Ok here you go! I have had a ton of requests for this test and I will be adding it into my official NEX-7 review as well. Many wanted to know how the newly released NEX-5n stacked up against the high megapixel but same size sensor NEX-7 in the high ISO department. So I set up my gorilla pod and did some testing.
DETAILS:
I shot the scene with both cameras using the Zeiss 24 at f/5.6. I shot one at ISO 100, then followed that with 800, 1600, 3200, 6400, and 12,800. I processed each file from RAW using Capture One which is supporting BOTH cameras. I left everything at default but turned off Noise Reduction so this was at 0 for each file. I wanted to show the RAW performance of each sensor not the JPEG Noise Reduction performance.
ALSO, I did NOT match shutter speeds. I shot these at f/5.6. Same lens as stated above. I did NOT match shutter speeds as I prefer to test a camera for real world use. In other words, if camera “A” chooses one exposure and camera “B” chooses a slightly different exposure then that is what you can expect from each camera. Matching shutter speeds when and if a camera chooses one slightly different is not showing what you will get from the camera when out shooting with it, and to me, this is the ONLY way one should test a camera. REAL WORLD as I have said from day one. Not “Scientific” as “Scientific” is not how we shoot.
I made 100% crops of each at full resolution. 24 megapixels of the 7 and the 16 of the 5n. I then resized the NEX-7 files to the same size of the 5n files to see how they stacked up in this way. Below are the results, let me know what you think!
First a web resize to 1800 pixels wide (click for that size). The NEX-7 at ISO 3200
–
and the same image with the 5n, resized to 1800 pixels wide – ISO 3200
–
and now the 100% crops from the NEX-7 and 5n at all ISO’s tested and at the native resolution of each camera
–
and finally, the 100% crops with the NEX-7 files resized to the same size as the 5n.
–
It is no question that the 5n does a bit better at high ISO. The NEX-7 is using the same size sensor but crammed way more pixels on to it. This will cause more noise which is why I am a big believer in large sensors and less pixels 🙂 But overall, they are not really THAT far off. Up to you to decide how important higher ISO shooting is. The files above were all straight RAW conversions and again, with ZERO noise reduction. Hope this helped some of you who were looking for this test!
[ad#Adsense Blog Sq Embed Image]
Sony Nex 5n is a lot better better camera, thanks for comparison it tells all we need.
more pixels less image quality , Sony is on the right track
Good job Steve.
THANKS TO ALL WHO RESPONDED TO MY POST ABOUT THE A-77. READ THE REVIEW , NOW I GUESS IT’S BACK TO THE NEX-7. LOW LIGHT IS MY FAVORITE TIME TO SHOOT. IT JUST MEANS MORE MONEY IN FAST LENS NOW. ALL GOOD THINGS COST SOMETHING , TIME OR MONEY. THANKS FOR THE AND GOOD REVIEW STEVE. I WILL BE WAITING TO SEE MORE.
Addendum: I screenshot the 100% crop at ISO 100 (!) and colorpicked it in photoshop.
The big letters on the label are about the same brightness for 5n and 7.
The shadows on the right side are BRIGHTER for the 7!
So I repeat: If same exposures would have been set, the picture of the 7 would have been OVERexposed, rendering any comparison meaningless.
Probably the contrast is set lower on the 7 and it meters to limit highlight blowout in the same way as the 5N.
IMHO Steve did it the right way -), ’nuff said…
I do large prints and take pictures with a tripod in low light so I suspect I’d get better results with the Nex-7 sensor than the Nex-5
From these tests, I would much rather have the 8MP resolution advantage over only slightly better noise at higher ISO’s. The difference is so marginal that I don’t think it matters at all. After doing some noise compression on both files it’s clear they will make out about the same. Might as well have a larger file if that’s the case.
seems strange as the noise almost appears worse than my M9!!!
Ozan, I’m not sure of your point here. I own older Nex-5, and I always shoot it in manual mode. The nex-7 is not aimed at beginners.
The bottom line is that, if you allow more light to hit the sensor of one camera, the comparison is unbalanced. It’s not as if I’m getting super techie and complaining about Steve not using measured, constant light sources or anything, but not even using the same exposure is a problem.
It’s only a problem if you are super nerdy and think high ISO is a competition. Does one really buy a camera just because it is the best high ISO camera? In reality most of the new cameras are great at high ISO. Seems silly to nitpick over 1/3-1 stop in high iso.
I don’t buy a camera for high ISO, and I rarely shoot over ISO 800-ish. Still, if you’re going to take the time to compare high ISO, it should be done with at least some control.
DF,
Obviously you didn’t read the text or the comments. Control was there. Read my explanation. For the last time, I ONLY review cameras on how they will be used and how they will perform when someone shoots with them. Just like this test. Period. I do not do hypotheticals or match exposure on cameras because if one camera is exposing differently here, it will be doing so when shooting it (and it does). These two cameras do not expose the same. Which one is correct? All personal pref on how you like your photos. What you see HERE is what you will get from each camera, not each camera matched to the other as that would be ridiculous to do because then the results would not be typical. Done with this explanation, and the comments are now being closed on this article. Why do you feel the need to keep on going with your comments that say the same thing over and over? This is the test. It was done the correct way as it SHOULD be done. End of story. Same lens, same aperture, same ISO’s. If the NEX-7 chooses a faster shutter speed then that is how the camera exposes the scene, and this is the results you will get. If you want to see the results of the NEX-7 when you overexpose it or bump up the Exposure Compensation then you can do that when you buy one or another site does this test. Thank you,
Steve
Thanks for the comparison pics Steve. They were informative and to the point. Not sure why some people get so worked up about the methodology given that it’s no secret that your evaluation is different than DPR and some others.
Did you see much of a difference in the JPG engines of the two? I thought the 5n was actually improved over the 3 and 5.
I agree 100% with Steve’s methology. If you look at DXO labs, you can see that camera “ISO” and actual ISO differ from camera to camera. Mostly, they are much less sensitive than they should.
When I look at the ISO 100 pictures, I think that both camera have the same level of brightness of the labels. If the 7 is more noisier, than this does not mean that it received less light. Instead, the camera design uses a higher amplification than the 5N. That is a design issue of the 7 and appears to be inline with Steve’s observation of frequent underexposures he experienced.
If Steve would have compensated manually, the labels would have been brighter than those of the 5N.
So probably the graduation curve is set differently, or the contrast in camera??
You should always shoot with same exposure if comparing. You can make up for curve or brightness differences in the raw converter.
But how to you decide exposure? By setting aperture and taking what the camera decides, right? So if the camera decides it needs less exposure, you say you will override manually and overexpose the bright sections?
My point being, that if the camera is built in such a way, you have two ways to fall of the horse. And I agree with Steve, that the real-world approach is to setup the camera and rely on the camera metering.
In consequence, you should disregard DPREVIEWS comparisons, too. Because they use different settings for each camera, trying to use the sweet spot for each. That is the way each owner would go for, too.
I set my Oly XZ1 at maximum of 4.5 aperture because that is the optimum for this camera. I wouldn’t dream of stopping at 4.5 with a APS-C camera. M43 are inbetween. Just take a look at the info button next to dpreviews comparison page.
The 7 and 5N have an electrically different sensor and circuitry. Its the same issue.
This is my 3rd post on this and nobody has yet to respond. As the Nex 7 seems to be about equal to a Micro 4/3’s high ISO the EP3, Can the Nex 7 do INCREMENTAL ISO? What pro camera CAN’T do that?. In that way, since you may only need say 1000 iso you HAVE to go all the way up to 1600 = NOISE,..or you need 2000 iso you HAVE to go ALL THE WAY up to 3200 ISO. TON OF NOISE unnecessarily.This is HUGE and a BIG disadvantage to a camera that is barely besting a Micro four which HAS incremental ISO. Hence if this 7 can’t do incremental seems a $699 NEX 5n makes a LOT more sense vs a $1,350 semi pro cam (with kit lens). Can ANYBODY please answer this question. Does the 7, DO Incremental and if not, Shouldn’t it? HELLO!!!
Thanks Steve for this interesting test… Even if I don’t really care about iso performance above 1600 as I use fast lenses…
Results are close and I see no point to get a 5n over the 7 ! I have a Nex-5 for a year now and all I need is in the 7 (finder, manual controls, hot shoe…). I don’t like controls of the 5 and I don’t think that a touchscreen is the solution (for me)…
When I read some comments I wonder if those people take pictures in “real life” or in a laboratory ?… 🙂
Steve, I can’t wait for your Nex-7 review with M lenses !! 😉
Thanks for this interesting comparison – even if the result was to be expected.
In my opinion, the 7 would be a much more attractive camera if Sony had chosen to reduce the number of pixels. My reference is still the amazing universe of the image quality of the Nikon D700 (which is only topped by the D3s’). Have I ever wished to have more pixels? Absolutely no.
Great comparison. Thanks for all your work
I was impressed that both cameras in raw were so much better than previous 5N pictures with noise reduction. I guess I will have to start shooting in raw.
thanks Steve
the high iso seems good enough on NEX7 , the fact that it is not the best shoud be quite irrelevant for most users, but I think most people you included would have preferred the 16 mp sensor in Nex 7.
What I am worried about is how slow will lightroom be with the 24mp raw files….
And one question for you: will the NEX 7 replace the EP3 as your take anywhere camera? is it that good? Maybe you could elaborate on that in a blog post, I am sure many would be curious to know…
From a “real world view” I would not discredit the 7 for its lesser high ISO performance which is still very good compared to last generation EVIL/DSLRs. “Scientific” tests and comparions between EVIL/DSLRs on other websites fail to mention the lenses used. I would expect that 1.8/24, 3.5/30 and 1.8/50 are better than the zoom lenses and the 2.8/16.
Ultimately, the decision between 7 and 5 is the user interface, size and price. I don’t mind the menu-based system on the 5 (I can change ISO faster than on my M6), but I prefer a see-though viewfinder at eyelevel over an LCD screen that I hold at arm’s lenght. The LCD is difficult to see and operate without reading glasses, the grip is less stable than with the camera pressed to the forehead and it places the camera closers to the subject which makes a noticeable difference with wideangle lenses at short distance. I would love to see a simple optical viewfinder with frames for 24, 30 and 50mm lenses, covering the most important focal lengths, just like the Leica M2. Perhaps Cosina/Voigtländer is listening.
It’s such a shame Sony don’t make a “Nex7s” with a 12 megapixel sensor and incredible low light capabilities. Who needs 24 megapixels? I don’t think many billboards will be shot with the Nex7…
I own three cameras: a Leica M8, a Panasonic GF1 and a Nikon D200 which each serve their own purpose. And while none of them are perfect, they have one thing in common: despite the highest pixel count being 12 megapixels in the GF1 I have never longed for more pixels in any of them. Not ever. And I tend to print quite large…
Now, I am not saying that there are no uses for 18, 24 or even more megapixels. It just seems strange to me that Sony seems to think that users of the Nex7 are interested in megapixels (and base their buying decision on that). I personally, and from what I can tell most readers of this site and others would agree, am much more interested in low light capabilities. Maybe it’s because megapixels are still “the thing” in Japan, I don’t know. Sad development though…
Thanks Steve, I found the comparison a useful one. As my choice is not between a 5N and a 7 but more likely a 7 and an EP3 or a 7 and the new Panasonic GX1 (which may have the GH2 or G3 sensor – unconfrmed), are you able to do a similar comparison against the EP3 with maybe your M9 thrown in for good measure… That would be awesome. Cheers
Steve, I appreciate your well written and timely reviews and tests. I’m really looking forward to your comparisons with various Leica lenses. In terms of this test, I think it is the term “real world” that is causing a difference of opinion. You state that you “Set the aperture and ISO and shoot.” I do things differently, so for me “real world” has a different meaning.
In my “real world” low light shooting, I would choose the aperture that would give me the DOF I wanted (or could live with) and the minimum shutter speed the situation called for. So the only things that would vary between two cameras I was comparing would be the ISO that each camera selected and possibly slight differences in the exposure (depending on each camera’s metering). In the “real world” I rarely set a specific sensitivity and then accept the range of apertures and shutter speeds available to me – unless I’m at the limit of acceptable image quality.
Only way to fix all this is for Sony to work very hard with the best minds in photography and come out with a FIRMWARE UPDATE for both 5n and 7 to make the pictures more accurate in color, right AWB, sharper and more life like Images, to keep all the critics quite and get back there number one place in Image-Quality. It’s sad Sony has lost its #1 place in Image-Quality which they had after releasing nex5.
Steve – Great test. I really appreciate the resized images, instead of just 100% crops. Personally, I think there is very little difference between the resized crops. Not enough to have me cancel my nex-7 order.
So if anyone here now feels that the nex-7 is not up to their standards, please let me know, I will gladly take it off your hands 🙂
Steve,Thanks for comparing 5n with 7. I am sure lot of guys are happy.
Please don’t stop at just high ISO comparison . Compare these two in full and complete. Compare the sharpness of images, accuracy of color, white balance, contrast, Dynamic Range and Exposure. Exposure may be an important comparison, as exposure in 7 is under question. With low-light pictures lets also have some Portrait, Landscape and fast moving like sports.
Also compare the speed of Auto-Focus on both,shutter lag and also how fast the image is written on to the memory card,time between each shoots, because of the increase in number of megapixel in 7.Thank you and keep up the good work.
I am really finding it interesting that people’s expectations for high ISO performance ignore the reality that high ISO is used for maybe 5% of anyone’s shots. Nobody should be surprised by the edge the 5n has in this department. I make large prints and with shots with ISO between 100 and 400 99% of the time. I’ll take the extra MP thanks. But hey, if you want to spend almost the same amount of dough for a 5n and a external viewfinder even though it lacks the usability advantages of the NEX-7, by all means, go right ahead.
5% of yours, I’m sure. The cleaner files can be appreciated even at more moderate isos.
The price difference is $250. I wouldn’t say that is insignificant.
The extra dials sound great if you’re going with Nex lenses. The lineup is just too small right now though to really be impressed by. The 5n controls just fine with manual focus lenses.
I just can’t get behind paying more for a sensor that has resolution I don’t need and worse ISO performance. If the 7 was a superior camera in every way but size, it’d be no contest.
As far as I can see the 5n’s results at 6400 compare to the 7’s resized 3200 results. The 7 does seem sharper in this comparison, but I think the 7’s between 2/3 to a full stop behind the 5n in terms of actual noise.
So let me get this straight, you’re saying you’re doing a test for REAL WORLD results… but you’re not actually doing a real world test but a controlled test.. This makes no sense.
What makes no sense? Real world as in what the camera will give you at whatever ISO. Set the aperture and ISO and shoot. The camera chooses exposure, not me. Every image in my review of the NEX-7 showed the real world images, this shows real world results. There are plenty of images at every ISO in the review if you want to see that. This was done for those wanting a comparison between the 5n and 7. No need to get rowdy or upset 🙂
Steve, stop replying to these comments. You’ve already more than explained your testing method. If other’s think it’s insufficient to their scientific needs, let them.
Steve I love your site here and how much time and effort you put in to all your work here.!! I have been bashing the 7 a bit and I guess I must admit that’s kind of DUMB as I can’t really bash a camera I haven’t used myself. Not really bashing, but just seems everybody WANTS the NEX-7 to be the Nex WINNER yet it keeps coming down to the 5N looks better at high isos and still “I don’t recall seeing that you ever said the the end results LOOK BETTER on the 7 vs the 5n”
. Simply may I ask..Controls aside, EVF aside, Camera LOOKS /Feel aside..
at the end of the day “WHICH pictures look best you” on average the 5N or he 7? Or a tie? I Know there are many other factors.I am SIMPLY asking at the end of the day is there a winner as to FINAL output if you HAD to choose not knowing which camera took the pictures. I am about to cancel my pre- order on the 7 but maybe I shouldn’t. Thx and thanks again for all your hard work here though it mist be sort of fun. 🙂
Thanks, Steve for the great practical and useful review. You would never satify all the people around, however, your site give another kind of review that I could not find elsewhere on the web.
If you did all the lab calibration in the world the other half would complain that’s not how the camera works… It’s a good useful comparison. Besides, they look the same to my eye, of course I carry a white red tipped cane when I walk around. 😉
I think either you do real world tests or you do a controlled test. You’re doing a weird hybrid here and I think that’s what people have a problem with.
How many people do 100% pixel comparisons of several ISOs between two cameras of a vitamin bottle? That’s not “REAL WORLD”. Which is Fine since what you’re trying to show is a controlled comparison. Either you should be scientific or you should just give your overall impressions and experience.
Trying to simulate a lab sort of test and getting the simple stuff wrong is just misleading.
So you never use exposure compensation and rely entirely on camera metering? Most cameras meter differently from each other, and it would make sense to equalize this in any real world test.
Interesting. The 5n is clearly better from 1600 up. While I briefly had a 5n, I shot an identically lit shot at 3200 with it and my D300 using an adapter and the same Nikon lens. That’s two (I think) generations difference in sensor/processor technology and frankly I didn’t see much to chose in noise levels. I didn’t match shutter speed either, but the two exposures were, to my eye, identical (as are those above). I’d love to have had a D7000 to add to the comparison as that’s reputed to be a full stop better than the D300 and also reputed to be the same base chip as in the 5n (Nikon, of course, adds their own tweaks).
I loved the size, balance, and feel of the Nikon glass (some D primes) on the 5n. All in all it was a sweet little set up, but the oled finder is just flat out not available anywhere and I really struggled focusing using peaking and the LCD.
I have seen several NEX5N Oled finders available NOW, on e-bay.but most are from Asia but IN stock.
Yeah, I should have said in US market. HK and I don’t do much business these days…
But you CAN get it. I have got stuff from HK on e bay with no issues. But don’t you feel IF you could get the finder, it’s a plus over the 7’s in that it it can be tilted UP? There have been a ton of times I did some shooting on timer to take pics of my wife and myself in an Arboretum type place where even on my K-5 I could not use the viewfinder as I had the camera placed rather low on a tripod.
With the Nex 5n finder IF you can get it :(..you could just tilt the EVF 90 degrees UP, look DOWN thru the finder and see your subject..Can’t do that with the 7. Plus you’ll have about $300 left after and better high ISO blahh 🙂
JUST ONE MORE FLY IN THE OINTMENT AS TO WHICH ONE WE SHOULD GO TO. NOW IT SEEMS LIKE I MIGHT WAIT TO SEE WHAT THE TEST ON THE SONY SLT- A77 SHOW. NO PROBLEM SINCE NONE ARE AVAILABLE YET. THANKS FOR THE GREAT REVIEW STEVE.
check dpreview, they just posted an a77 review. It’s noisier than the nex7 due to the mirror reducing light hitting the sensor.
Actually DP Review just posted their review of the 77. It ain’t Steve but it’s a full test.
Doesn’t the A77 have the semi translucent mirror? That means that there is light loss to the sensor, something that wouldn’t be true for the NEX-7. The A77 should have worse performance in that regard.
Steve,thanks for the detailed no-nonsense comparison test.
It just goes to show even now technology has taken camera performance to unprecidented new heights, that more pixels does not necessarily mean better ‘Cleaner’ pictures. I know which I prefer…..the NEX 5N in this test.
To be honest , I was expecting better from Sony with this 24 mp sensor, but when you compare it directly to one of their other products, the difference jumps right at you.
I guess the real crunch for me,would be do I sacrifice image quality in favour of better manual control offered by the NEX 7 as opposed to the 5N. Thats a toughey, lol
Still, both are fabulous additions to the mirrorless market. Certainly makes my Panasonic GF-1 look rather dated all of a sudden.
Further to that, I’ve just played with my friends new iPhone 4S, and was staggered even further with the advancement of digital image capture. Outstanding.
All the best
Gav.J
Gav..you just said it all..and my problem with the 7.
You wrote: “I guess the real crunch for me,would be do I sacrifice image quality in favour of better manual control offered by the NEX 7 as opposed to the 5N. Thats a toughey, lol”
If you read most of the replies here seems just about everybody is saying the same thing. The 7 is ALMOST as good as the 5N ..But Nobody EVER says/said the pic quality is BETTER on the 7 vs the 5N . Only physical aspects of the 7 are better like the EVF, built in flash Tri Navi but WHAT ABOUT THE END RESULT?? It’s the loser!! Again what’s the sense in great controls if the car is actually a Ford Pinto ha.
Not saying the 7 is junk. Matter of fact I’m having a tough time in choosing between the 7 and the 5n. But if I put how nice looking the 7 looks vs the 5..it’s controls, vs the 5 Put it all away and concentrate on my wallet and concentrate on the final product, “The Pictures” The 5N is the clear winner. Even my X100 as beautiful as it is? At the end of the day it’s it’s pictures that still wow me, not it’s good looks..or controls.
YAK ,I wish Sony didn’t put a 24 meg sensor in this 7. Ok I said enough..thx.
Yeah, the NEX-7 is going to be better throughout the low and mid ISO range and very competitive at the top. If you shoot ISO 1600 all the time, this probably isn’t the camera for you – the D3S is.
I’m starting not to GET this whole Nex thing. Meaning the NEX-7 clearly is inferior to the 5N re: iso noise. There is a new review at DP with the Sony SLT-A77. They keep on complaining about the NOISE and it’s SAME sensor in the 7. Yes with a mirror but gotta be pretty close I figure NOT good.
Not sure if it’s the raw conversion but the 7 is definitely showing a lot moire pattern issues in the crops. If it’s doing this to a can of beans what about a pair of jeans..it’s gonna be a mess MAYBE.
The problem for me is the 7 is rather expensive. With just an OK kit lens it’s $1,350 same as a Pentax K-5. When you get home at the end of the day and just LOOK at your output how to justify the high price on the 7? Seems the 5n’s IQ is as good in every way as the 7..better high ISO, smaller, actually if you add on the EVF ($349) yeah not great) is more useful in that it can be tilted up/down that can be a huge help in some situations. Yes sure it’s more convenient having the EVF being built in,. But the 7 is nearly twice as expensive as the 5N and pretty much it seems everybody is saying WELL it’s ALMOST as good as the MUCH less expensive 5N.?? Why do you want ALMOST as good and have to pay an extra $600 something for an inferior result. Sadly I think Sony blew it here. They COULD have had THE camera of the year,..instead in my opinion they did what Fuji did to their great EXR series. Sony put TOO many pixels on the sensor plain and simple!!. It’s over!
Sooner or later it’s going to bother people that this sensor is too noisy. That’ fine if it were $600 but almost $1,400 with kit lens? I gotta cancel my pre order and either buy the 5n or just forget the whole NEX thing. Great controls are GREAT, but controlling what? You can put great controls on a Kodak it’s still a Kodak 🙂 I say the 5N has the best sensor in the NEX series and Sony could have had it all had they put it in the 7.. Instead we have the NEX7 a bit large, too many pixels…TOO much money and what’s the point? It’s not pocket-able..what is it? Sorry I’m just disappointed..
The increased noise in the A77 isn’t due to the sensor, but the semitransparent mirror that diverts light away from the sensor. Because of this I expect the A77 to preform a bit worse in this regard.
we were typing same time, I was slower…
With its pellicle mirror, the 77 diverts some light away from the sensor and I’m not sure that that doesn’t hurt higher ISO performance.
Hmmm, I think I’ve been spoiled by my D700. Even ISO 800 on both of these makes me a bit squeamish. Still, not unreasonable, considering the pixel density.
Steve, would you say that this is also a fair comparison of AWB betweeen the two cams? I dont see a difference here. I didn’t like the results with AWB in my 5n. Did the raw conversion eliminate any difference in color?
Each camera was set on AWB and I did not change the WB during the RAW conversion. I find both are about the same in the AWB dept.
Did you use exactly the same exposure (f stop and shuuter speed) and the same lens for all direct comparisons?
See the description and the comment I left earlier for my thoughts on this.
As the owner of three rescue dogs, I’d like to see these tests done just with cans of Kibbles @ Bits, Alpo, and Pedigree.
Thanks Steve. Results are pretty much as we expected. That 5n is a killer little camera except for the deep menus.
If I need anything over ISO 800, I’m probably late for dinner and should pack it up.
best post I’ve read here on Huff’s website ever!
Excellent!
I think Camera manufactures lose their minds once they get one good product out. Once they get something really good, just hold on to it and make it even better by only fixing the things which are not good or needed. That’s what you think people would normally do. Make what you have into a perfect or the best and then go to the next level.
Making drastic changes trying to go higher and higher with no good results to me is insane. Nex 5 had very good Image Quality which made history and changed a lot in world of photography, which is made even better with 5n. They should keep that excellent image quality and work on fast Auto-Focus,in body image stabilizer, more Accurate Color reproduction like in x100 or better, more accurate White-Balance, Better or higher Contrast ratio, better Low Light performance,Sharp and crisp pictures. Work on better Lenses may be. To get more realistic pictures with better 3D pop with No flash or any other assist lights. Try and work on getting pictures, where people feel they can touch the person or the object in the picture they are seeing.
Canon I think is getting the Point, there new cameras or more focused on Image Quality and Low Light performance, and not on Mega-pixels. The new 1Dx is down to 18megapix and new Sx40HS is down to 12megapix. If Sony is making the Sensor for Nikon D3s which is the best DSLR ever, can’t they put the similar quality big sensor in Nex 7 with same great performance in a small body which people love.
People love the small size and light body of Nex series, very easy and convenient to carry. This small body combined with Performance of D3s will be the epic or the ultimate photographers dream. No matter how much a photographer will love his work, he will be tiered after sometime carrying big heavy gears. If cameras are small and convenient, every photographer for sure will shoot few more extra pictures. I am sure Steve will definitely agree with this.
The D3S sensor is not made by Sony. Nikon has stated it is designed internally and fabbed by an unidentified contract manufacturer, but they’ve stated Sony is not that manufacturer.
I think Nikon is stretching the truth. For ‘designed’ – read ‘specified’. Nor did they fully deny industry claims – i.e. it was not Sony manufactured.
Nikon has NO in-house fabrication facilities – so, as Nikon Rumors said some time ago, if not Sony, then who?
Not Panasonic (they claim) – obviously not Canon, Samsung, etc. The suggestion that some tiny nondescript electronics company in China, Thailand, Korea, Taiwan, or somewhere similar is sitting on an under-utilised, full frame sensor fabrication facility (of this quality) is laughable.
If true, dozens of manufacturers would be beating a path to their door – not wasting millions in R&D – and production facilities – developing sensor designs themselves.
The phrases “designed (specified) by Nikon” and “made by Sony” are NOT mutually exclusive – particularly when you are attempting to convince consumers you are not now a small fish in a big electronic pool.
But didn’t Sony do it? Nex 5 to Nex 5n with even better image quality, small changes in layout and touch screen added? The similar logic applied to 3 and C3.
The Nex 7 is another sub product line and maybe in a year comes Nex 7n with improved noise handling (but at 24MP).
There is not really anything to it. At higher ISOs they are very close, with a tiny advantage to 5N. And at lower ISO (where most photos are taken) Nex-7 has a higher resolution.
Add better build, built-in EVF, tri-dial navigation, and it’s clear: . Nex-7 is a better option (though at higher cost).
I don’t get the sense that the 7 has better build quality. They seem about the same. The 7 has the advantage of more dials and buttons though, but that’s more about handling than build quality.
I find the removeability of the evf is an advantage since you can choose to make your camera more portable if you need it to be… and its a really solid EVF since it screws in. The fact that you get tilt is a bonus too.
Advantage isn’t all that clear to me!
You must not ever use strobes if you don’t see the advantage in having the EVF built-in vs add on. However, for those of us that do use strobes an add on EVF makes a camera a total non starter.
You’re right, I don’t see myself using a bounce flash on a tiny camera, especially since it does ISO 3200 and 6400 so well. I reckon that an slr would be more appropriate if flash is such a big concern… The contrast detect af just doesnt compare in tricky lighting where you’d have to use a flash.
In the instances where i’d use a strobe the screen would be just as good. The viewfinder is best for daylight shooting. I mainly use and prefer optical viewfinder only cams but have found the tilt lcd to be a very flexible way to shoot with the camera, especially when size is more important that being able to put the camera up to your eye.
I rarely use bounce flashes either, but I do use RF triggers all the time. Also, since the NEX-7 has built in wireless control over Sony’s TTL I’m thinking about just giving that a shot so I no longer have to fiddle with triggers.
Also, I rarely ever use flashes because a scene is too dark; I use them to as fill light, or to improve on the natural light. As fas as DSLR’s being better for that, for now possibly, but as soon as global shutters arrive in CSC’s then that will no longer be true. Once the global shutter first makes its apperarence we will no longer have to worry about 1/250th and slower flash sync limits….finally.
The main advantage of having a small, light camera like the NEX isn’t all that important in a setting where you’d be in a studio or a controlled setting while using strobes, no?
These cameras seem to me to exist to either be a step up from compacts or to be a backup to a larger system camera. In both cases they’re meant to be a balance of size and quality, such that you can bring the camera with you everywhere and still have a decent sensor size.
I get it if you’re gunning for the NEX series to be your only camera, and that the 7 would be the only likely candidate since its feature set is the only one that really accommodates for this. But to me the whole point of these systems are compromises to make the cameras as small as possible while still keeping the highest possible image quality, and to me the 5n does that better than the 7 in both cases.
Great info. I bought the 5N as a low light high-ISO alternative to my Sony A-850. Basically for indoor stuff. I think your study confirms what I thought. The two NEX cameras are close with the 5N slightly better. I think this means that the jpeg engine in the 5N must be what accounts for the much greater differences I’ve seen between NEX-5N jpegs and NEX-7 jpegs (seen on several web-sites) than between the corresponding raw files. Thanks for your excellent work. And I’ll probably buy the NEX-7 too at some point to take advantage of the excellent controls.
Regards,
John
Based on the ISO 3200 examples, the exposures weren’t controlled as the NEX 7 received 1/3 of a stop *less* light. Which is always going to increase noise.
See my previous comment as to why I did not match shutter speeds.
I thought title of the article is “ISO wars between 5N and 7”, not “differences in metering at same ISO between 5N and 7 and result that is consequence of such metering”.
And while speaking of metering, if that is the subject, then we would be interested to know which one metered accurately (and underexposing does hide the noise).
Steve, could you list the shutter speeds for these? I assume that shutter speeds remained consistent between both cameras, right?
Absolutely not. I’m a big believer in testing a camera in the way it will perform for the person who is shooting with it. In the real world I would not be out trying to match the exposure of the 7 to the 5n. I would be letting the camera choose the exposure 95% of the time and then overriding if necessary with compensation. What you see here is how each camera exposed the scene and the resulting high ISO performance. Matching the 7 to the 5 is more of a scientific test and not one I agree with for real world use.
I appreciate the time, and it’s great that you provided same size crops, but not using the same exposure is an inherently flawed approach, both scientifically and from a user perspective.
Only shutter speed and aperture account for exposure in digital. ISO is simply an arbitrary gain number that varies from camera to camera.
To judge how any two cameras compare in lowlight, you must allow the same amount of light to hit the sensor, and then you can adjust ISO and/or exposure gain in the raw converter to make up any differences.
Unfortunately, what you tests tells us is that the NEX-7 shows more noise than the NEX-5n when the NEX-7 receives less light to the sensor, and we didn’t really need a test to show that. 🙂
I think you are just reaching to find a good reason for all of those MPs crammed in to that sensor, when you know deep down inside that the 5n has a better sensor and better noise control. Sorry the NEX-7 let you down. Sorry that you got duped into believing that more MP= a good thing. Don’t worry….99.9% of people think that too.
I don’t own either the NEX-5n or NEX-7, and this isn’t some kind emotional opinion. I’m simply stating facts about comparing cameras. The funny thing is that it seems the majority of enthusiasts still believe that more MP is bad, which isn’t necessarily the case (given the same sensor technology,) if output is compared at same size, which I applaud Steve for doing.
Trust me, I understand were you are coming from. But I disagree with your dogmatic approach that there is only one way to do such comparisons.
You see, on per body ounce comparison basis, I’m stronger than Mike Tyson. But I doubt this would matter in a real life situation.
But the way I tested here WOULD and DOES matter in a real life situation. If I went out with the NEX-7 and shot in A mode and let the camera choose the exposure I would get these results. If I shot the 5n in the same way, I would get the results you see here. I would NOT buy a 7 and go out and shoot it trying to match the exposure of other cameras. I would trust the in camera meter which is what most of us do, especially when buying a camera like this. I show what the CAMERA delivers when you buy it and use it. NOT WHAT it MAY do if you adjust it to match another model. I think that is silly, always have and always will.
Steves approach is correct in IMHO because that is how real life shooting would be done.
Steve that’s fine to make a review based on how you personally shoot the camera, since it is your blog, but you should at make it known in the review that you’re testing it in this unusual and unscientific way. Or, at least post the shutter speed numbers, so there isn’t any accidental duping going on. Your test seems to imply that the same exposure was used for both cameras, which isn’t the case.
Not unusual for me as I have been doing these kind of tests for 3 years now. I ALWAYS STATE I do not do scientific tests. Anyone who has followed this site knows this. I test for REAL WORLD usage, not SCIENTIFIC which means nothing when using the camera for taking photos. If the NEX-7 exposes like this when using it, THESE are the results you will get. Same with the 5n. NOWHERE do I state that the same shutter speed was used. Same lens? Yes. Same aperture? Yes. Same ISO’s (of course). Tripod? Yes. Shutter speed? NO. 95% of those who will buy the NEX-7 and especially the 5N will either shoot it in A mode or P mode and in some cases IAuto mode. In other words, the majority of us. This test shows what to expect when shooting the camera in these ways. Again, I do not do scientific tests. Matching the shutter speed tells me nothing about using the camera.
I agree with Steve here….these are results that you will SEE in REAL use! People don’t use their cameras as scientific tools in extremely controlled settings. They trust the camera to nail exposure properly and comparing what the two cameras give you is real life.
I think the test is fine if the exposure levels look the same out of the camera. This makes more sense than shooting at the same shutter speeds for all cameras and having to correct for the exposure differences in sensor sensitivity in post. Two different sensors often don’t expose ISO 100 the same, for instance.
Comparing two optimally exposed images at all isos like Steve’s done here makes sense to me.
I guess there is no “perfect” comparison test. Steve’s test assumes nex7 chooses a shutter speed that is consistently different from nex5n in all real life situations which can be infinite and therefore the test might lack consistency, whereas choosing the same shutter speed for both cameras would correspond to only a small set of “possible” real life situations among infinite and would not tell us at all the amount of noise in an actual shoot. Given both cameras are from the same manufacturer and their relative sensor sizes are obviously fixed I’d go with Steve’s approach on this one.
And exactly why when you read other review sites and they have their “test” shots, which are very scientifically controlled, has never translated to real world shots for me. I bought two cameras because of these sites – both were sold on craigslist in a month – the real world performance sucked even though they were “technically” superior from the test photos – ISO performance, low light , etc – Guess if I was in a studio shooting this could mater since its a controlled shoot but otherwise helps me personally very little.
DF makes a point here, but the point is more valid to comparing a manual setting VF styled camera like the Fuji x100 to an Leica M9 than it is with a more consumer aimed camera – which I believe will be the majority of NEX-7 and NEX-5n users.
In the case of cameras like this where a majority of the users will be using the classic “P” (for “P”rofessional) mode and depending on their situation (if they shoot mostly night or day) Steve’s article is most useful in helping them decide on which model suits them better.
In the case of an x100/M9 full manual “shoot-out” (hint, hint Steve!) then shutter speed, ASA/ISO, f-stop value, WB would be an absolute necessity for a proper basis of comparison.
Nice work on the ISO war Steve!
Doc.
Thanks for taking time to shoot and publish this *tremendously useful* comparison Steve.
I don’t suppose you had your X100 on hand when you shot these? It’d be great to how the 7 output resized stacks up to the excellent output of the X100 at 100, 1600, 3200, 6400 ISO.
The NEX 7 seems to create angled lines on the blue label at ISO800 up to 6400. Is this moire? It is not at ISO100, which shows that it is not a detail on the label itself. D!RK
ok, 5n is winning on noise but wouldn’t 7 deliver more detail at low iso ?
like for example with trees and feathers; I know you can combat noise and imho the picture quality in terms of delivering fine details and good dynamic range is more important
Much appreciated, thanks!
To my eyes, up to 3200 the differences are minimal enough not to matter (to me). At ISO 6400, the NEX-7 loses it, while the 5N still hangs on. Since I have never shot over 3200, and rarely at 1600, these results are perfectly acceptable for me. I’ll be keeping my pre-order.
Yep pixels have been on a stable or even downwards trend in the last few years seems that SOny is bringning it back. The benefit is in biright light where the resolution will be very good especially if you try to crop to isolate your subject, I know, it is like the crappy digital zoom but more pixels give you the chance to do it and still have a good 8X10 print to put on your wall and in the end, I don’t care what camera took it, if i put it on my wall and people ask about it (where it was, How it was taken, bla bla) then that is what counts. The last pics to hit my walls were all from the Canon G3… Time to shake that hard drive to get a few more….
It just goes to show you….better pixels is better than more pixels. Seems there was someone spouting off around here that there is always benefits to going up to more MPs, and I think that’s clearly wrong. Canon’s doing it right with their flagship now….
I was one who said “I will take the more pixels”. As I see it, at low ISO I’m fine and at high ISO I can bring the images back to match the 16mp of the 5N and they aren’t that different.
oh, i forgot to say, thanks for the great comparisons, esp with the resizing done too!
Thanks!
I’m curious to know if there is a difference in the noise of the NEX-7 if the resolution is lowered to 4240 x 2832 or 4240 x 2400?
Overall, looking at 3200 on the NEX-7, it’s still pretty darn good! We tend to be spoiled with very good high ISO in most cameras these days. I shoot some film at 1600 and either of these provides a much clearer image.
I’m no expert but apart from nicer looking noise in 5n the amount of noise is not that much different between those two. Or maybe I’m too blind to see the difference 😉
Overall, looks nice and usable 🙂
I should’ve been more precise, up to 3200 the difference is not that great but after that 5n clearly wins.
But for me 6400 is not that important, after one year with nex-5 there was only one time where such high ISO was really needed.
“24 megapixels of the 7 and the 12 of the 5n”
You mean 16mp.
Ooops, just fixed that. Thanks!
OMG! i almost peed when i saw the noisey high iso banner for the title of the article!
too funny! 🙂 thats why i keep coming back here and to other choice blog/review sites… the real person humor and silliness in them!
Strange how the dark noise looks similar up to ISO 1600 but the light areas on the label in the NEX-7 in the ISO 1600 show a definite moire pattern you don’t see in the 5n shot. I don’t think I’d ever use either one of these over ISO 1600, but that still much better than the ISO 640 I get from my M8.
I think we may be seeing the printing dots.
These are out of camera JPEG – how can they be ‘printing dots’? Need to be printed to show ink-jet print ‘dots’ – neither of which would show up anyway unless you magnified it thousands of times … inkjet dots are measured in microns … i.e. microscopically small! This is original output digital noise ….
The printing dots on the label.
No, they’re not printing dots on the label. If they were the effect would be even more pronounced at ISO100.
Commercial printing is done by offset litho process – which, to oversimplify, is a little like glorified potato printing … except it is a rotating roller that transfers a film of ink to the object. Definitely NO printing dots involved, at any time, anywhere, no way. It’s noise (and as ‘doug’ says – would be more pronounced at ISO 100 anyway)
Thanks Steve,
I have the 7 on pre-order, but it looks like the 5N will be hanging around too.
thank you steve for this test – now I have a reason not to drool too much over the nex7 (I just bought the nex5n …)