Prime and zoom comparison

Discussion in 'Sony Alpha E-Mount Lenses' started by addieleman, May 31, 2016.

  1. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    Yesterday I bought a used FE 2.8/35mm, primarily intended for use in the smallest possible kit as a snapshot lens when I want to carry the minimum weight and size and still have a serious camera with me. But is it also to be preferred as a landscape lens over the FE 4/24-70mm that I have used up to now? The prime should be better than the zoom, right? Let's find out...

    All images shown at 100 % viewing, in the true spirit of the pixel-peeper that is me :).

    i-P5xx7xG.
    Surprisingly, not much of a difference, the zoom may be just a little sharper despite being wide-open. Center crop at f/8 is essentially the same.

    i-zJmxQVQ.
    Now there's a surprise: the zoom is much better near the extreme corner. Left upper corner shows the same kind of difference.

    i-jWfc4GV.
    And at f/8 the same story: the zoom is clearly better.

    Want to see more? Check out all full-resolution images.

    So the 2.8/35 I bought is a dud? Don't really think so. When going up close it shows some decentering, resulting in a worse right edge, but it's not visible further out in these test shots. The 2.8/35 I borrowed last Sunday at the Sony booth showed more or less the same behaviour: sharp at the focus point, but the focus plane certainly isn't flat. A third copy I tried a year ago didn't show excellent corner sharpness either in a flat subject, the reason I didn't buy it then. Lloyd Chambers says the same in his evaluation of this lens, which is much more thorough than mine of course.

    This is in fact an evaluation of the camera (A7R II, firmware V3.20) in combination with the lenses. I used autofocus throughout and I confirmed the well-documented focussing behaviour of both lenses: the 2.8/35 focusses wide-open irrespective of the aperture set, the 4/24-70 focusses with the shooting aperture as long as it's f/11 or larger; when stopping down below f/11 it opens up a bit (to f/11?) for focussing. This might explain the somewhat better sharpness of the 4/24-70 at f/4, the 2.8/35 could have been at a disadvantage due to focus shift when focussing at f/2.8 and shooting at f/4. I was pleasantly surprised by the sharpness wide-open of the 2.8/35 and I don't really see the sharpness at smaller apertures others rave about. I hope that at some point in time there will be an option to tune this focussing behaviour to your own preferences via additional menu items. If focus shift adversely affects the 2.8/35's performance, manual focus should improve things; I'm going to try that later.

    Also it is my experience that it's important to switch off image stabilization for optimal image quality of the 4/24-70 (even more so for the FE 4/16-35, especially at 35mm). With Steadyshot on you run the risk of ending up with pictures unsharp at one edge. Here's hoping that there'll be an option of disabling OSS while keeping the in-body stabilization. Of course the 2.8/35 is not affected, not having OSS.

    So there you have it. That 4/24-70 zoom once again proves it can be an excellent performer and with these high-resolution cameras the challenge of bringing out the best possible image quality becomes ever greater. On the other hand, a little less than optimum performance becomes less and less interesting for many applications. I'm sure that I'll be happy to print big even if pixel-peeping shows some defects like shown here.
     
    Last edited: May 31, 2016
    • Informative Informative x 4
  2. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    So I tried AF-S vs. MF at the shooting aperture (Live View Display set to ON).

    i-bxTfLVH.

    The difference is on the verge of visibility, I'm not sure the MF picture is better. In my experience the A7R II's autofocus does a better job than me manually focussing at the largest possible magnification combined with peaking, so I conclude there's not much to be gained here by MF. None of this matters where I intend to use the FE 2.8/35 for, I was just curious to see if focus shift was relevant for this lens and it doesn't look like it. That is confirmed by focussing at max. magnification, I hardly have to adjust focussing when stopping down if at all; might just be me trying to find optimum focus. If I do the same with e.g. the FE 4/24-70mm there's a noticeable readjustment necessary, like for many legacy lenses.

    So the FE 2.8/35 is just fine for street-shooting, snapshots and candids and the FE 4/24-70 will continue to be used for landscape/cityscape photography. Case closed.
     
    • Informative Informative x 6
    • Like Like x 1
  3. roundball

    roundball TalkEmount Hall of Famer

    Oct 8, 2013
    USA
    Good project.
    As others have, I've always heard that primes were better than zooms, but I've personally never seen anything to suggest that.
    Not being a pixel peeper may account for that, but I have become pretty critical of my own work and use manual Canon zooms a lot.
     
    • Like Like x 2
  4. WoodWorks

    WoodWorks Super Moderator

    Dec 12, 2012
    Ashland, OR, USA
    David
    I've got to say that your results are surprising to me, Ad. I think I'll go do a similar test with my 35 vs. the 24-240. Is it OK with you if I post my results here?
     
    • Like Like x 1
  5. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    By all means, am curious!
     
    • Agree Agree x 1
  6. WestOkid

    WestOkid TalkEmount All-Pro

    Jan 25, 2014
    New Jersey, USA
    Gary
    David even though the 16-35 is known to be weakest at 35, I (and I think others since many of us have the 1635Z) would be interested in that comparison. That said, I don't like to give others work so don't go out of your way.
     
    Last edited: May 31, 2016
  7. WoodWorks

    WoodWorks Super Moderator

    Dec 12, 2012
    Ashland, OR, USA
    David
    OK, here you go. The scene. Shot on a tripod, IBIS off, 2 sec. timer.

    Scene.

    100% crops, from left to right: SEL24240, SEL35F28Z, SEL1635Z.

    Upper left corner @ f/4
    4UL.
    Center @f/4
    4C.
    Upper left corner @f/8
    8UL.
    Center @f/8
    8C.

    I'd say the 35 prime is the winner in this contest. But at f/8 the difference isn't all that significant.
     

    Attached Files:

    • Informative Informative x 7
  8. WestOkid

    WestOkid TalkEmount All-Pro

    Jan 25, 2014
    New Jersey, USA
    Gary
    Thanks David.

    Overall the prime wins, certainly at F4. However, it looks to me that the 1635Z wins at F8. I have also thought my 1635 did well @ F8, but I wasn't expecting that to show in this test since everyone always complains about it being terrible at 35mm. Either way, what this confirms for me is that the perspective one has on a lens may be more about individual shooting style. People who shoot wide open will have one view of a lens because it just doesn't work for the situation they need it to work. People who shoot mid-aperture with the same lens will have a totally different view. People who shoot portraits where corners don't matter vs people who shoot landscapes where corners do. That's why brick walls were never my thing.
     
    • Agree Agree x 1
  9. WoodWorks

    WoodWorks Super Moderator

    Dec 12, 2012
    Ashland, OR, USA
    David
    Mine neither. And I agree 100% about shooting style determining one's estimation of a lens. I doubt that I'll ever shoot the 24-240 at f/4, so how it performs at that aperture is irrelevant to me. As long as it can run with the primes at ~f/8 it'll find a place in my bag. And now that I've seen how well the 16-35 performs against my 35, that Loxia 21 lust I have is likely to go unrequited. :rolleyes:
     
    • Like Like x 1
  10. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    Thanks from me too. Interesting! I agree with Gary that the 4/16-35mm looks better in the corner at f/8 than the 2.8/35 and that there's no contest at f/4 where the prime is clearly better. That roughly corresponds with my experience with my copies (sample variation, always something to keep in mind) where the 16-35 is good but the 24-70 is slightly better at 35mm and f/8. I get the idea that my 2.8/35 copy isn't as good as David's but that this lens is not the best option anyway for landscape photography where I often want good sharpness across the frame at apertures f/8...11.
     
    • Agree Agree x 1
  11. WoodWorks

    WoodWorks Super Moderator

    Dec 12, 2012
    Ashland, OR, USA
    David
    While I was out there with the 35, I shot a frame at f/2.8 as well, but didn't include it because the other two lenses don't reach f/2.8. Here's the result, left to right SEL24240, SEL35F28Z, SEL1635Z, all shot wide open, with 35 at f/2.8 and the other two lenses at f/4.

    Upper Left Corner
    428UL.

    Center
    428C.

    With all three lenses wide open, I'd say the prime outperforms the other two by a lot in the corners, and noticeably in the center.
     
    • Informative Informative x 4
    • Agree Agree x 1
  12. pbizarro

    pbizarro TalkEmount Veteran

    358
    Nov 24, 2014
    Portugal
  13. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    Being an engineer I have a reasonable understanding of the meaning of an MTF graph, but I'm never able to translate that into what it means for me using a lens. Moreover, Sony supplies calculated MTF graphs showing the ideal behaviour and, to put it mildly, that doesn't always represent the performance of your own copy. Full-res image samples on the 'net of the kind of scenes I intend to use the lens for, are much more useful to me. And then there's sample variation so you can end up with a worse than average copy, as seems to be the case with my 2.8/35.
     
  14. pbizarro

    pbizarro TalkEmount Veteran

    358
    Nov 24, 2014
    Portugal
    In this specific case, it means that as one moves away from the centre of the image, the separation between the lines for sagittal and meridional measurements indicates that the performance (for contrast and resolution) deteriorates. A good primer here:

    How to Read MTF Charts

    As an example, just compare the MTF for the Sony Zeiss 35 f2.8 with the one for the Zeiss Batis 25:

    https://www.zeiss.com/content/dam/P...downloadcenter/datasheets_batis/batis_225.pdf

    You can see that the Batis 25 holds onto contrast and resolution much better towards the corners, even wide open at f2. BTW, MTFs from Zeiss are actually measured, not simulated (afterall, they invented the thing).

    How this translates in real life: towards the corners, the contrast and resolution of the Sony Zeiss 35 f2.8 lens are not as good as in the centre, at all apertures. Hence, this is not a lens suited for subjects where corner performance is critical. Overall, it is a great little lens, suitable for travel, people, events, etc. Whereas the Batis 25, with a much more constant performance across the frame, will be also suitable for subjects where that kind of performance is required.
     
    • Informative Informative x 1
  15. Chris Munden

    Chris Munden TalkEmount Regular

    42
    May 13, 2016
    Peter Bower
    Its easy to get swept up in Zeiss mania, yes, they (some zooms possibly made by Tamron) make some very good lenses but they are generally not consistent with their zooms, primes are a different ball game and are generally excellent. Having lived through the 35mm SLR days of the 60/70/80/90, Zeiss primes (zooms were rare then and pretty awful ) were consistently good but were often out performed optically by Canon FD and Minolta Rokkor lenses. My old manual focus Canon FD 50mm F/1.8 was a tad better than my 50mm F/1.8 Zeiss Planar on a Contax. Its worth bearing in mind that in those days the 'Tessar' labelled Zeiss lenses were at the budget end of the range where as Sonnar, Planar and Distagon were more high end in performance. As for German lenses in those days Leitz were probably the best followed closely or equalled on a few occasions by Zeiss, but even then, Leitz had to turn to Minolta for help.
     
    • Informative Informative x 1
  16. chalkdust

    chalkdust TalkEmount Veteran

    285
    Sep 25, 2015
    Bert Cheney
    In one thread we have managed to incorporate
    prime vs zoom
    critical image examination (pixel peeping) vs MTF graph
    Zeiss vs Canon, Tamron, Leitz,...
    portrait vs landscape (not referring to rectangular orientation)

    Thanks to all the contributors for thoughtful, useful information without ranting.
     
    • Appreciate Appreciate x 1