New Items

One, Single Samsung NX-1 Test

Published December 9, 2014

I just want to be clear, before we get started, this is not an NX-1 review. It’s not even a complete optical test. It’s simply me answering a single question: is this new Samsung worth further investigation from an optics/image quality standpoint?

Samsung’s 28.2 megapixel, APS-C backlit sensor should capture more light with less noise, with a bit higher resolution than the 24 megapixel sensors in the competition. The hybrid AF system with striped illuminator should be quick and accurate. The short lag time 2.36 mpix OLED, low-lag-time EVF should be great. I know the built-in WiFi will actually be useful, because I’ve used lesser Samsung cameras in the past just for that feature.

All of those things are well and good. But none of them are even worth investigating until one simple question is answered: is this camera system (meaing camera and lenses) competitive with other APS-C cameras I can buy? Because I’ve sung the “Baby, I Love Your Sensor, but you Just Ain’t Got No Optics Blues” to too many cameras already.

Photo credit: Roger Cicala


So when a Samsung NX-1 and the new 16-50 f/2.0 – 2.8 OIS lens showed up in the lab, I told them I’d need it for just an hour to run one test. I was going to directly compare the Samsung camera and lens to the best APS-C combo that I could think of in 10 seconds or less, the Nikon D7100 with 14-24 f/2.8 lens. If the Samsung was competitive, then further investigations were warranted. If it wasn’t, really I wouldn’t be all that interested in it.


Photo Credit: Roger Cicala


We could be pretty sure the Samsung sensor was at least very good, so the camera starts with a slight advantage over the Nikon. But, I expected the excellent Nikon optic would be a bit better than Samsung’s wide angle lens. So my expectations (I like to put those out on the front end) were that the outcome might be even, but I wouldn’t have been surprised if the Nikon smoked the Samsung given that I expected Nikon’s optics were far superior.

The Test

Again, I had no intention of doing a complete test and analysis. We set up our Imatest lab to test both systems at 20mm zoom length (to avoid being at either lens’s extreme), shooting both sytems at ISO 400, aperture f2.8. I should mention that our newer high-resolution targets were used, so you can’t directly compare these results to our results of a year or more ago.

Only one copy of the NX-1 and 16-50 f/2-2.8 OIS was available, so this is not our usual multi-sample test. Multiple focus-bracketed shots were taken with each system.

The results are pretty simple. The table below shows center resolution, average resolution over the entire lens front, and average of the 4 corner resolution (performed about 85% of the way to the absolute corner) for MTF 50. I’ve also included the same Nikon lens tested on a 16.2 megapixel D7000 body, to give you a bit of perspective about how more megapixels in the body affect the Imatest results of the lens-body combination.

  Samsung NX-1 / 16-50 Nikon D7100 / 14-24 Nikon D7000 / 14-24
Corner Avg. 770

Don’t read too much into this. It’s a test of one lens at one focal length and one aperture. It doesn’t mean the Samsung 16-50 f/2.0-2.8 Β is as good as the Nikon 14-24/f 2.8, or the camera is better, or anything else fanboyish.

What does this mean? Very simple, the 16-50 f/2-2.8 OIS lens is at least pretty good, and the camera-lens combination is very good. If it wasn’t, the Samsung system wouldn’t slightly out-resolve the Nikon like it does. Which answers my question — the new Samsung NX-1 system and lenses are certainly worth further investigation.

Roger Cicala and Aaron Closz

December, 2014

Author: Roger Cicala

I’m Roger and I am the founder of Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in New Items
  • Holger

    The 14-24 isn’t a new lens (I think 2007) and designed for FF sensors, with lesser pixel pitch. As far as I understand, mounting lenses designed for FF on smaller sensors could result in suboptimal lp/ph figures, as soon as the sensor outresolves the lens (depends on aperture, as discussed in a famous luminous landscape article)..
    Suppose it is designed to be able to resolve 80lp/mm, you basically can get about 1248 lp/ph on APSC, but 1872 on FF (x1.5). The smaller pixel pitch on the APSC cameras doesn’t help much here, which is why, due to the larger size of the sensor, FF lenses don’t need to be as sharp as lenses on smaller formats (the Zeiss lens manual talks about that, too). Tamron claims to construct the new 15-30/2.8 for 50MP FF cameras, so it should be sharper than the Nikon. Wouldn’t it be more fair to use a newer lens for comparison, or is my reasoning not correct?

  • R

    If you’re reluctant to use the Nikon 17-55, you could compare the NX1 to the Sony a77 (II?) with the 16-50 f/2.8, that’s a fairly similar lens with a rather good reputation.

  • Dave Sucsy

    Thanks for this, and all your great tests and info that you share with us.
    Perfect test, perfect question, perfect premise. I’m glad that you continue to partially ignore (with your great sense of humor) all of us Monday-morning quarterbacks!
    Is the new Samsung SYSTEM even going to be worth considering/further investigation? (How does it stack up against two worthy elements of what is arguably the comparable de-facto best (no fanboy here) SYSTEM?)
    Apparently, the answer is YES – it is worth further investigation.
    Extremely valid question and mini-test!

    But I still don’t understand why you don’t devote 48/14 of your life to fulfill all of our requests! πŸ˜‰

  • Roger Cicala

    Gordon, I should have said “set up to test both lenses set to 20mm zoom length”.

  • >>We set up our Imatest lab to test both systems at 20mm focus distance

    Twenty millimeters is less than an inch. Did you mean 2 meters (roughly 6 feet)?

  • adrian

    While it’s always nice to see hard numbers, and i hope Roger can find the time to include that body in the tests, i think your question is rather simple… the image resolution is only limited by the sensor, so the numbers will sit in between (probably precisely at half) the numbers from the 16MPx D7000 and the 24MPx D7100.

  • Martin

    1) Interesting, overall the NX1 sure look like Samsung means business, and it seem they have got a fair idea how to build lenses too. Not surprising, but good to know.
    2) And of course I have to point out what I assume is a typo: “test both systems at 20mm focus distance” … I guess that was supposed to be “20 mm focal length”? πŸ™‚

  • Marco

    I own this lens and I think it’s very good even on my NX300, which is “only” 20MP and far from being the best sensor around. Could you please make a similar test with also the previous generation of NX cameras (NX300/NX30) to see how much more resolution the NX1 can produce (with the same lens)?


  • intrnst

    Roger, maybe instead of “system (meaing camera and lenses” you should use “system (meaning camera and lenses”.

    Thanks for the pit stop.

  • John Pittman

    The Nikon 14-24 is probably the best wide lens Nikon or anyone else makes so if the NX1 and the 16-50 outperform it on the D7100 what makes you think lesser lenses like the 17-55 or Sigma 17-50 will perform better? The 14-24 is a full frame lens on an APS-C camera so the results should be even better than the lens used on a full frame since it’s eliminating the corners. Frankly, I’m shocked the 16-50 performed so well against the 14-24. It’s a fine performance by the Samsung and I’m excited to see additional tests.

  • Roger Cicala

    Andrei, I felt the 17-55 f/2.8, being a much older design, wouldn’t have held up in the corners. I wanted to see if the Samsung could run with the best. The Sigma 18-35 would be another great option and if it ever gets out in Samsung mount a great direct comparison of the two sensors. But mostly I wanted to see if the Samsung lens, which they are very proud of, was really good.

  • Roger Cicala


    That’s exactly my point – it’s worth further testing. I wouldn’t have been shocked if the Samsung lens was weak enough to make it not worth further testing because I picked the 14-24 precisely because I thought it the sharpest wide-angle lens available. Like I said, this was NOT a fanboy contest. It was simply to see if maybe this new Samsung can run with the big dogs, and it appears that maybe it can.


  • Alliumnsk

    It is completely fair to stop down lenses to same aperture. If we need f-stop faster than f/2.8, the Samsung wins because the Nikon has no such aperture at all.

  • Toshik

    First of all, Roger thanks for a quick the test.

    But i think comparison is a bit unfair:

    1). As mentioned in the comment above, 24Mpx sensor in Nikon resolves slightly less than 28Mpx in Samsung.

    2). Why 14-24 is choosen for a comparison? Yes, it’s a very good lens optically. But 14-24mm doesn’t really coincide with 16-50 focal length.

    I think Sigma 17-50mm OS better suits here.

    Also Samsung lens was stopped down a bit (2.0->2.8), Nikon was wide open.

    So as is pointed in the last paragraph, it’s difficult to judge.

    Samsung sensor maybe resolves slightly better, their lens seems to be not bad too. But this needs to be thoroughly tested futher.

  • Tim Marshall

    Really dont see why the author chose that lens, it invalidates the test completely. One is a full frame wide angle lens being used on an APSC Nikon. The other is a APSC standard zoom lens being used on an APSC camera.

    How about the Nikon 17-55 F2.8 at 20mm vs the
    Samsung 16-50 F2.8 at 20mm. Better yet try them at 17mm, 25mm, 40mm and 50mm

  • AndreiK

    I know how Roger “loves” suggestions for further tests, but will propose adding few more competitors too.
    First of all Nikon’s comparable lens is 17-55/2.8 so i’d better see comparisons with it. Second Sony (a6000 generation) and Fuji have some very good sensors, so would love to see comparisons with them and best compatible lenses. Also Sigma 18-35/1.8 also deserves to participate, either on top of Sigma’s own DSLR with a Foveon sensor or coupled with d7100 / 7dII.
    Also if the point is to check best possible resolution the system is capable to, then why not to shoot at base ISO at f/5.6?
    Thanks again for your great tests and insights, even if you don’t work 24/7 fullfilling our requests πŸ™‚

  • CarVac

    That’s a pretty good lens: if the sensor is limiting both, then sqrt(28/24)*1300 gives 1404 line pairs/picture height. It’s very close.

Follow on Feedly