The Great A900 Lens Shootout

<p>I happen to have 11 lenses that fit the A900, mostly older Minolta lenses I used on film era Dynax 7 and then the 7D while waiting for a good DSLR from Minolta or Sony. Obviously the A900 is demanding on lenses and I wanted to make a comparison, entirely for my own purposes, to see what lenses to keep and use, and which to replace. I am posting here the results, not trying to prove any point, just out of community spirit since this might be a subject that interests other photographers who are either new to the Sony system or look for second hand lenses. This is not a scientific test by any means and is not a demonstration on what these particular lenses are capable of. It has clarified my thinking so at least it has served that purpose. To keep it simple, I made a few simplifications. I shot highest quality jpegs with standard settings in the camera. Raw processing could probably extract some more information and judicious sharpening would make the images better as well. But I doubt the difference or ranking between the individual lenses would be affected. I shoot mostly landscapes and architecture. I don't really need wide open performance, so I shot all images on F/5.6. Some lenses might be better when stopped down another stop or two. Certainly the results would be different wide open. I just did not have the time or interest in going through the whole exercise with different apertures as well. It is all right when comparing two lenses but I had more than ten to go through. I have a total of 5 lenses that cover approximately 28mm focal length on the Sony. Just out of curiosity I wanted to also test a few other cameras at this focal length, namely the Ricoh GRD with its well regarded fixed focal length lens (6mm on the tiny 8Mp sensor) and the Sigma DP1 with it equally highly regarded 16mm F/4 lens. The Sigma is a 4.5Mp camera but is sold by its manufacturer as a 13Mp one and has been touted as equal in quality to Canon 5D with excellent upressing potential. The Sigma DP1 was the only exception where I shot the original as RAW file. This camera is just hopeless with jpegs, and the Sigma raw converter gives the option to export the image as 13Mp file, which I did, as highest quality jpeg. I also made a comparison with the 10Mp Olympus E-3 using their newest wideangle zoom, the well regarded but amateur grade 9-18. I shot all images using the lowest native ISO speed in each camera, 200 for Sony, 100 for Sigma and Olympus and 64 for Ricoh. I find it difficult to compare side by side images that are of different size. The native resolution of the Sony A900 is 24.5Mp. If one wants to make a really big print, one needs to increase the resolution to offer at least about 150 dpi to the printer. There are several programs that do this upressing, Genuine Fractals etc, and some do a better job than others. Again, to keep it simple I just used photoshop and increased the resolution in one go to comparable 24-25mp numbers from the GRD, DP1 and E-3. Better results could no doubt be achieved by using different programs, doing it stepwise, or sharpening in between or at least at the end. But this was not the purpose of my comparison, adding these other cameras was just a fun exercise that did not add a lot of workload to what I was really attempting, comparing the lenses I have for the Sony A900.</p>

<p>Without further ado, here are the comparison images. These are all 100% crops from approximately half way between the center and edge of the image (unless otherwise mentioned). For 28mm, I made a separate set of images at the very edge of the image area. Some comments on the lenses in the first comparison: The Minolta 2.8/24 is the first version of the lens from 1980s, I did not bother to upres or enlarge this image since 24 is quite close to 28, but the image crop is a bit smaller than for the other lenses. The Minolta 4/24-50 is the first version of the lens, from late 1980s, bought second hand to work as a small and light normal zoom for the crop sensor 7D, and from my experience it did well on that 6Mp APS body. The Konica-Minolta 2.8/28-75 and 2.8-4/17-35 are often said to have Tamron heritage and just about equal lenses are available under that brand name. I have the KM versions and cannot comment hopw the Tamron lenses would do. The Schneider 2.8/28 Super Angulon is a shift lens originally built for Leica R but available in different mounts including the Minolta/Sony AF. </p><div></div>

<p>Here are the 28mm crops from the edge of the image area.</p>

<p>My opinions: 28-75 and the 28 Super Angulon seem to be the best. Minolta 2.8/24 is pretty good as well. 24-50 is the weakest, especially on the edge. My first impression was that the 17-35 is not too good either, but on second look it is actually not that bad, though clearly weaker than the top three mentioned above. DP1 is surprisingly good, clearly better than the GRD and even 10Mp E-3 with the 9-18 zoom, but no match to the A900 even with a mediocre lens. My conclusions: I will keep my eyes open for a 17-35G, and failing that either get a 20mm (hoping it will be good) or the new Zeiss wide angle zoom. 24-50 will have to go, or stay with the 7D.</p><div></div>

<p>Between 50-80mm I have again several lenses. The Minolta 1.7/50 is the latest version. Sigma 2.8/70-200 APO zoom is the older version, just before the digitally coated model came out that also focuses a bit closer. The 2.8/80 Hartblei is in my understanding the normal lens for Kiev medium format camera made in Ukraine and mounted on a tilt and shift mechanism. This is not the Hartblei Super Rotator model which I found unnecessarily complex, but very similar to the current Canon and Nikon TS lenses in its overall construction. One problem with this lens is that it does not rotate (from vertical to horizontal for example) on the A900 because of its large prism. The lens needs to be removed and the mount manually rotated before remounting the lens. Not really a big deal since the use of TS lens requires tripod and contemplative approach anyway. But it rotates on the full frame film Dynax 7 for example.</p>

<p>My conclusions: At 50, again the 24-50 is the weakest. 1.7/50 may be a tad better than the 28-75. On longer focal lengths, the Hartblei is clearly #1, followed by the Sigma. The 28-75 is not bad but clearly third in this group. The Hartblei is a really positive surprise. The 2.8/28-75 is also a positive suprise and well worth the good publicity it has received. Still, not the absolute best, as expected. Buyers of the 24-70 Zeiss probably got their money's worth.</p><div></div>

<p>At 100mm I have three lenses that cover this focal length and also made a comparison using the dedicated Sigma 1.4x APO teleconverter with the 2.8/70-200 zoom. The Minolta 2.8/100 macro is the second version of the lens, from 1990s. The Minolta 4.5-5.6/100-300 APO is the latest D version. I bought this lens because it is much smaller an dlighter than the 2.8/70-200 Sigma and expected it to be just about as good since it is an APO Minolta and slower in speed. Having now compared them, in my opinion this lens was a disappointment. The 100 macro is the best, closely followed by the Sigma zoom, which is good even with the 1.4 extender attached. At 200mm the results are similar and the 100-300 remains weakest.</p><div></div>

<p>In the last comparisons I have two images for each lens, the aircon from nearly center and the pipes closer to edge. The last comparison is the Sigma 2.8/70-200 APO zoom with its Sigma 1.4x APO tele converter against the Minolta 100-300 APO telezoom at its longest setting. Out of curiosity, I again made another comparison against the Olympus E-3 with its 70-300 zoom (of Sigma ancestry) and ressed up to similar 24mp size. With no Sony/Minolta lens to compare against, I also include the Minolta 8/500 AF Reflex mirror telephoto and a comparison against Olympus E-3 with its 70-300 lens at equivalent 250mm setting. These I have in two ways, E-3 upressed to 24Mp and Sony downressed to 10Mp.</p>

<p>My conclusions: Now the 100-300 starts to shine and is almost as good as the Sigma. Both give more detail than the 10Mp Olympus with a zoom at 150mm setting. At 500mm, the Reflex Minolta gives more detail than a shorter lens cropped. This is not a bad lens overall. The Olympus 70-300 is also very good at its native resolution, but not really a match to more megapixels when the full image area is needed. </p><div></div>

<p>Thank you, Sir.</p>

<p>The 50mm 1.7 did better than I thought it would. The 100-300 APO did not live up to its APO status. I have both lenses that I have used on a film system. My 50mm 1.7 also does pretty well on film, just short of my Contax 50mm 1.7 and 50mm 1.4s. I was disappointed in my 100-300 APO and get the same type of results as you do. The only digital body I have that fits these lenses is the Minolta 5D which does not have enough definition to show up the weaknesses in any lens I have that fits the KM mount.</p>

<p>It was nice of you to share your results with photonet members.</p>

<p>Tom</p>

<p> </p>

<p>Couple of final comments. All images were shot on tripod with mirror lock up. I usually shot more than one image focusing again to prevent any focus or other errors. For meaningful comparison, you really need to look at the images in their native resolution, not cropped to fit a small screen. At least in my computer there is a small square button hovering above the image that enlarges it. Or you can of course download and save the images using your browser and then look at them at 100%.</p>

<p>Tom, I suspect the 1.7/50 would do less well wide open but it is not at all bad stopped down. And yes, the 100-300 was the biggest disappointment and surprise to me. I should have sold my lens before publishing these results : )</p>

<p>Thanks Ilkka,<br>

I don't come around Photo.net as much as I used to, but for some reason figured I'd read a bit online today - and saw your post. I also enjoyed doing lens tests (back in 7D days) and found similar results regarding the 50mm, 28-75 and 100-300 APO D. The 28-75, although no match for something like the Zeiss 24-70, was very good for its going price. The 100-300 APO D was always somewhat of a let down to me, especially after shooting a good prime such as the 100mm macro.<br>

I'm drooling over an A-900...you are a lucky man at the moment with that camera! :-)<br>

All the best,<br>

Jed</p>

<p>ilkka, there is a lot of detail in these posts. thanks for taking the time to post the results here.</p>

<p>Why do two images with the same camera, but with different lenses, have such different color?</p>

<p>What images are you talking about? Sometimes I had to wait for clouds to come so that the comparison would be more even, and sometimes after waiting a long time the clouds never came and I just shot in somewhat subdued sunlight. That is probably the main difference. It took about 2 hours all in all so there is some difference in colour temp and all were shot in auto white balance so there could be some variation due to that as well. In hindsight I should have done the whole test on a cloudy day.</p>

<p>O.K. Thanks. That explains it. That little Sigma DP1 doesn't look bad.</p>

<p>Wonderful test Ilkka, thanks very much! I am disappointed that my 24-50/4 does not really seem good enough for the A900. And even with film, both my 70-210 lenses are visibly worse. I was really happy after replacing a 50/1.7 with 50/1.4, but it has problems as well according to the dpreview.com test.<br>

Benny, the Sigma DP1 and Oly E3 or lens have a warm (red) color balance. It is not uncommon for lenses to have unique color rendition. Note blueish corners of the 24-50/4. The prime lenses have more saturated colors than the zooms.</p>