the canon 17-85. My copy provided after CA correction sharp results from corner to corner at 17mm, which I proved to Klaus in a high res 15mp sample. The lens problem at 17mm is high field curvature, which hardly becomes visible in 3D environments)
Therefore I strongly suggest to make your own real life tests and buy according to your own experience. Furthermore: I dont think that colour or contrast is a meaningful criteria to differntiate lens performance anymore. All modern lenses are multicoated and provide a very high degree of contrast in most situtations. colour differences are minimal and strongly dependet on the light. This is why your photos look great: because they are shot in great light. In any event, if colour diffences of lenses exist, are irrelevant if you process from raw.
OK, so for those out there who rely (solely) on Photozone's resolution numbers, here are my own non-scientific results at 35mm (large files 6MB) - about 2m from subject to lens (60x focal length):
18-35mm @f4.5
http://www.studor13.si/music/photos/PZ/1835-35-f45.JPG
35mm AIS @f4
http://www.studor13.si/music/photos/PZ/35AIS-f2-f4.JPG
18-55mm II @f4.5
http://www.studor13.si/music/photos/PZ/35Kit-f45.JPG
18-55mm VR @f5
http://www.studor13.si/music/photos/PZ/35VR-f5.JPG
The results at other focal lengths are in favor of the 18-35mm universally, except wide open at 18mm.
As I have said (twice), it is beyond believe that PZ can claim that the resolution in the corners of the 18-55 VR at 35mm wide open can be almost double that of the 18-35 also at 35mm and also wide open!
[/quote]
I would not give too much about lens test. I test all my lenses in reallife and give them back if I dont like what I am seeing. There is not a single formal test procedure which gives fully reliable results for every lens they test or fully mirrors real life performance when shootimg 3D subjects as opposed to 2D test targets. Mostl formal tests nevertheless provide some adequate guideline what to expect from a lens, though not always. That being said, I found photozone to be more reliable than others. (Even though I fundamentally disagree on the canon 17-85. My copy provided after CA correction sharp results from corner to corner at 17mm, which I proved to Klaus in a high res 15mp sample. The lens problem at 17mm is high field curvature, which hardly becomes visible in 3D environments)
Therefore I strongly suggest to make your own real life tests and buy according to your own experience. Furthermore: I dont think that colour or contrast is a meaningful criteria to differntiate lens performance anymore. All modern lenses are multicoated and provide a very high degree of contrast in most situtations. colour differences are minimal and strongly dependet on the light. This is why your photos look great: because they are shot in great light. In any event, if colour diffences of lenses exist, are irrelevant if you process from raw.