p.3 #1 · Sony A7 - first impressions and some tests
rscheffler wrote:
The kit zoom's results remind me of what I got with it from my infinity test shots. Just click on my Flickr link/icon and dig through the image, it's towards the end of the city skyline series. Will be interesting to see Phillip's infinity shots from it, and how they compare.
IMO, it's not a great lens for technical applications, but should be fine for most day to day stuff...
45/2 does look great!
Ron, I completely forgot that your test included kit lens, thanks for reminding! indeed, it'd be interesting to see how lens compare at midrange/infinity.
p.3 #2 · Sony A7 - first impressions and some tests
carstenw wrote:
I find it very suspicious that virtually nothing here looks any good before f/4. What is going on? Very weird.
Is it really that surprising? The last high resolution sensor that the corners of many of these lenses saw was Velvia 50... Obviously the Sony sensor is a particular challenge but I doubt these lenses would fare much better on a Canon FF sensor. Only side by side comparisons would reveal the truth but the A7/R is the only game in town for short registration FF.
PS edit. Thankyou Phillip for doing this. I cant wait to see the images you create with this amazing little camera.
p.3 #4 · Sony A7 - first impressions and some tests
carstenw wrote:
I find it very suspicious that virtually nothing here looks any good before f/4. What is going on? Very weird.
I suspect there are a lot of possibilities, foremost could be the one both you and I have mentioned and Phillip is aware of - testing distance. Many of these lenses are not going to perform at their best 1m away. The distant testing should reveal more.
p.3 #5 · Sony A7 - first impressions and some tests
It might be interesting if there was a separate single thread that contained a huge matrix of all the lens compatibility with links to results from other threads or sites. It's getting hard to keep track of all the different tests out there, and sometimes it can get confusing which lens works for which camera and why. Unfortunately, I don't care enough about non-native FE lenses or non-modern lenses to create this thread and maintain it. If someone had a burning desire to track all this, it might help those who have a burning desire to use old lenses.
Suggested format:
Works on A7? .... Notes .... Link to test ------ Works on A7r? .... Notes .... Link to test
lens 1 @f/
lens 1 @f/
lens 2 @f/
lens 2 @f/
It would be great for someone to just look at this matrix and see which lens to start investigating.
p.3 #6 · Sony A7 - first impressions and some tests
I don't see anything confusing here, not for close-up performance, except that the 35-70/3.5 macro looks to be an even bigger bargain than we thought. Phillip's Rokkor 58mm looks the same on his a7 as mine does on the a99 at that distance. And if you've ever used Canon's EF 35/2 (pre-USM/IS), it's basically the same optics as the nFD 35/2. Good for the money, but not the perfect all-arounder. I think some people are going to have to reset their expectations on the performance of older lenses in general.
Thanks for doing this, Phillip. I'm even more anxiously awaiting the arrival of the 7r now.
p.3 #7 · Sony A7 - first impressions and some tests
thanks phillip!
carstenw wrote:
I find it very suspicious that virtually nothing here looks any good before f/4. What is going on? Very weird.
not sure what your surprised by? all the minolta glass and the contax g 45/2 look just like they do on film and other FF digital cameras. which ones were you expecting to do much better?
p.3 #9 · Sony A7 - first impressions and some tests
Hmmm, those test results seem to be not right at all especially the corner performance. Maybe the camera lens mount is bent or maybe the lens mount is misaligned. At f5.6 or f8, it should be sharp - just basing this from my Nikon results i.e. the 50mm f/1.8 corners are dang sharp at f5.6.
p.3 #11 · Sony A7 - first impressions and some tests
It's (the sensor) a mild update of what FLM and I use in the a99, believe me there is nothing wrong with good older lenses on it..testing distance is the 'usual suspect' here...Zeiss used to publish different MTFs for near and infinity - with very different results, even for a 'macro' lens. Much poorer closer.
p.3 #12 · Sony A7 - first impressions and some tests
sebboh wrote:
thanks phillip!
not sure what your surprised by? all the minolta glass and the contax g 45/2 look just like they do on film and other FF digital cameras. which ones were you expecting to do much better?
Agreed. I think many have forgotten how relatively new the idea of corner to corner sharpness is, especially with lenses wider than 50mm.
p.3 #13 · Sony A7 - first impressions and some tests
I would like to add to the excellent spot on remarks above the sensor toppings factor. Sony has traditionally used quite thick material so far. The A900 was a disaster in the corners even with highly regarded lenses. It had smearing even with the 135 ZA. I have the feeling they are still using the same glass thickness. Anyhow an infinity test should be very revealing.
p.3 #14 · Sony A7 - first impressions and some tests
Thank you so much for creating all these samples and presenting them so clearly. A lot of work, I'm sure.
One question -- what version of the Canon FD 35mm f2 are you testing? The "radioactive concave front element" version, or one of the later conventional versions?
p.3 #15 · Sony A7 - first impressions and some tests
I hope that there may be some factor influencing negatively performance of all tested lenses.But,if these results will be confirmed later on,bye bye A7/R. Thank you Philip for what you've done.
p.3 #16 · Sony A7 - first impressions and some tests
douglasf13 wrote:
Agreed. I think many have forgotten how relatively new the idea of corner to corner sharpness is, especially with lenses wider than 50mm.
indeed, even many modern zeiss lenses look bad in this kind of test that shows a 100% crop from the extreme corner e.g. the ZE, 28/2, 50/1.4, 85/1.4 actually look noticeably worse than some of these in the corners (even the ZE 25/2 and 50/2 MP don't look so hot in the extreme corner).
i'm quite impressed with the g45, rokkor 35-70, and rokkor 50/1.4.
with regard to the slr wide angles, with very few exceptions most of them kinda sucked towards the edges till the 90s.
p.3 #17 · Sony A7 - first impressions and some tests
Kurt Munger looked at the ZA135 on a900 in a review:
'In the last row I show the difference between the corner shot at F/5.6, and the center shot at F/5.6. I see virtually no difference as compared to the centers, super good full frame performance.'
Of course this is hard to test as all *lenses* fall away at the extreme corners, the best (flattest) bench MTF I have seen is the 100MP, but even it loses ~10% of centre MTF, for which you cannot fairly blame the sensor. My a900 and a99 have 'smeared' their way to some very impressive corner performances. I might be less fussy
p.3 #18 · Sony A7 - first impressions and some tests
sebboh wrote:
thanks phillip!
not sure what your surprised by? all the minolta glass and the contax g 45/2 look just like they do on film and other FF digital cameras. which ones were you expecting to do much better?
I was expecting them to clear up faster in general, and especially the performance of the kit zoom is highly disappointing. The results of most of these tests are so bad that I wouldn't bother shooting with these lenses at all. What bothers me the most is that not a single of these lenses is great, or even uniformly good. It makes me wonder about the sensor toppings and the strength of the A7R as an alt platform. More results are needed with stronger lenses, but for now I have lowered my expectations. I do hope the A7 is better.
p.3 #19 · Sony A7 - first impressions and some tests
carstenw wrote:
More results are needed with stronger lenses, but for now I have lowered my expectations. I do hope the A7 is better.
This is the A7.Witout the R.
p.3 #20 · Sony A7 - first impressions and some tests
Phillip Reeve wrote:
...my chart which is about 40comx60cm. I am pretty sure that this is partially to blame for the bad results, the lenses are pretty close to the target. Most of them performed much better when I tested them at a longer distance on the a7r some weeks ago.
I suggest you edit your OP to highlight this. Lenses do often perform differently at typical testing distances of ~3m vs. infinity, and that's bad enough. But in this case, particularly with the wides, you are even closer. I am afraid I think your test results are simply invalid.
What you could do is tape 5 copies of your test chart on a wall at the centre and each corner, and shoot that. In case your wall is not flat, refocus for each test patch - you will then not pick up "softness" due to curvature of field, but then again there's a school of thought that you don't want to.
Thank you Philip for what you've done.