Hi Everybody! Just got this month's issue of CI. They have tested both these lenses, so I though a lot of people might be interested in knowing their conclusions. I know tests are always subject to debate, but for what it's worth, here it goes (note: all comparisons are based on previous CI tests):
CANON EF 24-70 f/2.8 L 24 mm: Distortion (barrel: 0.6%) is hardly visible and comparable to that of the 24 mm prime lens (also 0.6%). Vignetting at full aperture (0.6 IL difference between center and corners) is noticeable and also comparable to that of the 24 mm prime (also 0.6IL). It completely disappears at f/5.6. Resolution at full aperture is slightly lower than that of the prime lens. Resolution reaches it peak at f/8 50mm: Distortion (pincushion) is not measurable (28-70 f/2.8L: pincushion, not measurable). Vignetting at full aperture (0.5 IL difference between center and corners) is noticeable (28-70 f/2.8L: 0.6IL). . It completely disappears at f/4. Resolution at full aperture is comparable to that of the 28-70 f/2.8L. At all other apertures, the resolution of the 28-70 f/2.8L is slightly better (no, I made no mistake, thats what the tests say). However I doubt it's noticeable in real situations. 70mm: Distortion (pincushion: 0.4%) is not noticeable (28-70 f/2.8L: pincushion, 0.4%). Vignetting at full aperture (0.4 IL difference between center and corners) is considered slight (28-70 f/2.8L: 0.4IL). . It completely disappears at f/4. Resolution at full aperture and all other apertures is comparable to that of the 28-70 f/2.8L. Sorry, not comparison between the 28mm or 35mm setting, as CI tested 24-35-50-70 focals on the 24-70 f/2.8 and 28-50-70 on the 28-70 f/2.8. However, distortion and vignetting on the 24 mm setting of the 24-70 zoom are the same as on the 28mm setting of the 28-70 zoom. Conclusion: they conclude that the optical quality of the new zoom(24-70) is comparable to that of the old zoom (28-70). SIGMA 120-300 f/2.8 APO EX HSM IF 120mm: Distortion is not measurable. Vignetting at full aperture (0.5 IL difference between center and corners) is considered visible. It disappears at f/5.6. 200mm: Distortion (pincushion: 0.4%) is hardly visible and comparable to that of the Canon Ef 200 mm f/2.8L II prime lens (also 0.4%). Vignetting at full aperture (0.6 IL difference between center and corners) is noticeable and just slightly higher to that of the Canon prime (0.5 IL). It completely disappears at f/5.6. Resolution at full aperture is comparable to that of the Canon prime lens. At all other apertures, resolution is just just marginally lower than that of the Canon prime lens (hardly any difference. I'm quite sure it won't be noticeable in most real situations). 300mm (the focal that we all are interested in!): Distortion (pincushion: 0.4%) is hardly visible and slightly higher than that of the Canon EF 300mm f/2.8 IS (0.3%) and the Sigma 300 f/2.8 EX prime (0.3%). Vignetting at full aperture (0.7 IL difference between center and corners) is noticeable and higher to that of the Canon prime (0.5 IL) and the Sigma prime (0.5 IL). It completely disappears at f/5.6. Resolution at full aperture is lower to that of the Canon prime lens and comparable to that of the Sigma 300 f/2.8 EX prime. At all other apertures, resolution is lower than that of the Canon prime lens and comparable to that of the Sigma 300 f/2.8 EX prime. Conclusion: They conclude that this zoom can replace a classical 300 mm f/2.8 prime Please note that even tough the Canon prime outperforms the Sigma zoom in every aspect, the difference between the two can not be considered "huge". That's it. Of course the whole test is much more complete (it has resolution and color rendition charts, plus other information and it's respective interpretation). It's not my purpose to provide all test details, just to give an idea of the tests to those who can't read French or have no means of buying the magazine. I just wish that Sigma zoom started at 100mm and had OS... * **** ******* *********************************************************** * For list instructions, including unsubscribe, see: * http://www.a1.nl/phomepag/markerink/eos_list.htm ***********************************************************
