1. Reminder: Please use our affiliate links for holiday shopping!

Featured: 'One More Goofy Shootout: NEX-7 vs A7 At 50mm Equivalent' by Jefenator

Discussion in 'Open Discussion' started by Jefenator, Jan 4, 2014.

  1. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    The object of this test was to see if there was ANY halfway-practical scenario where I could get my NEX-7 to match my A7 at the pixel level at comparable viewing angles. I've done a bunch of A/Bing over the last couple of weeks and so far, it hasn't even been close.

    I got a glimmer of hope the other day while running some 50mm infinity focus tests. It looked like there might almost be some overlap between the old Leica 35mm lens on the NEX-7 and some of the lesser 50mm lenses on the A7.

    So for this showdown, I chose the Leica 35mm lens for the NEX-7 - the same one that did so well in my 35mm equivalent comparison. (My Sigma 30mm is significantly wider - not so good for a direct comparison - and it seems softer under all this new scrutiny. Go figure - I used to think of it as my tack-sharp reference lens.)

    For the A7 I chose an Olympus 50/1.8 which was given to me this summer. It's not my favorite all-round 50mm but it's not especially bad, either. (I'm coming to appreciate it for how small and light it is.)

    The best 35 money could buy from the '70s versus a free lens - that's about as good a fighting chance as my NEX-7 is going to get!

    Ironically, with these pairings, the NEX-7 actually becomes bigger and heavier than the full-frame rig! :biggrin:

    11745950664_ee1db96f42_z.
    NEX-7 / Summicron-R 35mm vs A7 / Olympus 50/1.8 by Jeff Addicott, on Flickr

    Hey, I said "halfway-practical", not "sensible". :p

    Let's start with the infinity focus crops, shot from the optimal f/5.6 all the way down to f/16:

    11745610126_7cf6d46913_o.
    50mm Equivalent Shootout - Infinity/Diffraction by Jeff Addicott, on Flickr

    Well? Close - very close - but no cigar. :( The A7 still has a noticeable edge at the pixel level.

    Note how the full-frame A7 captures degrade less with diffraction. Seems to be about a 2 stop advantage. f/16 is much more usable and at that setting, the entire frame tidies up rather nicely as I'll show later.

    Here's another center crop comparison, with the optimal sharpness at 100%, then the f/16 shots sampled from screen resolution. It's subtle, but there is a visible difference.

    11745571454_99ffce2471_o.
    50mm Equivalent Shootout - Center by Jeff Addicott, on Flickr

    Corners at their optimal f/8 seem sort of close. Can't tell the difference, downscaled:

    11745261024_0321a56c1a_o.
    50mm Equivalent Shootout - Corners by Jeff Addicott, on Flickr

    Focused up closer, the difference in the center is still subtle but unmistakable:

    11744856395_33c822db5c_o.
    50mm Equivalent Shootout - Still Life by Jeff Addicott, on Flickr

    If you compare the full image downscaled, f/8 on APS-C versus f/16 on full-frame, the quality is about the same in the middle but on the full-frame, the detail is preserved all the way out to the corners!

    11745097303_11bed564f7_z.
    NEX-7 / Summicron-R 35mm @ f/8 by Jeff Addicott, on Flickr

    11744860425_c65f5de0a7_z.
    A7 / Olympus OM 50/1.8 @ f/16 by Jeff Addicott, on Flickr

    Weird…

    What would a crop vs full-frame comparison be without a BOKEH test? I like how the viewing angles here are so very similar (finally…)

    11745263794_cd5d4db721_o.
    50mm Equivalent Shootout - BOKEH by Jeff Addicott, on Flickr

    f/2 on APS-C seems to translate to roughly f/3.5 on full-frame for BOKEH/DoF purposes. (But note the 50mm lens renders the focus target more cleanly since it's stopped down a bit.) As expected, the full-frame wide open yields that extra reserve of subject isolation.

    If you're still reading, thanks! :)

    I said right in the title this would be a goofy shootout. For a much more sensible high-performance normal-length option on the NEX-7, I would be curious to see how the Zeiss Touit 32mm lens would fare. It has been called the sharpest lens for APS-C e-mount. I may wind up with that lens eventually if I hang on to my NEX (Lord help me, I'm becoming more "normal" as I age.)

    11745951424_10304a9a25_z.
    NEX-7 / Summicron-R 35mm vs A7 / Olympus 50/1.8 by Jeff Addicott, on Flickr
     
    • Like Like x 13
  2. Tabibito

    Tabibito TalkEmount Regular

    177
    Apr 1, 2013
    If I didn't know better I'd think the NEX-7 is the full-frame one because of the bulk. :)
     
    • Like Like x 1
  3. davect01

    davect01 Super Moderator

    Aug 20, 2011
    Fountain Hills, AZ
    Dave
    Great comparisons
     
    • Like Like x 1
  4. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    Nice comparison and write-up Jeff. It looks as if the A7 shots are sharpened a bit more. Did you experiment with sharpening to get equivalent results for the NEX-7 and A7? Maybe the NEX-7's AA filter is stronger than the A7's, who'll say, that could account for the differences. It would be a bit hard to believe that a Leica Summicron runs out of resolution at f/5.6 in the center, especially not compared to a middle-of-the-road 50mm (OK, you Olympus lovers, prove me wrong! :)). Anyway, as you said the differences are subtle but remarkable, I'd have expected very similar performance between the two cameras under these circumstances.
     
    • Like Like x 1
  5. tommie

    tommie TalkEmount Rookie

    14
    Mar 7, 2013
    Stockholm, Sweden
    Goofy it may be but real world comparisons are always great! Do post more if you can. :)

    I often find that same sort of softness with my NEX-6 and that's one of the reasons i'm considering the A7.
     
    • Like Like x 1
  6. WoodWorks

    WoodWorks Super Moderator

    Dec 12, 2012
    Ashland, OR, USA
    David
    Great test, Jeff. I would not have expected such a noticeable difference. But with the A7, "good enough" just got bumped up a notch for me.
     
    • Like Like x 1
  7. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    Whoa - featured? Really? Cool!! (Where my road trip to the Eastern Sierra failed, some test shots out in the yard finally succeeded. :cool:)

    When I pulled the trigger on getting my A7, I honestly had no idea how things would test out. The NEX-7 always seemed like such a hot rod - I was expecting less of a difference, if any. Indeed, when the entire image is downscaled for screen viewing, the differences in sharpness can be practically invisible. But you never know which nuances might survive the reduction. (It could just be me, but it seems like my recent frames have just a bit more "snap".)

    Ultimately, if you are keen for acuity at the pixel level (and when I got my NEX-7 that was a significant selling point) for a 24MP capture I can now safely say that full-frame is a no-brainer. 24MP feels like a "sweet spot" on full-frame while being, frankly, a bit of a strain optically with APS-C. (There might be a good reason why Canon & Fuji have stayed in the MP teens with their crop sensors.)

    Don't get me wrong - I am still a big fan of APS-C. It's a great size/IQ blend and I think it should be viable - even dominant - for a long time to come. I plan to keep at least one APS-C body, configured to take full advantage of the compactness.

    Format aside, I'm noticing a few other things with the A7, in particular the color response. The A7 seems to consistently do a nicer, more natural blue sky than the NEX-7. I have noticed this with various different lenses and calibration methods. That alone might be worth the price of admission for some (hopefully the upcoming NEX-7 replacement will feature similar refinements in that area). It seems like the A7 has a more life-like palette - one foot out of the "clinical" realm without being exaggerated (like Canon).
     
    • Like Like x 1
  8. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    Both cameras received identical sharpening treatment in post. (LR defaults.) Exposure was reduced slightly in a few instances to compensate for differences in the cameras' metering. Rather than sneak in an AWB comparison like I did the last time, I calibrated each camera & lens with my WhiBal card.
     
    • Like Like x 1
  9. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    That might explain why the NEX-7 comes out worse than might be necessary. For my NEX-6 I use Sharpening 60/1.0/25/50 (Amount/Radius/Detail/Masking) and for the A7 40/1.0/25/50 is enough.
     
  10. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    There's certainly some wiggle room in post for all kinds of parameters. And who knows what's really going on with the optical filters and RAW cooking?
     
  11. Deadbear77

    Deadbear77 TalkEmount Hall of Famer

    Sep 14, 2012
    Northeast Ohio
    Kevin
    Nice post. I was thinking of turning in my nex 6 for a 7.... Now I'm thinking differently.

    Still debating about the a7 or a7r.

    Thinking of going with the a7 just because of the front curtain shutter.
     
    • Like Like x 1
  12. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    No regrets so far, settling on the A7. I wondered if having larger pixels might allow me to "let my gut out" more as it were, and thus far I'd have to say: yes! It most definitely does! :biggrin:
     
  13. tommie

    tommie TalkEmount Rookie

    14
    Mar 7, 2013
    Stockholm, Sweden
    Care to elaborate? Unfortunately i don't understand your reasoning. My logic tells me that if the A7 require less sharpening how is that not the same as the A7 being sharper to begin with?
    So how is it then wrong, in a test scenario, to use the exact same amount for sharpening?

    If we were to apply 0 sharpening it stands to reason that the A7 would still be sharper?

    Edit:
    This test seem to correlate to the article "The Full-Frame Advantage" written by Ken Rockwell in 2007.
    Especially see the chapter "Cheap lens on full-frame vs my best lens on DX".
    Link: http://www.kenrockwell.com/tech/full-frame-advantage.htm

    The results of his test is a tad more clear but it's still the same sort of increase in overall detail.
     
    • Like Like x 1
  14. addieleman

    addieleman Passionate amateur

    Nov 13, 2012
    Netherlands
    Ad Dieleman
    I agree with you that the A7 is sharper to begin with based on these tests, where the processing is kept the same. However, each camera has its own optimum sharpening parameters when processed and if you follow the approach that you want the camera to look as good as it can, you'd probably have to try different sharpening for each camera. Having done that, there may still be a difference but it may be smaller.

    In practice you don't do comparative testing but you try to make your pics look as good as possible so different processing may be applicable for each camera. Of course, a box of Pandora is opened up that way: what is the optimum processing for each camera? What criteria are used to determine optimum processing? Matters of taste come into the equation as well.

    So I don't think Jeff's test is wrong. I was just speculating on the cause of the (for me) unexpected differences.
     
    • Like Like x 1
  15. Jefenator

    Jefenator TalkEmount Top Veteran

    876
    Nov 23, 2012
    Oregon, USA
    Jeff
    If I can convince one other soul to support the A7/r series, then it was all worth it! ;)

    I was able to produce some visible blurring by disabling the front curtain on my NEX-7, but only under some very specific conditions. (Long lens, flimsy tripod w/ center column extended, portrait orientation, certain range of shutter speeds, basically going way out my way to see if I could make it happen at all…)

    My take is, if someone really wants that extra resolution they should definitely just go for the A7r. I don't have the need (or the optics) for that extra resolution and I do like the extra bit of peace & quiet.

     
  16. radimere

    radimere TalkEmount Regular

    71
    Apr 2, 2013
    Excellent test, Jeff. A definite eye-opener. "It's not the sensor; it's the lens" just got roundly debunked. Either that, or the OM is a better lens than the 'Cron. Unlikely but who knows? After all, the OM is a well-regarded lens.

    The diffraction advantage is real, though. That's why MF shooters can go down to ƒ/32. My D700 also produces sharp files at ƒ/22, but the NEX-7 turns out mush at that setting.

    It'd be interesting to see a controlled experiment featuring the same lens(es) on the two bodies :)
     
  17. Gandalf

    Gandalf TalkEmount Regular

    59
    Sep 5, 2013
    Nice writeup, and probably as good an apples-to-oranges comparison as we're likely to see.

    The "infinity focus crops" in particular, give an excellent comparison of diffraction effects. Through all of that series, the NEX-7 resolution at each f/stop is almost indistinguishable from the A-7 at one stop smaller. This is exactly as the math predicts.

    Good job!
     
    • Like Like x 1
  18. lenshoarder

    lenshoarder TalkEmount Veteran

    297
    Feb 7, 2012
    • Like Like x 1
  19. alaios

    alaios TalkEmount All-Pro

    Jan 11, 2013
    Düsseldorf
    Alex
    Thanks for all the work you did. Keep testing
    Alex
     
    • Like Like x 1
  20. alaios

    alaios TalkEmount All-Pro

    Jan 11, 2013
    Düsseldorf
    Alex
    Hi there,
    I waswondering if there is also any comparison between A7 and high end DSLR/APS-C regarding iso comparison.

    Regards
    A.