Crazy Comparison! NEX-7, E-P3 and X100 high ISO examples and crops.

Sony NEX-7, Olympus E-P3 and Fuji X100 High ISO Comparisons & more thoughts on the 7

Had about 40+ e-mails asking me for some high ISO last night so I figured I would throw something together quickly. I have had the NEX-7 for less than 24 hours and have taken about 40 shots with it outside of my house so far. Thats it. I shot it last night in the dark with the Zeiss 24 1.8 and wanted to see what kind of results I could get. I will post some of those later but for now I wanted to post some HIGH ISO comparison shots from these three cameras that I took this morning in my kitchen. These shots will be posted below but 1st I wanted to touch on some things with this camera.

Some more initial thoughts on the NEX-7

I noticed many have been complaining about the Zeiss 24 being so large. Well, it is but it really isn’t. A 24mm 1.8 lens of high quality has to be sort of large. Look at the Leica 24 Summilux which is MUCH larger than this Zeiss and has no AF built inside. Sure the Leica is 1.4 but this Zeiss is 1.8, still pretty fast. Also note that the NEX-7 and Zeiss 24 1.8 is about the same weight or a tad lighter than the Olympus E-P3 and Panaosnic 20 1.7. The lens is NOT heavy.

Things I have noticed while shooting the NEX-7 are that the E-P3 focuses faster but the X100 is slower than the 7. The EVF in the NEX is the best I have ever seen and the configurability of the NEX is amazing. If customizing is your thing, the 7 can do just about anything in that area. It can get a little intimidating and a couple times I wondered if Sony did not cram TOO much into the 7, but after some use I realized that like with any camera, once you set it, you can forget it. The 7 has all of the external controls you could ever need or ask for. It’s also very speedy to shoot. Aim, focus fire and it’s ready to go again. The X100 feels painfully slow in comparison. 

Also, I have been getting requests to test loads of Leica lenses on this camera. All I can say to that is I will test what I have. I do not have wide angles or 90’s, can’t afford it! I will test what I can and will also shoot with the kit zoom as well. Another thing is about the size of the body itself. It is small but larger than the 5n (thankfully). It feels wonderful in the hand and the grip is just right. The camera is solid but light and it is not even close to DSLR size, even with the lens. The lenses for the NEX system can never be as small as the lenses for the Micro 4/3 system as the sensor is much larger than the Micro 4/3 sensor but Sony is aware of the desire for smaller lenses and are working on it. It is not as cut and dry as many think. As for a full frame sensor in a NEX, Sony said if they did that then they would have to make a whole new lens mount and it would not be the “E” mount. But they did say “anything is possible”. 🙂


Below are 100% crops from the NEX-7, X100 and E-P3 (cameras I have on hand right now)  6400. Noise reduction was set to low on the 7 and X100 and OFF on the Olympus. You can not turn it off on the 7, only set it to WEAK. ALL CAMERAS WERE SET TO F/2!

(had to remove the full size files due to site crashing…will post full samples in the full review which will be after my server upgrade).

It is interesting how each cameras metering exposed the scene. I set the lens for f/2 on each camera and let the camera choose the shutter speed. The NEX chose 1/2000, the X100 chose 1/640 and the E-P3 chose 1/1000! Very strange. Seeing that exposure can have an effect on the noise level we see I then did one more test and set each camera to f/2, ISO 6400 and 1/1000s as I found the Olympus to be the most accurate in regards to exposure. Why the Sony chose 1/2000 I have no idea as it seems a bit underexposed. The Fuji seemed to slightly go brighter at 640. Hmmmm.

So to be 100% fair, let’s see the crops at 6400, f/2, 1000s…

So the same result really. The E-P3 does the worst and that is just due to it’s smaller sensor. It can not keep up with the bigger APS-C cameras in the high ISO noise dept. Plain and simple. The NEX-7 does well but the X100 seems to have the finest and most film like grain of the bunch. Overall though, from what I saw shooting real images last night, the NEX-7 will have no problem shooting at 3200 or 6400 if needed. More to come…

[ad#Adsense Blog Sq Embed Image]


  1. It’s probably just me, but I like the look of the iso noise of the ep3.Not for every photo but it can be used creatively. I would rather have the noise than the loss of detal with noise reduction. That being said, the x100 does a fine job at all iso ranges its hard not to love the out of sensor.

  2. Someone already posted but it bears repeating – the X100 is not meant to be used for macro shots and does not produce optimum results wide open at f2 at macro distances – this was stated in the manual as well. This test is inherently unfair to the X100 but of course, Steve already said this is purely an ISO test so we shouldn’t be commenting on sharpness but ISO quality instead.

    If this test were to be performed at a focus distance of 1.5m and beyond, the X100 at f2 will really shine. Let’s remember almost no Leica lenses focus below 0.7m, so this is not unique to the x100.

    As we can all see, the m43 camera is the worst performer where noise is concerned; the X100 the best. But the NEX 7 might well be much better with a good RAW processor and downsized to 12mp…

  3. It’s amazing to see that the Ep3 produces the sharpest results. Does this mean that the ep3 has a thinner AA filter? Or just different way of Jpeg processing?

  4. I wonder why the E-P3 noise reduction was set to OFF? To me this looks a little bit like a deliberate attempt to make the smaller sensor look much worse?

    No, I’m not an Oly fanboy, and would have loved to see the NEX 5n in the comparison as well. But that would showed that Sony didn’t put the best sensor into their best-equipped camera and wouldn’t make Sony look good, as well. I think Sony wouldn’t have been happy with that, right?

    Thanks anyway for the comparison – I’m quite surprised how well the E-P3 holds up against it’s competitors – also considering the very pleasing colors and obviously most reliable exposure.

  5. hm.. interesting responses…. though, considering E-P3 sensor is so dated but still be able to keep up with these newer sensors, I think E-P3 wins! 🙂

    • x100 is ultimate retro precision digital camera.
      Taking picture with unique style. A fashion statement! In-front and behind the lens

      NEX-7 is a new fantastic all around gadget camera.
      EP3 is a just another digital camera that is doing well
      M9 is the ultimate dream machine. as long as you will never wake up.

    • I would call the X100 an almost perfect digital camera, if it wasn’t so slow to operate. It has it’s place in my camera bag and its IQ is phenomenal but it can’t keep up with the photographer behind the lens – something the 7 looks like it might be able to do.

      I am witholding judgement until I have one in hand but based on Steve’s comments it looks to be as responsive as one could hope for.

      I’m less concerned about how a camera looks as I am about how it handles and performs but as you say – personal preference.

  6. NEX-7 has a same problem just like the previous versions did, feels/look like a toy! X100 feels like a proper camera when you use it. I’ll be keeping my X100 for a long time… right until X200 comes 🙂


    • Looks can be deceiving…and based on all the hands on accounts I’ve read, the 7 feels very solid in hand – a real photographer’s tool. Looking forward to comparing it to my X100. I know the 5N certainly doesn’t feel like a toy…

  7. An IQ comparison between the Ricoh GXR w/50 mm module & a leica 50 mm lens, and the NEX 7 with its equivalent 50 mm lens & a leica 50 mm…..tested under diff conditions:

    * Macro
    * Low Light
    * Portrait
    * Daylight

    PLEASE 🙂

  8. Give me a break, I can’t take any APS-C sensor seriously. FF/FX is the only way to go for low light. The D3S showed the way!

    • Well, I think you should take APS-C or for that matter any sensor seriously 🙂

      Nowadays good results can be achived with many sensors that are smaller than FF. But I agree that D3/D700 and not least D3S (which is the low light champion) are better than all the smaller sensors.

      But usually very high ISO is only needed in very special conditions at indoor sports events, theater, concerts and similar.

      I own a D700, but is perfectly happy with the high ISO capability of my X100 which is not quite as clean as the D700, but it is still extremely good for everyday use.

  9. If you want to compare the best of small APS-C cameras to the best of Micro 4/3 in high ISO performance, you shouldn’t be comparing to the E-P3. Many reviews have shown that the Panasonic G-3 and the GH-2 have the best high ISO performance of Micro 4/3. That would have been a good comparison. The sensor on the E-P3 is outdated and for all practical purposes the same as in the original E-P1. It is now three Panasonic sensor generations old. GH-1 even does better than the E-P3 sensor.

  10. I don’t think it is the jpeg engine of the nex that is blurring the image.
    To me it looks like diagonal motion blur on the nex7 image from camera shake,
    or it could be effects of electronic first curtain, if you are certain there is no camera shake.
    If you can turn off electronic first curtain and re-shoot it would be interesting to see the results, to say the least.

  11. @ Mike Costa.

    I have handled both NEX-7 and EP-3. The weight difference is irrelevant. The handling of NEX-7 is much better than EP-3. EPM and EPL-3 are even worse than EP-3.

    Pocketability. Only an S95 is really pocketable. Neither NEX-7 nor EP-3 is really pocketable. It is true that EP-3 with pana 20mm f1.7 is a bit smaller than NEX-7 with Zeiss 24mm f1.8, but it doesnt make a big difference. Both cameras must hang from you neck or be in a bag. But the strain on the neck and the size of the bag will be the same.

    3 dials is not too many. On the contrary. It is a dream come true. Clicking to have to switch function of one common dial is nowhere nearly as intuitive and fast. Notice how entry level DSLRs all have only 1 dial. Enthusiast DSLRs have 2 dials. And Pro DSLRs have 3 dials. There must be a reason dont you think.

    Overall, the only thing lacking for NEX is a range of good but reasonably small and reasonably affordable prime lenses.

    • Agree with you, well partially…

      Weight: EP-3 and GF1 are far to be pocketable (very far from S95, although the mini-pen +14mm is very close), but are far smaller than a light SLR.
      Sony A35 weights 473gr (with battery & media), add a 18-55 lens and you get 683gr, as close as the Nex-7 is to a GF1+20mm or an X100. Ok, 100gr doesn’t make a great difference, but stating that Nex-7 + Zeiss 24 is lighter than EP-3 + 20mm is unreal.

      Usability: i have to touch the camera with my hands to really test its usability. For my taste (from the cameras i own/ed) i prefer a real good dial to switching from aperture/shutter plus a top one for ISO (that i touch less frequently). Anyway your good impressions on the 3-dials navi are encouraging.

      IQ: really hope it will be better than A77! I see some JPEG from that camera, and it does not bright compared to other APS-C (not to mention FF here). The noise reduction applied even at low iso is too aggressive for my taste. Anyway you all can check it by comparing with other cameras on the DPreview widget. I really do not understand why Sony haven’t put its 16mp sensor into the Nex-7, maybe something about marketing…

      Lenses: we all agree here… Having a 24mm prime at 999 bucks (ok, it’s a Zeiss, but also the Lumix-Leica 25mm or Olympus 12mm costs less and have similar construction quality, maybe the Oly is even better) and 6.5 cm length as the only valid AF alternative is disappointing…

      Anyway, this camera has sold very well, so we only have to wait about some owners reviews (Steve included) and some good in-hands test to express a final judgment.


  12. Mmmhhh… what i see here is a great camera, but far from pocketable, at least with that protruding Zeiss glass attached to it!

    I was already skeptic on paper, at 6.5cm the lens is too long for a compact camera, but after seeing it on the video i must say it resembles more like an A55 than a M43 (or an X100)…

    The weight topic: claims 350gr for the body including battery and card, add 225gr of the Zeiss jewel and you get 575gr. states 369gr + 100gr of the 20mm F.1.7 you get 469gr, so the advantage is clearly on the M43 side here…

    Anyway, i’ll wait till i can get hands on it, but i must admit i have some reserves for now:
    1. pocketability (more about general equilibrium than weight)
    2. Zeiss lens price (ok, they never be cheap, but at 999$ is too close to others FF Sony-Zeiss lenses)
    3. IQ: hoping it’ll be better than A77, by using the DPreview widget is clear how the Nex 5n is better, even at low ISO…
    4. usabilty: aren’t 3 dials too many? How about only one good dial that switch functions on-click (not the crappy one i have on my GF1, but the same concept), and a top dial for ISO (why the hell the Canon G12 has it, while other “PRO” camera doesn’t?)

    At this time i’ll wait for the next Panny GFX1 (GH2 18MP IQ is identical than the best APS-C sensors, anyway until 3200 ISO) and M43 lenses are great (also read about a 16mm F1.4 by Panasonic) and/or for a (hoping big!) Canon entry in the mirrorless market.

    Mike Costa

  13. I seriously believe that Steve Huff will sell his M9 when Sony will offer a NEX-7N without an AA filter.
    Oh, and there’s really no need for the NEX-7N to have 24MP. 21MP, or even 18MP, should be enough.

  14. I agree. I want to like the NEX…I really do…but it is a clunker with a not-so-sharp lens. How can Sony mess this up. I own a Zeiss 21mm ZE for my Canon 5D Mark II…It is one of the sharpest lenses on the planet. This “Zeiss” lens is NOT in that league…not even close. An look at the size of it???
    It is totally amazing to me that the meek little Panasonic 20mm pancake blows both of these lenses away in the category of sharpness (that is what a lens is supposed to do, right?). Give me the MFT camera over either one of these…with a wide selection (and growing very quickly) of autofocus lenses for every pocketbook. I rarely shoot @ 1600 just tone shoot with lower ISO…and realize that eventually Panasonic may produce a new sensor that performs better in the near future….
    And…this is the Zeiss lens on the Sony…what about the others….???

  15. I think I’ll skip the NEX-7 and wait for next year. Here are some tips for Sony on NEX7N:

    1. Get rid of the AA filter entirely and do it using a DSP (like Leica does).
    2. Allow NR to be turned off completely.
    3. Get rid of the built-in EVF and make the LCD better. Many people might disagree with this, but I still believe that framing using the LCD display is a better choice and allows more freedom when capturing a shot.
    4. Add capacitive touch-screen capablity, like on the NEX-5N. BUT make sure it’s as responsive as possible.
    5. Add the 3-way hinge, just like the one on the A77. Shouldn’t be a problem since that EVF won’t be in the way.

    Oh, almost forgot: Perform a physical miracle and offer a 18-135mm f/1.8 lens for $299.

  16. Steve, thanks for the enjoyable reading. I follow your site from faraway Japan.
    I guess there is an interest to compare NEX-7+1855zoom with the NEX-5N+Zeiss24mm1.8, as both combinations seem to cost around the same amount when adding up the separate components, disregarding special package prices.
    Of course, if someone wants manual adjustment, then NEX-7 is the baby.
    But just for picture quality, disregarding anything else, what would be the best bang for the buck?

  17. I have to compliment Steve on doing the comparison at exactly the same exposure conditions, rather than just accepting the automatic metering chosen by the camera. It is exactly this that so many other “comparisons” fail to do.

  18. Steve, your site is all about Leica M equipment. I think you would be doing the Leica M community a great service if you can do comparison on these cameras: EP3, X100, 5N, 7, and the GXR M in the areas of resolution, high ISO performance, easy of focus, handling of ultra wide angle aka 21MM, in addition to the usual 35, 50 (except for the X100 of course).

    • Thanks Matthew but contrary to popular belief, i do not have a stash of Leica glass here. I have NO wide angles and for some reason its getting tougher to get review samples from Leica these days. Also, I do not always have a stash of cameras. I have to send cameras back after a certain amount of time and never have all of these at the same time. I do not have a 5N right now and the GXR has to go back in the next few days. But the issue with all of this is the lenses. I just do not have the Leica glass to do all of these tests.

      • Steve, thanks for the reply. I just wish that I live close to you and can lend you some of M glasses for your testing. Anyway, any Leica glasses on the available bodies would still be great.

  19. Thanks for making effort to post something so quickly. I will be following your endevours closely.
    A lot of web talk about Sony adding 1/3rd manual ISO control and improving bracketing ability. when you have the time could you comment on these given that you have final firmware in your camera.
    P.s. you must be really thrilled to be able to get a decent time with the camera so soon.
    Glen Wells – UK

  20. Thanks for posting this Steve.
    What are the frames per second in RAW mode vs. jpeg ?

    I had the Fuji x100,loved the IQ but really prefer fa$t interchangeable lenses.

  21. The X100 lens is not really designed to be shot close range at f2. Hence the glowy look, it just doesn’t perform well in those conditions (this was mentioned on the website way before the camera’s release). A shot from at least two meters would look better.

  22. In terms of size, please compare the 1.8/24 with the Summicron-M 2/35 which covers a larger sensor.
    I remember that at introduction of four thirds many were also disappointed by the relative bulk of the lenses. Even the Summilux 1.4/25 for micro four thirds is as large as the M 1.4/50. Lumix 2.5/14 and 1.7/40 demonstrate that compact lenses with decent image quality are doable.

    After so many years of APS-C DSLRs, only Nikon and Sigma bothered to offer a fast standard prime, three times the price of the ubiquitous 1.8/50, but still no fast wide-angle prime. Perhaps in case of Sony e-mount we should be glad to get something at all.

  23. Ummm …… X100 and EP3 look really classic. That huge zeiss lens makes NEX7 a scientific tool rather than a camera.

  24. Are the NEX 7 skin tones as yellow or orange as the NEX5N? I am looking for great skin tones OOC jepgs. I want my wife looking pink healthy not old yellow sick.

  25. I crazy or is the EP3 twice as sharp as the other two cameras even with all the extra noise?
    It is isn’t it?

    • Well I think the winner as far as noise is concerned is the Fudji, the EP3 is sharper but if you want to get rid of the noise then you will smudge the sharpness… So what is a guy to do? Sharper then smudged, sharper lower ISO but risk of movement or best all around but not great and not be able to use any other lense with it….There is no perfect solution, when I was shooting film, I would not change the film based on light but when I was done with it… So High ISO in bright light…done, pushed low ISO films, done, pull Hi iso films, done, une low iso in dim light …sure…

  26. The X100 performed exceptionally even though it’s “only” 12mp. This comparison shows that megapixels aren’t everything, and that 12mp will be more than enough for 99% of the photogs out there.

    I’m still not convinced with the NEX series of cameras without an optical vf.

  27. Do a review between nex5n vs nex7 the ISO comparison

    I believe most of us wants to know is NEX-5n really better than NEX7 in terms of low noise performance

    • Check out Imaging Resource. They have their studio comparison up and you can do a side by side with the 5N and 7. I don’t see a difference at 3200 which is great news for the 7 sensor but keep in mind they are JPEGS. Would still love to see RAW files and edit them myself.

  28. The NEX 7 images look smeared a bit at the high ISOs….disappointing to my eyes. The X100 is definitely a low light champ here, and probably still wins the IQ battle of this gang, albeit with the substantial limitation of a fixed lens and slow AF

    I really wish that the NEX left out the AA Filter or weakened it, and used algorithms that didn’t cause the smearing that I see. The noisier EP 3 images look sharper to me, and I prefer sharp to smeared….

    That being said, none of this probably matters for web sized prints…or prints on paper, for that matter…I still wonder if adapted Leica lenses lose their microcontrast look on this sensor…..

    I may be watchin’ the NEX 7 battle from the sidelines after all…

    • I assume they will. And this is why I ordered a GXR with M module last night. And you should take some credit for it as your excellent review of this combo eventually made me pull the trigger :).

    • I wouldn’t be too concerned at this unless you are a jpeg shooter – Sony is all about NR in jpeg… I’m more interested in why all the crops are the same size when the NEX-7 has twice the pixels…

      Most accounts also indicate it will have a very light AA filter.

    • Ashwin,

      I can’t compare the 7N to the X100 – since I don’t have one – but the high ISO comparison shots at Imaging Resource between the 5N and the 7 are very, very close AND I do have an x100 so that is an easy comparison to make. The 5N looks equal to my eye when shooting RAW – maybe even a hair better (so the 7 should do just as well) – but if you shoot JPEG only, then the X100 clearly wins. SONY is heavy handed with their JPEG noise reduction at higher ISOs.

      You don’t strike me as someone who shoots JPEG at high ISO…

      Also read an interesting article over on TOP that discusses the real impact of AA filters. I certainly don’t think the AA filter has anything to do with the “smearing” at high ISO – that is all SONY software.

  29. The E-P3 clearly has more noise at ISO 6400 than the Nex-7 (or X100) but in my opinion the E-P3 is rendering more detail than even the NEX-7. Look at the edge definition of the printed letters on those labels … they look sharper while the NEX-7 looks smudged.

    I think that high ISO smudging is what an earlier commenter was talking about when he said the NEX-7 image had motion blur. It almost looks like motion blur but it’s just detail smudging due to noise reduction. I would be interested in seeing if the NEX-7 renders more detail with noise reduction disabled.

    In any case, this is not as impressive as I expected from a camera with 24 megapixels and a $1,000 lens. I still have pre-orders in place for the NEX-7 kit, 24mm and 50mm lenses, but if I keep seeing samples like this I’ll probably cancel my pre-orders.

  30. I don’t know anything about noise reduction, post processing, raw or high ISOs but the EP-3 pictures look better than the others to me. Thanks for doing the test.

  31. Since you spoke so highly of the Nex 5n, I would like to see some comparison images of the 5n and 7.


  32. The Nex 7 setup is MUCH more expensive than the X100.. so the X100 does really well here (the x100 costs almost the same as just the zeiss lens on its own)

    I’m still also convinced that the NEX 5n is the better camera image quality wise and therefore I think it’s the better choice overall. Yes the handling is better on the 7, but since you can get the viewfinder for the 5n, is it really worth giving up IQ for better handling? The 5n is also smaller and less expensive.

    Also, I’m not sure the size comparison of the 24mm lenses you mention makes sense. The zeiss is an APS-C lens @ 1.8. The Leica is a full frame 24mm @ 1.4. Stated length of the zeiss is 65.5mm, vs 58.5 mm for the Leica, both without hoods. I’m not sure what the length of the zeiss is with its hood, but I’d guess they’d be similar or the same length. Diameter wise they seem about the same as well. I really do think the zeiss lens should be either smaller or faster for what it is.

    • Well, focal length is focal length, so there shouldn’t really be any reason that an aps-c 24mm should be much shorter than a 25mm full frame lens, if they’re both retrofocal. Maybe the lens could be skinnier, but the AF mechanism and auto aperture cause issues there. Also, the Leica has nearly an extra 10mm of registration distance in the camera, so it is actually longer from sensor to the tip of the lens.

      IQ wise, unless you output the images from the 5N and 7 to the same size, it isn’t worth comparing, and the 7 certainly competes with the 5N at high ISO when doing this, yet the 7 has the resolution advantage when needed. It’s a net gain for the 7.

      • An APS-C lens should be smaller since the image circle necessary to cover is smaller. Something like a Sigma 30mm 1.4 is much smaller than a 28mm 1.4 or 35mm 1.4, for instance.

        I think it is worth comparing. To me I’d rather have the ISO performance than extra resolution. Sony clearly thinks the 7 is a stop behind, since it limits auto iso to 1600. This will affect the IQ at all isos, not just higher ones. A cleaner lower resolution image sounds more useful than a smeary or noisy high resolution one. I think the 24megapixels is just a waste of space, especially on an APS-C sized sensor.

  33. Steve,

    I think for the sake of sanity, you just need to replace the first three shots with the three “corrected” exposures you did with exact same shutter speeds. People are making ridiculous pronouncements based on unequal samples.

  34. I am impressed …. by the X100. I expected more from the NEX-7 – in spite of the high pixel count, the sensor is supposed to be more modern. But let’s wait for the RAW stuff ..

  35. Steve. In all your tests please compare Nex 5n v/s Nex 7 with Sony E lenses 18-55,50,18-200. especially at Low Light. . That’s what everyone in the world is looking for. We know you do wonderful extensive tests and very good comparison.And you also tell the result clearly in laymen terms what is good and what is not. Most of us visit your site to read the reviews. when you reviewed Nex 5n Your site was #1 on top of Google search result page for Nex 5n. Keep up the good work. There is world of interest in Nex Series cameras,we know you are big fan of Nex series too. So in all tests please include Nex 5n. People will greatly appreciate.

  36. Steve, Can you please also do some “P” including Auto ISO mode comparison in exact lighting situation? indoors and outdoors! This may be a very interesting comparison.

  37. The thing with the x100 though is that it INCLUDES A BADASS LENS! It’s worth the price of admission just for the outstanding 35mm equiv f/2 beauty!!

  38. I wish people would read the whole text before posting negative comments and moaning when at the end of the day – they themselves got it wrong! Just too quick to step in when they think they see something wrong…

  39. “Similar lenses” You said in your test you set all the cameras at F2 and let the camera pick the shutter speed – or did I read it wrong?

    • You just didn’t keep reading…the crops at the bottom were all same shutter speed as well. Same aperture, same shutter speed, same equiv focal length on the X100 and NEX and similar on the E. But that doesn’t matter for an ISO test.

      • Hi Steve,
        being a pure amateur it’s not up to me to criticise your approach, but I’d really like to see how the PEN fares with a similar level of NR as obviously used by the NEX and the X100.

  40. P.S., these tests are meaningless if you don’t use similar lenses, manual exposure, and exactly the same shutter speed, aperture & ISO.

    • its called “crazy comparison” for a reason and its fun to read and gives some rough comparison ideas for those interested in getting one of these cameras.
      – dont get too worked out over it…Steve put this up quite fast and we get to enjoy it.

    • The test would be meaningless to most users if you used the same lens on each camera (if that were even possible).

      if you stuck a 50mm Summilux on each of these cameras what would that prove ? you need to remember that most lenses are designed to work with a specific system or sensor, so using the same lens on three very different systems would only confirm which of these the Summilux works best on.

      a great number of users of each of these cameras will use the lenses provided in these test shots than a 3rd party lens designed for a different sensor/system

  41. For me, the Fuji X100 image at ISO6400 is more pleasing to they eye – although that could just be the difference in white balance and noise reduction.

    There is no question the images are sharper with the NEX-7, much sharper.

    The EP-3 in my opinion is completely outclassed by both the NEX-7 and X100 in these sample images.

    • I can’t quite follow you. If you rev up the NR of the PEN to an extent that the letters are as unsharp as they are in the crops of the NEX and the X100, the noise arguably also would be much less visible. Which doesn’t mean that it’s high-ISO performance is on a par with the one of the other two cams, but as the PEN crop os clearly sharper and shows more details, this comparison doesn’t do the PEN justice, IMHO.

    • I agree. Unless it is camera shake (which doesn’t seem possible at those shutter speeds)…There is no comparison in the lenses. The Panasonic 20mm wins hands down.

      • What contest are you talking about Sanford?

        These soup can test shots are essentially comparing JPEG engines at high ISO, not the native sharpness of the lens. how could you compare lens sharpness when the software is mushing up the details? Shoot the same images in RAW and then you would have a closer comparison.

        And even if you were able to tell which lens is sharper – which you can’t from this test- there are other components of lens/sensor IQ that are equally important. Look at the “tree” shot Steve did yesterday and you notice things like OOF rendering, and dynamic range where the Oly sensor and 40mm just can’t hang.

        • Good call Chad…using Raw would be better…who knows what each individual jpeg engine is doing to the image…not a good way to compare. (who shoots jpegs anyway…LOL!)

          • Bob,

            The funny thing is, I’ve taken to shooting JPEG with the 5N (at least below 800)! Those files have SO much latitude and with the portrait setting at low contrast you have plenty of headroom to push the file around. I’ve been a RAW snob for years so this is a real revelation that so far is specific to the 5N – can’t do it with my X100.

            But yes, over 800, RAW is the only way to go – especially with SONY!

  42. Hi Steve,

    Would you do a quick review with the 35mm summilux and the nex-7 ? I am very curious if the nex-7 will be a good camera for leica lens.

    Thanks in advance.

  43. I’d like to see iso comparison to the nex 5n if you have it. That would tell us if how the 16mp sensor compares to a 24mp sensor and if more mp does indeed mean more noise.

    One other thing is that I heard the nex7 can only go up to ISO 1600 in video mode (according to eoshd). Is it possible to use iso 6400 in video mode for the nex 7 or is 1600 the tops?

    • All I can tell you is that the 7 I have was set to ISO 6400 when shooting video. Wether or not it takes it down to 1600 I do not know yet. I will ask Sony for clarification. The video did look grainy for 1600 though…so I am not so sure.

      • just put the next in manual mode and put a high blue in ISO (6400).
        Then press rec and see the on screen display (press the wright button to see ALL the details).
        You ll probably see ISO 1600 … its the same with sony nex 5n (tops at 3200 ).

      • Read somewhere (can’t remember where – but had all the test readings n stuff) that Sony sensors – i.e. as used by both Fuji and Nex range – naturally ‘top out’ at 1000 ISO and that any ‘higher’ ISO value is achieved by an electronic boost to this S/N ratio rather than further fixed logarithmic exposure increases.

        Put simply, the sensors apparently have a maximum upper physical S/N ratio of 1000 ISO and any extra sensitivity is pure software manipulation of that base reading – like 1000 ISO x 2 etc. (All ISO settings are actually created by electronically boosting from a ‘base’ sensitivity point; so this is no revelation, merely a technical limitation)

  44. All three “looking good enough”. However, good old X100 is impressive. Still the best IQ of all compact large sensor cameras.

  45. I am happy with what I see, the NEX7 pre-order stays. 6400 is mighty high in film I never shot over 3200 and I hated the results. On the M8.2 the ISO was intentionally blocked at 800 because the pics were odd above that.

    Cheers! Thanks for sharing and it is hard to believe how poeple get into technicalities…try this vs. that gheez! The perfect camera is not out there, yet, could be a 5n in Nex 7 body but then, in bright light the high resolution gives you so darn much reframing options…

  46. I think we should be able to fix most noise problem or exposure problem with Lightroom or Aperture…
    But doesn’t NEX7 and EP3 looked “FLAT” in this high ISO test? only X100 has that 3D look to it. This problem can not be corrected with post photo editing!!!

    • That extra ‘3D’ effect on the Fuji is just because it overexposed, I think. Warm colours ‘pop out’ at you, cool colours seem to recede. The warm reds and yellows of the subjects are brightest in the X100 shot, therefore they jump out most in that shot. Of course, any 3D effect is an optical illusion, and I think the extra 3D of the Fuji is just due to the colours, which could be replicated by tweaking the exposures of the others in post.

      • @ Chris,
        Sorry I can not agree with you on this…
        I assume all pictures were taken in the exact light environment
        if you must look very closely for this 3D look, look beyond color presentation.
        ex. look between Campbell’s soup and X100 black box. Look for shadow presentation (Dynamic Range) for all cameras shots. NEX-7 and EP3 only showed black shadow, but X100 sees the gradual change of its shadow.
        Dynamic Range is only one thing that makes any picture 3D, yes color also, so is sharpness, and noise level….etc. It is “overall picture presentation” to the eyes of viewers.

        If slight over exposure can capture more details and gather more attention… why not?

        ps. I do believed, NEX-7 is a better, faster and more user friendly gadget camera.

    • I guess that as usual, Zeiss designs its lenses differently than the others. Their wide angle lenses are almost always become bigger than e.g. a Nikon or Leica equivalent. Different elements and more of them can be found if we look at their diagrams.

      Maybe these are needed to
      – make a lens sharper (more asph., ED etc. elements)
      – reduce vignetting, coma and all kinds of reflections and geometric distortion
      – put (silent) AF motors somewhere under the lens barrel, as well. (Maybe floating rear elements happen to be here).

      How about that? Steve mentioned something about the lens design.

      • The Zeiss glass I’ve had on my D700 (back to the latest Nikkors now) was quite big, but optical quality was exceptional, some chromatic aberration, more than I liked, apart.

        • Zeiss lenses are not generally longer they are just super wide. Like the 35mm Nikon f1.4 and ZM 35mm 1.4 the Zeiss glass is much stubbier looking vs the nikon

      • ZM lenses are not significantly bigger than leica lenses, and contax G lens line up was pretty small, and autofocus (ok, in-body af and poor close up focus, but still). Contax g was a great system camera, and seems less bulky with the 35mm F2 than the nex 7 with zeiss 24 (and was full frame !)

        • But i sometimes wonder if it’s not a design/marketing decision, sony digicam have a long history of that kind of form factor, so maybe they want to keep a strong design signature to make a significantly difference with other manufacturer ?

  47. It’s weird, but I’m almost as excited that you have this camera as I would be if it were in my hands. I guess it’s because you test the same way I would, but your critical eye for composition will show off what the highest potential of the camera is.

      • Mine required a hard reset every time I changed the aperture in aperture priority mode :/ I was so sad to return it. I want it love it and cherish it but it was just too slow and faulty compared to my Nikon that just worked and worked and worked and did not skip a beat.

  48. Thanks for sharing! Could the NEX have been exposing to protect the highlights on the olive oil bottle? I think I like the EP3 exposure the best.

    NEX 7 at high ISO looks good enough for me!


  49. Nex 7 looks the best in the first test, but unfortunately there is motion blur in the second test shown in the nex crop.

      • There is definitely motion blur in the NEX-7 shot – look closely at the X at the bottom of the crop in “LH-X100” and you will see it in the stroke of the X going from lower left to upper right. I think the camera moved slightly from upper left to lower right during the exposure.

  50. One explanation: The reason for the cameras choosing different speeds is possibly that the ISOs are not accurate. DxOMark has shown this repeatedly in their analysis. When I compared my Pany G2 with my Oly 620, I found that if I choose same Fstop and same shutter speed, that to get similar exposures I needed ISO 400 on the 620 and ISO 200 on the Panasonic (or, was it the other way around). That is because Pany was rounding the actual ISO up and Oly was rounding down. In reality both cams were shooting at an ISO of 300.

    So the camera of yours that shot at the lowest shutter speed may actually be at a lower real ISO than what the setting is….

  51. The NEX-7’s pictures look terribly mushy, which keeps me off of it. But I believe it has a 12MP RAW mode. Could you do a comparison between 12 and 24 MP RAW files in your final review, please?

    • I personally think that’s due to Sony’s JPEG engine. Their JPEG’s have always been a bit splotchy for my taste at high ISO. Once Lights gains RAW support for the NEX-7 I’d be willing to bet ISO6400 won’t be an issue at all for those of us that shoot RAW.

      Either way, I consider anything over ISO1600 a bonus, not really a necessity, so this looks like it”s going to be one amazing camera for me.

      • Funny how we all used to be contented on ISO Under 1600 or yet 800 back in the days (D300 for me). But now a usable 6400 is a must for most. But I guess that’s technology doing its thing on us.

        • Exactly. I really don’t understand what all the fuss is about. If you’re a sports photographer shooting basketball games in low light I understand the need for clean ISO6400, However for normally every day use it’s not that big of a deal. I just got back from a 2 week road trip where I shot over 700 photos and never once needed anything higher than ISO800.

          • Once you get spoiled by really fast glass and D700 you will be surprised where and in what light conditions you can shoot. Especially sports are pretty epic indoors, and of course weddings :P. But this is more of a daily walk-around street camera so high ISO is not that important really. I like the EP3 as it has that Oly JPEG look and the camera feels more like it has soul vs being all digital gadget.

      • Very True.

        I just ran a test shot between my X100 and the 5N with a C/Y Zeiss 25mm, both at 3200, f/5.6 1/125 – RAW. I then copied the RAW develop settings from the X100 file – including minimal ACR noise adjustments to the 5N file. I attempted to equalize WB between the files as well by using the eyedropper on a neutral element in the photos.

        What I found was that at 3200 they are much closer than the JPEG examples in Steve’s test above. I give the advantage to the 5N for what looks like better dynamic range, slightly better detail and an overall “more lively” look. Keep in mind, all ACR settings were equalized between them. Noise was so comparable that I’d be splitting hairs, but needless to say they were both worlds better than their respective JPEG versions. Both held excellent detail.

        These JPEG comparisons are valid for some but completely irrelevant if you shoot RAW.

  52. Great to see some nice examples. Truthfully I think the muddy high ISO scare was more due to the slow kit lens than the NEX-7, with the Zeiss 1.8 it can stretch its legs so to speak. It certainly is a touch noisier than the superb Nex5n and C3 16mp sensor, but hardly enough to be worth weighting it over.

    • Stepped away and coming back with fresh eyes, I have to say I much prefer the NEX-7 treatment… of these JPGs. Maybe the NEX has saturation up a notch which accounts for some of the difference but I’ve always found that my X100 looks softer at higher ISO and this photo illustrates that. There’s something of a halo around the lettering on the clove can, less so on the tomato soup but still present. Everything looks crisper. The damage to the O and L on the Cloves is rendered with much more detail on the NEX7 image.

      On the down side the “Condensed” lettering on the soup can is pretty mushy, almost painterly looking on the NEX. Maybe that has more to do with the point of focus and depth of field.

      Anyway… from a high ISO perspective I’m feeling reassured that I won’t hate the NEX7 compared to the X100.

      Hopefully Steve has some ZM glass kicking around for some tests. The 25mm would be nice to check out compared to the Sony Zeiss 24.

  53. E-P3 @6400 looks the best of all three at detail preservation, NEX-7 has too much in camera noise removal! Look at small text on the bottle (8.5OL OZ…) or small text X100 box (includes ADAPTER…) NEX-7 shots are almost unreadable, E-P3 seems OK 😉

    When you do some calculations… the NEX-7 pixel size is the same as E-P3!

    In fact you get sharper and noise free photo with E-P3 + photoshop NR. See URL!


5 Trackbacks / Pingbacks

  1. Steve Huff first look: NEX 7 w/ comparison shots to E-P3 - Page 6 - Micro Four Thirds User Forum
  2. Sony NEX 7 Sample Images and High ISO test « NEW CAMERA
  3. ソニーNEX-7 / X100 / E-P3高感度比較( | YOUのデジタルマニアックス
  4. What is Image Quality? - Page 8 - Micro Four Thirds User Forum
  5. Tanding Antara m43, Nikon V1 dan NEX-7 « Kamera Kamera

Leave a Reply

Your email address will not be published.