Friday, February 15, 2013

Which 35mm lens is the best? Battle of 35mm lenses on Sony NEX 7. Part I Bokeh.

Here it is - a comparison contest between 20 lenses at 35mm focal length, mounted on Sony NEX 7.

Why those 20 lenses? Why many 35mm lenses are missing?

The answers is simple - I have those 20 lenses and do not have others.

If anyone, who will read this review, have one of those many missing 35mm, that can be mounted on NEX, and want to either send it to me for the test, or test it himself and send me the images, we can let this site grow. (I will probably add new Sigma Art 35 f1.4 in Nikon mount when it comes here, and Samyang 35 f1.4  in Nikon mount, when I find it in my home chaos).

I was also writing to SLR magic, asking them to give me the quote and most importantly, the delivery estimation of their 35mm f0.95, but I didn't receive decisive answer.

So far, I also couldn't figure out, if Mitakon 35mm 0.95 is the same lens as upcoming  SLR Magic (considering that they are produced in the same factory and designed by the same people), and in one of my previous posts I received valuable comment regarding - not so great performance of Mitakon. I decided therefore, not to wait longer for those Chinese lenses, and will probably never review them. (Unless producer does not send me testing samples, but I doubt they will).
I am kind of sorry because of that, as I think that SLR Magic did and probably will deliver some interesting lenses.

Why NEX 7?

NEX, because of the short flange distance, that makes most lenses possible to mount, and 7, because of the high pixel dense sensor, that has not much mercy with poor optics.
Of course, NEX 7 has several flaws itself, but, I will not compare magenta cast in this test, so it should not affect image quality in most cases.
If any lens can be disadvantaged by NEX 7, it will probably be the only RF lens in this test - VC 35 f1.2 II AS.

So, which lenses are included...

  1. Canon EF 50mm f/1.2 L USM + Metabones Speed Booster (This combination results on APS-C in 35.5mm f/0.85)
  2. Voigtländer 35mm f/1.2 AS II (Using Metabones - Leica M/NEX adapter)
  3. Minolta AF 35mm f/1.4 (Using LA-EA2 adapter)
  4. Canon EF 50mm f/1.8 II + Metabones Speed Booster (This combination results in 35.5mm and f/1.3)
  5. Olympus OM-Zuiko 50mm f2.0 Auto Macro + Metabones Speed Booster (using no name - Olympus/Canon EOS adapter. This combination results in 35.5mm and f1.4)
  6. Sony SEL 35mm f1.8 OSS
  7. Fujian 35mm f1.7 (Using no name C-mount/NEX adapter)
  8. Zeiss Distagon 35mm T* f/2 ZE (Using Metabones - Smart Adapter version I)
  9. Carl Zeiss Jena Tevidon 35mm f/1.9 (Using no name C-mount/NEX adapter)
  10. Canon FDn 35mm f2 (Using Kipon - Canon FD/NEX adapter)
  11. Canon FD 35 f2 (Concave front element - Radiocative. Using Kipon - Canon FD/NEX adapter)
  12. AF-S Nikkor 24-70mm f/2.8G ED (Using no name (zebra color) Nikon G/NEX adapter)
  13. Contax G Planar 35mm f2 T* (Using Metabones - Contax G/NEX adapter)
  14. Auto Chinon MC 35mm f2.8 (Using Kiwi - Pentax K/NEX adapter)
  15. Leitz Wetzlar Elmarit R 35mm f2.8 (Using Kipon - Leica R/NEX adapter)
  16. Sigma 30mm f2.8 EX DN E (This is only lens that differ in FL. I included it, because it is probably the sharpest native E-mount lens, and incredible value for money those days.)
  17. Leica Vario-Elmar R 35-70 f4 (Set at 35mm. Using Kipon - Leica R/NEX adapter)
  18. Canon FD 24-35mm f3.5 L (Using Kipon - Canon FD/NEX adapter)
  19. Sony SEL 18-55 f3.5-5.6 OSS (Set at 35mm)
  20.  Minolta AF 28-80mm f/4-5.6 (Using LA-EA2 adapter.)
In this first part of the comparison, I arranged a scene that should give you an idea of the OOF rendering differences. Every lens was shot wide open, but because their maximum apertures vary, subject isolations varies too. It is however good to see, what you can expect from each lens and the difference in DOF, related to maximum apertures.

If you want to learn more about bokeh, I strongly recommend you to check this article - On Bokeh from Jakub Trávník.  

For the bokeh test, it doesn't make much sense to compare crops.
Therefore, I made a small video, where I arranged lenses from the "smoothest" subject separation to the "roughest" one. It is of course very subjective and YMMV.

(Video is slightly over 3 minutes)

Bokeh Test

Following files were used for the video.

Shot in RAW, imported to LR 4.3, exported to Photoshop CS 5.1, added titles, downsized to 2600 px width at 72 DPI using Bicubic smoother algorithm, and posted here. In a single case, WB was corrected. (Canon FD 35 f2 concave, which due to the radioactive glass element, yellowed by time, and I was lazy to cure it). Otherwise, no corrections were applied. 
Sturdy tripod, remote control, OSS off, MF, 10 shots of each lens wide open, with re-focusing and selected the sharpest ones etc...

Warning! This is not a sharpness test.
I am preparing sharpness test too, but it will be published in a new post soon.

Voigtlander Nokton Aspherical 35mm f1/.2 II

Olympus was used with a Metabones Speed Booster and no name Olympus/Canon EOS adapter

Sony SEL 35mm f/1.8 OSS E

Canon FDn 35mm f/2

Canon FD 35mm f/2 - rare earth Thorium element

Carl Zeiss Planar T* 35mm f/2 (Contax G)

In this test, there are no winners or losers, and I mean it. "Bokeh" is very subjective aspect of photography, often misinterpreted as a lens quality.
In fact it is more about lens character, influenced by lens design.
Some designs, such as a classic sonnar, does have many aberrations wide open and thus creates smooth soft look, which I personally like. Metabones Smart adapter brings effect called "blooming", that makes resulting image "glow" around the contrast areas.
Finally, I can see from those results (and not only from those) that the cheapest lens on the earth (well certainly close to) - Fujian 35 f1.7 is capable of interesting bokeh and good subject isolation.

So, regarding "bokeh" - it is all about the look, you are after.

Instead of spending thousands of dollars for the fastest lens, you can learn how to use cheapest ones, think about your shot before you lift the camera to your eye, use distance from the subject and subject from the background, use light and structure of the background in your favor and create images that will stand out from masses.

I will mention few surprises though...

The biggest surprise for me was overall look of the Nikkor 24-70 f2.8G AF-S ED. You don't want to use this lens on NEX trust me, but comparing it to so many primes, Nikkor has least aberrations wide open, very sharp focus area and pleasant, smooth OOF rendering.

Another pleasant surprise is Auto Chinon 35mm f2.8. I was rather disappointed with Chinon 50 f1.4 (Auto Chinon 50 f1.4 review), but its wider brother is great performer. And it can be found for peanuts.

Finally, Metabones Speed Booster, works really well as a shallow DOF maker. It might have few bugs - vignetting, purple fringing, infinity problem with some lenses and probably few firmware bugs, but it gives something I had been dreaming off - 35.5mm at f0.85 in terms of DOF (with Canon EF 50 f1.2 L USM on APS-C).
While, it is certainly not a cheap combo, it deliver unique subject isolation, that will be hard to beat at this FL.

I uploaded few original RAW files, all taken wide open with following lenses:
Sony SEL 35mm f/1.8 OSS, Minolta AF 35mm f/1.4, CV 35mm f1.2 AS II and Canons EF 50mm f1.8 and 1.2 L with Metabones Speed Booster. 
You can download those files here RAW files and use them for your own evaluation, but you are not allowed to use this files on public, without my permission.
Thank you for understanding.

Links to tested 35mm lenses: (tests are in progress. data will be updated and new lenses added, so be sure to check from time to time, if interested)

Sony SEL 35mm f/1.8 OSS
Carl Zeiss Planar T* 35mm f/2 (Contax G)
Canon FDn 35mm f/2
Canon FD 35mm f/2 Thorium
A.Schacht Ulm Travegon 35mm f3.5 R
Voigtlander Nokton Aspherical 35mm f/1.2 II


  1. Yeah, as a general rule while looking through your test, the aperture really makes a difference. Sure, there are some that surprise, but the wider aperture plays a significant role. And the Canon 50/1.2 at f/0.85 is just sick. I was thinking of buying a 5DIII just to get the 50/1.2 as my main lens since I like shooting wide open, but you make me think about just adding the lens and speed booster to my 5N. Hmmm. Thanks for the test.

  2. Thanks for the nice test and effort. I am still drooling to see what selection of lenses you
    Any hint where to get Speedbooster in the Europe?

    1. I believe the eBay seller "Lensadapter4U" has some deal with Metabones, but you won't gain anything, as their offer is more expensive. I bought my adapters from eBay seller Lasjashk, the official Metabones seller. You can also write directly to them, and they will send you invoice.
      If that is too expensive, did you consider cheaper - Techart or RJ adapter?
      Kipon is also supposed to join the club shortly.

  3. I bought RJ adapter and tested Carl Zeiss Distagon T* 3,5/18 ZE With my Sony NEX-FS100.
    This didnt worked at all. Please dont buy this adapter.

    Zafar Anjum

  4. I love your site! Thank you so much for doing so many amazing tests and sharing the results. Quick question - did you have any clearance issues with the Speed Booster and the Zuiko Macro 50 F2? Some Zuiko's have lens protrusions that cause problems and the 50 f2 looked questionable.

    1. Thanks. Zuiko 50/2 is on the edge, but it can be mounted as you can see from enclosed images. Just be careful, because it might fit just so,so. After taking it off, I haven't seen any marks on the SB glass, neither on the Zuikos. Therefore, if it does get in contact, it's probably at the mechanical part.

    2. Thanks! Lots of my Zuikos look really close but the 24 2.8 and 50 1.8 are the only ones that don't fit so far. 50 1.2, 1.4 85 f2 and 35 f2 are all good.

  5. Thanks for review, it was excellent and very lenses
    thank you :)