Sony a7s scores as the best low-light camera at DxOMark

Sony a7 camera test at DxOMark
The Sony a7s camera ($2,498.00) scored as the best low-light camera at DxOMark beating the Nikon Df and D4s (they still got a better overall score):

Sony’s A7 and A7R models have not unexpectedly generated a lot of interest with enthusiasts, and the new addition to the range looks equally intriguing, if perhaps for different reasons.

While it’s true the short register and low pixel density makes this model potentially more suitable with rangefinder type (non-retrofocus) wide-angle lenses, there’s little doubt it will appeal to those who value extreme low light performance over high pixel density. And then there’s the additional attraction of video, with 4K as an option (albeit with a third-party recorder). Even the price doesn’t look too bad against the Nikon Df, which is not only more expensive but can’t quite match the Sony’s low light capabilities and shuns video capture completely.

This entry was posted in Sony and tagged . Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • TinusVerdino

    Quite logical 12 vs 16 mp with the same sensortech.

    • Erik

      Most recent sensors from 16 to 36 (nikon) have almost had the same ISO performance. This is a step up.

      • TinusVerdino

        That would be explained by concessions in color depth and dynamic range. Less important for video.

        • NerdBuster

          I see, it all makes sense in your “brain”. Did you get past elementary school?

        • Please

          Right, that was precious. Please give us some more.

          • TinusVerdino

            A bully. Being a bully does not make you right. It makes you stupid.

  • Arn

    Less dynamic rangethan a aps-c like a K5 or a D7000… strange.
    Did they sacrifice those parameters to hi-iso performances ?

    • E-Nonymouse A

      Most likely some sort of design compromise to squeeze 4k vid and many other features into a small package, 5 FPS is pretty sad for such a low density sensor.

    • E-Nonymouse A

      Most likely some sort of design compromise to squeeze 4k vid and many other features into a small package, 5 FPS is pretty sad for such a low density sensor.

      • The 5fps is probably related to the output bandwidth. They read the whole sensor at 30fps or so for 4k so technically a high frame rate is possible. If you really want more fps, get an external 4k recorder and extract frames. Won’t be too different in resolution to full sensor. Not raw though…

        • E-Nonymouse A

          I was thinking more in relation to using as a still mode with such good SNR on a camera like this it’s a shame it can’t handle faster rates. Still sounds like a winner for non-terrestrial fun.

  • Guest

    I would not say that. Most recent sensors from 16 to 36 (nikon) have almost had the same ISO performance. This is a step up.

  • E-Nonymouse A

    Much as I may dislike the external design of the DF in many ways the Df and the A7s are closely matched matched in specs, to declare a clear winner. At best the A7s has a marginal advantage. and probably features some more UI design related enhancements that would put it ahead of the Df. Df loses primarily for its clunky, poorly design body and bad button placement. The imaging sensor inside however is still the venerable old D4 sensor.
    At the end of the day specs be damned, its the visceral experience of hands on usage that determines who really wins.
    A7s is a great first entry in competition with Nikon’s ‘light king’ sensor but needs more refinements, hopefully they do. If they do then they will most certainly be eating the lunch of several major camera producers.
    Depending on real world results one or both cameras may turn out to be fantastic astrophotography cameras, this is a category where higher ISO is common unlike terrestrial applications where it’s the exception.

    • doge

      Isn’t the main problem with Sony’s A7 cameras the lack of good glass? It seems like they’re hitting most of the correct notes on body/sensor design, but they don’t have a glass catalog anywhere remotely close to their FF competition.

      • If you like the sensor that much, you do realise that you can use lenses from Canon, Nikon and everyone else via an adapter? With Canon lenses you even get the image stabilisation.

        • d

          But, with the adapter, it would either be manual focus only, or the AF will be terrible

          • Manual focus isn’t manual focus any more the way it used to make people miss focus and kick chairs and flip tables once they got the envelop of prints back from the lab back in the day.

            Have you properly tried getting the hang of manual focusing with a A7 series camera? If you have, you wouldn’t say what you said.

      • E-Nonymouse A

        Considering that 3rd party makes do a much better job and the fact that as geno says, it fits the best glass ever made from any 3rd party (Leica, Schneider, Canon, Nikon, Zeiss). Zeiss does offer sony glass with AF on it IIRC. Manual focus is usually not an issue if your a landscape guy, i’ve seen studio guys who tackle MF with aplomb while snapping away pictures in a modeling shoot. Some adaptors allow almost full usage of 3rd party lense features.

    • Me

      Why does there ‘need’ to be a clear winner?

      • E-Nonymouse A

        There does not I just thought it was silly that someone would declare a winner based on such closely matched comparison, not likely see any visual difference at all in the picture taking process except under extreme circumstances.

  • PGi

    What a huge disappointment this junk was!!!!! I am so sad now, expected 2 stops better than the df. Now I must debate this vs df to back up my A7r.

    • Global

      Save $500 = get an A7S.
      Use your same lenses = get an A7S.
      Better ISO than Df = get an A7S.
      Better shutter speed (1/8000 vs. 1/4000) = get an A7S.
      Better manual focusing (peaking vs. a dot) = get an A7S.
      Infinitely better video (vs. no video at all) = get an A7S.

      I don’t know what you’re debating. The Df’s old fashioned dials that conflict with the menu system and are less functional than Fujis? Your primary camera is an A7R — get an A7S.

      • No one said it’ll be 2 stops better than anything. So you’ve disappointed yourself 🙂

  • stoooopid

    This looks like a fantastic sensor. Sony – please don’t sell this sensor to Nikon or Pentax so they can put it in a proper body with 12 fps and a proper AF system – my bank account could not handle such an awesome camera.

    • TheInconvenientRuth

      made me LOL

    • alreadyupsidedown

      Don’t worry, Nikon would never release such an obvious camera. Too many people would actually want it, a phenomenon completely against Nikon’s marketing vision.

      • hahaha so true. It’s almost like Nikon knows what the market wants, but they release the exact opposite hahaha

        • george

          Just a friendly reminder; the D800 was probably the most desired camera ever made, by any manufacturer (the same held true at a slightly lesser degree when the D700 came out). And they were both released by Nikon 🙂

          • bobgrant

            Nikon has done very well….D800 is the resolution king and what I see most in pro kits here in NY. The Df has also proven popular, especially for a 2nd body and the D610 is a terrific package for a low price. It’s been a long time since I even thought about my old 5DII.

    • Steve

      Yes Sony, don’t make money, who cares about millions of loss in sales, remember that you compete for pride, who cares about revenues.

    • Reality Check

      Exactly, if this sensor were in a D750 Noct specialized for low light photography it would sell like hotcakes. But no worries of that being made, Nikon is clueless about what their users really want.

  • Sonikon

    Remember kids what makes the A7s special is it has full sensor read out in video on a full frame sensor. no line skipping like every other full frame camera on the market.

    Where does the A7s destroy other cameras? in high iso Video. And par it up with the Atomos Ninja Star and you now have a 4K low light beast that nobody can touch.

    for photos. 12MP is plenty.

    • Ninja Star doesn’t do 4K, does it?

  • Clint

    ISO ratings…surely one of the most overated specs IMO. I would much rather have the resolution and dynamic range of the A7R over the crazy high ISO capability of the A7S.

    Aside from a very few exceptions, almost all cameras have ISO abilities good enough for my purposes.

    • Clint

      Also…am I the only one shocked to see that the A7R with 3x the pixels of the A7S actually has the better dynamic range??

      • The a7r sensor is very efficient. Also DxOmark’s DR measurement is silly anyway.

      • Thom Hogan

        There are a lot of folk who were shocked, but those that understand how sensors work probably weren’t. What surprises me is how flat the DR decline is overall, yet that it isn’t actually a straight line. That suggests that there is something else going on than the standard gain approach.

    • Mardock

      Depends on what kind of photography you’re doing. If you’re still shooting for theatrical, film, or television, this camera (with the silent shutter mode) is a godsend.

      Though I grant you that I’m describing a niche market … but then the A7 cameras are very much a niche product (so far).

      • NerdBuster

        Really? Have you got smart advice for different kinds of photography?

        • Experts

          LOL all sorts of experts here. Even selective experts 🙂

  • DuncanM

    Yeah, but a7 cameras have about as many lenses as PS3 has games.

    • Dadan Ardi Nugraha

      sony & zeiss are supposedly release a bunch of lenses at photokina….. well supposedly…

  • MB

    So Sony still makes good old D3 sensor …

    • Neopulse
      • MB

        I know that, D3 sensor was developed by Nikon and made by Renesas but Renesas is in financial difficulties for the last couple of years … the point was maybe Nikon should have improved that sensor and fab it by Toshiba to be able to compete with Sony …

        • Neopulse

          How would one understand “your point” if you write just one sentence that doesn’t mention anything about Nikon that they should have improved on the sensor and etc. You need to write more to make an actual opinion about something.

  • dxo

    Yawn dxocrap is always dxocrap. It’s so bad I am staying away from their software, too.

    • Eno

      I petty you troll!

      • dxo

        You petty me? Or you would like to pet me? Choose between ignorant and disgusting. I blame cheap schools…

      • NiagaraTim

        petty? did you mean pity? lol

      • greywoody

        Some pitiful petting there.

      • greywoody

        Some pitiful petting there.

  • BJ

    Purposely putting less than average number of megapixels into a camera with today’s high ISO technology is bound to get this result.

    • Global

      Is this supposed to sound bad? It sounds great for anyone not printing posters.

  • Global

    I wouldn’t compare the Df & Sony A7S — the Sony A7S is a much better camera, ADJUSTED for price!

    Not only is it -20%- less expensive at launch, its has a relevant competitive advantage in low-light against the Nikon Df, its lighter and smaller than the Df, and of course it is infinitely better at video for obvious reasons that Nikon crippled the Df intentionally this way; the Sony may even better than the Df with many manual focus lenses, including using Nikons on a Sony adapter:

    Nikon handicapped the Df by failing to offer a Split Screen for manual focusing. A huge disappointment for many users who had no support for 1 year, until a Taiwanese company started producing low-quality make-shift ones. Katzeye can’t make high-quality ones, due to Nikon failing to engineer the Df AF & metering system properly (using recycled lower technology, instead of making a proper design for the unique camera). So the Df doesn’t cope well with focus screens, preventing a proper solution. Nikon could have designed their metering and AF around this issue (they have the technology), knowing how their users would use the Df — but Nikon simply chose not to do it, because selling Df’s to AF users was just easier, and marketing the illusion of Manual focus historical uses was mostly just a gimmick. While every camera has its pluses and minuses, the Sony A7S has focus peaking and also magnified focusing for MF. As long as your lens isn’t extremely hazy or covered in jam and as long as you mind your shutter speeds, the Sony looks like an excellent effort for considerably less money with considerably more features.

    If it was $500 more expensive, one could nit-pick the small issues with the Sony — but it costs LESS than a Nikon D610 with better sensor ISO than the Df. So you can’t judge it any more strongly than an ENTRY level fullframe! For an “entry level” fullframe, this is truly excellent! And it proves, once again, that the Df was way overpriced and should have cost LESS than a D610, not more (or at least equal, considering the poor AF and other crippled factors like shutter/synch speeds like 1/4000; something like a 1/8000 is common sense when having just released extremely sharp wide-open f/1.4 and f/1.8 lenses.. seems they forgot about the needs of their new lenses AND of their historical manual lenses AND of video users on the Df).

    For the price, this SONY “entry level” fullframe is an excellent bargain!

  • nexus

    I’m saying that DX0 Mark is wrong on this, I’m just saying how sad it is that the ‘internet’ crowd makes ‘all’ there decisions by what these guys say ( suppose they were offered a secret kick-back one day ? )

    • Eno

      Again stupid comments based on emotional response.
      Dxomark objectively reflects the sensor performances of the cameras that they sport in DXO Optics pro software. Those dates are real and without them there will be no raw suport for that particular camera in the program. Get a life and don’t cry if sometimes you camera is inferior to a newer one.

      • dxo

        Get your elementary school cert, then you can talk.

        • Eno

          What a stupid comment, most of the people here are not native English speakers so get lost!

          My comment stays valid, if you take raw pictures for any site of any camera and edit them in LR you get the same results as Dxomark says in their sensor reviews.

          • dxo

            Really? Have you tried? How many cameras? How many brick walls? Entertain us.

          • Eno

            I smell you trolling so it has no sens “to ague with the wind that is passing”.

          • dxo

            What a pity, I wanted to learn something…

          • Eno

            Learn how to behave!

          • Ano

            Sorry mom…

        • bas076

          You got it an it wasn’t help you to became a human being. So, who told you that if you recognizing people language and standing on two you already part of people society?

      • 24×36

        Interestingly, the A7S and D3S measurements on DxO are basically an overlay for SNR @ 18%, Tonal Range, and Color Sensitivity, with the Sony just having some extension of the range. The Sony only really beats the D3S in Dynamic Range, which it should after nearly 5 years of sensor tech development. Based on the IR still life images, the stupid high ISOs just look like a massive smear of detail (i.e., heavy-handed noise reduction being applied), and thus appear quite useless.

        Still another potentially interesting sensor to put into Nikon DSLRs. Adds more to the argument for making a few different sensor/body combinations at each price point, to leverage R&D over a wider range of products appealing to more of their user base.

    • bas076

      Wrong on what? On your possibility to understand what is the meaning of science experiment?

    • E-Nonymouse A

      Actually most of the reviews from various sources that I took the time to find and read at length prior to purchasing what I own now are pretty damn accurate, DXO is subjective for the most part but still useful as long as its paired up with a few real world hands in depth dissections. Some sites are kind enough to offer them.

  • El Aura

    Some discussion and further data analysis in this (sub)thread (weaker CFA filters are part of it and some fancy stuff do reduce high ISO read noise to below 1 e- (but the read noise curve has several inflection points that need explanation):

  • El Aura

    The big difference is in high ISO read noise (that affects high ISO shadows and thus high DR). It is however not clear what Sony is doing to achieve this (ie, are they cheating a bit?):

  • Ufupuw
    • At 12,800, I actually prefer the Nikon. 400k ISO, no doubt, the Sony looks cleaner, but it’s mush vs. salt-and-pepper.

      • E-Nonymouse A

        The highest i’d go is probably 6400 but it’s nice to see that its still fairly clean at 12k.

    • E-Nonymouse A

      cool thanks for the comparison! I really haven’t seen any hands on stuff with the 7s and i’d love to atleast rent it for a week or two.

    • TeaBreak

      Nikon is way better than Sony. Even @ 400k! Denoise this and you’ll still be able to see texture whereas Sony is mud.

  • Xavier Duquesne

    Perfect ! Wait a few weeks and let’s see where the D810 stands, or just a few more weeks/month to see what will come up next as I am sure Nikon will be inspired 😉

  • Zibai

    Release the Kraken D900…..

  • AlanC

    DxO ratings are HIGHLY suspect. (Some even use the word “fraud”.) For instance they “normalize” their noise ratings by “adjusting” them for resolution. That would be something like Hummers “normalizing” their gas milage figures for vehicle weight or some other way to misrepresent a common metric.

    • MrSkelter

      You’re either very misguided or being deliberately unfair. There’s perfect logic to normalizing for resolution. Photos are printed at sizes based on need, not on sensor resolution. Failing to normalize would actually hurt cameras like the D800 because the ‘quality’ of each sensel is much less important than the quality of the image.

      DXO compares images. People online are fond of blowing things up to 100% and comparing pixels. It’s pointless. I’ve never seen a single pixel on a wall, or an image printed so that individual pixels were visible under normal viewing conditions.

      • TeaBreak

        Then perhaps you should look closer. Of course 100% view matters, indeed it’s the only thing that truely matters. Small or scaled down they do it all – therefore you don’t need to compare anything.

  • TeaBreak

    No wonder! Sony’s fooling us again. Regarding the DxO pictures it’s evident that they’ve doubled the sensor size without telling us. 😉

  • Back to top