philip_pj wrote:
I can't access the main a7/a7r thread - no, not the weepy hand-wringing one regarding wide angle M lenses and their travails on this most modern of all cameras - but the main one concerning the cameras themselves...
it happened again. replace the "topic" in the url with a "z" to be able to see posts beyond the first on that last page. i thought fred had tracked down the problem...
uhoh7 wrote:
"My point here is not to say which is better. There are many other factors to take into account (number of blades, build quality, ease of use …) But here are 3 of the best 35mm lenses the world has ever know and the Sony/Zeiss 35mm F/2.8 Sonnar is easily as good on the MTF charts."
Edgars Kalnins wrote:
Are sony zeiss MTF charts based on real lens test now? If not, the article seems pointless, doesn't it!?
No, Sony Zeiss MTF's are calculated, which should be obvious to anyone just by looking at the MTF (because central/0 values at 100% are impossible!). A curve that starts at 100% at F8 is not of this world. Diffraction alone would prevent such a real world measurement. I believe someone suggested dropping everything down by 20% at least to get a rough idea of performance (I guess that's if manufacturing tolerances and so forth are ideal).
I rather like the passionate tone of the dear susan person, and MTF is a valuable tool, as a 'guide', so we can talk about Leica's mid field dips, for example. What inevitably happens in the absence of consistent methodologies, is that people will use whatever is around. This happens with all kinds of data. Sony are far from the only offenders in using 'promotional' MTF data.
The thing the market really needs is something comprehensive - with an optical bench for infinity testing and and flexibility for near distance testing and outcome reporting, based on a good sized sample of all lenses. The Imatest is another which has some controversy attached to it but is also still very useful.
Aberration analysis (beyond distortion and vignetting, the two least troublesome aberrations) would also be very helpful, as would actual images that show lens problems. Roger Cicala is on the case, here is a recent comment made on his site and a link to the article above it:
'The latter lets us take out paper and pencil and make a map of field curvature. More powerful benches will actually plot out the field curvature map for you.'
cputeq wrote:
Man oh man those images look great to me! I love the very subtle and neutral rendering. I actually downloaded the original of the 2nd shot, added just very minor tweaks (minor sharpness and minor vignette and a spot removal) and added it as my desktop on my 15" retina Macbook -- it looks amazing (I even kept the watermark, as I love the choice of font).
Hope you don't mind! Very nice.
I typically have a rule that the only desktop wallpapers on my computers are images shot by me (hey, it's a pride thing!), but I made an exception to the carrots...I just love the light and feel of the images. Okay enough gushing. ...Show more →
cputeq, thanks for your nice comments!
I will try to take more test images with a different lens and upload in the next few days, which shouldn't be difficult as I can take this camera with me anywhere I go. I actually thought I left the camera at home today until I had another look into my bag. It is so light (compared to full-frame DSLRs) and it still gives the best full-frame image quality available today. Impressive. Handling/usability is nice and well-thought as well, at least for my use.
lostinjapan wrote:
Here is a quick test shot from this evening. I used the A7r with the Canon 17mm TSE at full shift. I also bought the camera here in Japan, though I bought the adapter from BH, even with shipping and tax it was about 100 bucks cheaper.
Ryan
Hi Ryan,
Have you taken any shots with the 17mm TSE and examined corner detail when shifted? Any comparisons with a FF Canon?
Edgars Kalnins wrote:
Are sony zeiss MTF charts based on real lens test now? If not, the article seems pointless, doesn't it!?
Did you read the article?
The point is simply: is the 35/2.8 too expensive?
No, it's a great lens.[says the author] Do you disagree?
I discounted the lens a bit at first, RF snob that I am. Now that I've seen a thousand samples, I think it's a fantastic lens. Well worth 800usd.
But I'm happy to be proved naive, if you have some evidence.....
Now the 35/2 asph is looking killer as is the lowly cv 35/1.4, and the hulky cv 35/1.2 and I've heard the latest 35 lux is good but haven't seen that yet.
The picture that I posted on page 2 of this thread was taken with the 17tse at full shift. The corners are sharp, though regardless I always stop down to f/11 when using that lens at full shift regardless of which body I am using it on. I also tested in daylight as well and the results looked perfect.
At least on the A7r it works very well and the vignetting is much better than I had hoped.
Below is one more image I took with the 17mm tse last night. Nothing special, still a work in progress, mainly I took it out to see how it performs.
As quite a few people have tried Canon L lenses already on A7r (also all of my L lenses perform well on A7r as expected), today I went ahead with the more compact consumer grade lens, EF 85mm 1.8 on A7r.
Take handheld at f2, ISO 400. RAW conversion with Lightroom 5.3 RC. No sharpness added, no saturation/contrast added, only a tiny bit of white balance fix. It seems like EF 85/1.8 produces less bokeh fringing on A7r than on Canon cameras. Of course I have to perform side-by-side comparisons to verify this but so far, I'm very satisfied with its performance on A7r. I used Metabones Smart Adapter Mark III. AF works but it's very slow.
The original size is available at: http://www.flickr.com/photos/ippei-janine/10944698735/ http://farm4.staticflickr.com/3698/10944698735_a1dc450699_h.jpg
The picture that I posted on page 2 of this thread was taken with the 17tse at full shift. The corners are sharp, though regardless I always stop down to f/11 when using that lens at full shift regardless of which body I am using it on. I also tested in daylight as well and the results looked perfect.
At least on the A7r it works very well and the vignetting is much better than I had hoped.
Below is one more image I took with the 17mm tse last night. Nothing special, still a work in progress, mainly I took it out to see how it performs.
I've got my old SLR stuff in the wings for backup, esp wide angle.
nFD 20
Kiron 28 and PC nikkor 28
some nice 35s too, but not looking like I'll need them.
Lots of cropped images. Planar 1,4/50 and Distagon 2,8/28 within the Flickr pool, perhaps others. I've written him to see if he'd post some uncropped images from the D28.
Nice shot... I have been waiting for someone to get this combination. I am very interested in it myself - hope you don't mind if I ask some questions:
How is the live view (vs the Canon 5d mk 2)? The screen is OK in daylight?
Have you tried the 24 ts-e also?
Does the Metabones adapter feel like it would stand up to a lot of lens changes every day?
Thanks a lot in advance for your help...
Hope you will post some more soon.
Simon
Hi Simon,
I think the live view is better the 5d2 but not as good as the 5d3. The 5d3 has much better live view, but the tiltable screen is quite nice. I tried the 24 ts-e as well, and like the 17mm it was sharp into the corners on my quick test. The raws look very nice, much more DR than either of the Canons, with nice clean shadows. (which is my one big gripe with the 5Ds) There is no question the Canons have better build quality to them.
The adapter is my only concern at this point. It seems very well constructed and the finish is very very good, but I do worry about the lens release button and how it will hold up in the long term. It works flawlessly with all my lenses. I only have ZE primes and the TSEs so I can't test the autofocus. But it works exactly the way I wanted it.
The weather here has been horrible for the few days, but starting tomorrow night it should start getting better. So I am planning on taking the camera out with all the lenses this weekend and spending some time getting to know it better.
I believe someone suggested dropping everything down by 20% at least to get a rough idea of performance (I guess that's if manufacturing tolerances and so forth are ideal).
