Michael Reichmann from Luminous Landscape (Click here) posted his Sony A7s first impression review. Here are just some of his key considerations:
– I have to say that the Sony A7s produces some of the cleanest, most pleasant images I’ve seen this side of a medium format system.
– On the basis of these tests I would judge the A7s to have almost a three stop advantage over the A7r between ISO 6,400 (which is as high as I would use on the A7r) and ISO 51,200 where the A7s reaches its useful limit (for my type of shooting).
– My biggest beef, already mentioned, is that Sony forgot to allow Silent Mode / Electronic Shutter to be available as a custom function.
– Otherwise, the A7s has all of the goodness as well as minor flaws of its predecessors.
– My parting thought is that the A7s is neither the holy grail of cameras, nor a whimsical outlier. It’s going to take a special photographer with keen image perception to appreciate what the A7s has to offer. Thanks Sony, for catering to more than the lowest common denominator.
Sony A7s preorder in USA/CA at Amazon, Adorama, BHphoto, Sony USA, Sony Canada.
Sony A7s EU pages: Sony DE, UK, ES, FR, IT, NL, BE, AT, CH, SE, NO, FI, PL. And Wex UK.
RX100M3 preorders in USA at Amazon, Adorama, BHphoto, Sony USA.
RX100M3 preorders in Eu at Sony DE, UK, ES, FR, IT, NL, BE, AT, CH, SE, NO, FI, PT. And Wex UK,
RX100M3 in Asia at Amazon Japan and Sony Australia.
Sony A77 body in USA at Amazon (Click here), Adorama (Click here), Sony Store (Click here) and BHphoto (Click here).
Sony A77 with 16-50mm lens at Amazon (Click here), Adorama (Click here), Sony Store (Click here) and BHphoto (Click here).
Sony A77 body in Europe at Sony Germany, Sony UK, Sony France, Sony IT, ES, NL, BE, AT, CH, PT, NO, SE, FI and Wex UK.
Sony A77 in Asia at Digitalrev and Sony Australia.
0 Responses
“produces some of the cleanest, most pleasant images I’ve seen”
what ? you have only phones?
Keep on trolling on dude… You obviously have nothing else going on in your life…
ok 🙂
That was a very well written article. I was shocked that the silent shutter couldn’t be custom assigned, I hadn’t seen that in other reviews. Hopefully in the final firmware it will be different but it’s that sort of feature Sony that usually adds in the first firmware update.
Often Sony will allow the C buttons to have assignments the Fn button grid is not. A good example would be Eye AF or Application list. You are not allowed to set them to the Fn grid but you may certainly set them to C1, C2 or C3. I wonder if the Electronic/Silent Shutter will be treated in a similar manner….
i wonder how accurate his FE lens prediction roadmap is 🙂 Couple of really exciting options listed there…hope they’re true.
Its not.
Its according to the fake roadmap, he is wrong
http://photorumors.com/2014/01/27/updated-sony-fe-lens-roadmap/
The 85mm cz is unlikely to have oss IMHO, given Zeiss’ position on oss in primes. I’d expect the macro to be oss (otherwise it’s going to have limited appeal). The 35/2 G oss is a new one for me – can’t recall seeing any rumours about G primes. Couple that with a 50/1.8 G oss and I’d seriously think about buying an a7…
It’s quite obviously wrong. Too many incosistencies.
Looking at the photos of the books I am not seeing 3 stops better noise performance. In the article the author said it was more like 2.5 stops but it looks more like 1 – 1.5 stops to me.
Did I miss something?
Nope – I’m with you. About 1.5 stops I’d say.
Interesting, but I don’t see where he gets three stops advantage. To my eye, the A7S ISO 51200 image is considerable worse than the A7R at ISO12800. Look at the brown veneer of the shelf for example.
I think when the measurements come out, at equal output size, we’ll see about a 1.5 stop improvement at higher ISOs on the A7S, which is still pretty impressive. It will move it ahead of the Nikon 4.
nb. the article seems to confuse electronic first curtain with a fully electronic shutter. The A7 (and some other Sony cameras) have an electronic front (first) curtain, but that doesn’t make it fully electronic. That needs an electronic “back” (second) curtain which reads the last line of the virtual “exposed slit”.
Am I right in thinking the a7 also has EFCS? The review implies it doesn’t…
You are correct. The author has confused EFCS with a fully electronic shutter. The EFCS only resets the photosites (and is available on some other Sony system cameras). A fully electronic shutter (of this type) also requires an electronic back curtain that reads the last line of the “virtual exposed slit”. On earlier Sony cameras, the read rate just wasn’t fast enough to allow an acceptable traversal time. (The EFCS has much less work to do – it’s “merely” dumping the accumulate photosite charge, and not having to go through the whole exercise of reading it and going through the ADC exercise).
Thanks – very helpful!
However you measure the noise advantage over the 7r, its there. This disproves the oft repeated theory that the 7r’s 36 megapixels downsized to 12 will be as noisy as the 7s
It depends what your measuring. The 18% Grey SNR numbers will come out very similar. That’s because noise in the “mid” areas is dominated by photon shot noise, and the mathematics is such that it averages out.
Dynamic range at higher ISOs is a different matter. In the dark areas, it’s not photon shot noise which dominates, it’s read noise from the sensor. The latter (for similar technology) tends to be fixed so the noise from multiple photosites adds so the SNR in dark areas of the images does not improve by mathematically combining the output of multiple photosites.
The upshot is, dynamic range at higher ISOs will tend to favour the larger photosites, but it equals out in mid-range. At low ISOs, there’s precious little difference.
Very useful. Thank you
I assume photon shot noise, being ambient and random, is unavoidable, whereas the sensor read noise has decreased and will likely cobtinue to decreaae with better electronics. Correct? Or can photon shot noise somehow be filtered out?
That’s a reasonable summary. Photon shot noise is simply the randomness in the arrival of photons and is inherent to the nature of light and is inescapable. It basically goes as to the square root of the arrival rate. So if you receive 4 x the number of photons, the shot noise doubles whilst the signal strength quadruples. That means, overall, the signal-to-noise ratio doubles with a quadrupling of photon count. It doesn’t matter if that’s one single photosensor, or four smaller ones combined. It’s a bit more complex with bayer-pattern sensors, but it’s a reasonably accurate summary.
Read noise is, as you say, a function of the quality of the electronics and, for a given technology, very similar for different size pixels. So add up multiple sites, and you add up that base noise so combining multiple readings doesn’t help that base noise level. It’s that base noise level which governs the bottom end of dynamic range.It can be improved, but even then there are limits. For instance, there’s something called thermionic noise which is to do with random behavior of electrons with temperature. To overcome that you have to cool the sensor, which is fin if you are an astronomer with a great big telescope and access to liquid nitrogen, but not much use in a hand-held camera.
There’s actually quite a lot more to noise than all that though. Modern cameras have to compensate for the differences in sensitivity of different photosensors, dark noise and so on. Much of that can be cancelled out by clever processing, but once those are done, the two critical principles are the ones above.
Mathematically combining multiple photo sites *does* reduce the contribution of read noise to total noise. Downsampling reduces noise, no matter the source. It just doesn’t reduce it enough to make it equivalent to using fewer and larger photo sites in the first place.
Simply speaking, not it doesn’t. It’s the SNR that matters.
First take the low intensity case where (to a first approximation) noise is a constant R and the signal is S. The SNR for each photosite is therefore S/R. Combine 4 sites, then the SNR is just 4S/4R which simplifies to S/R – that is unchanged (and so on for any number of sites).
In contrast, in the brighter areas where shot noise dominates, the noise is the square root of the signal which means, of course, that the SNR is the square root of the signal S; so SNR = sqrt(S). If we now combine
Of course, neither the (largely fixed) read noise or signal-variant shot noise completely disappears, it just becomes relatively less important as you get into dark and brighter areas. The above is simply a first approximation, which illustrates the extremes. A more complete mathematical analysis can be found in the following article which basically says that (for equal sized sensors and technology levels), those with larger photosites will have higher DR at low light levels, and that the smaller photosites could get a small advantage at low ISO.
I’ve just simplified it a bit to illustrate what happens at the extremes.
http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p3.html#pixelsize
Note the same article explains why, apparently, lossy RAW compression systems don’t actually lose image information (that’s assuming they are implemented properly). All interesting stuff for those interested in science and engineering but not, maybe, for those into actual photography.
Check your reference. Noise doesn’t add linearly. It adds in quadrature: N = sqrt[ (N1)^2+(N2)^2+(N3)^2+(N4)^2 ]
BTW, your reference is a really good read with lots of information. Especially interesting is the supplemental page talking about the difference between “sensor dynamic range” and the delivered
camera dynamic range
at any given ISO. The author theorizes about a future sensor with dual amplifiers and ADCs. Such a sensor now exists: http://www.scmos.com/downloads/ I wonder if Sony has done something similar in the a7S, or if they’ve found a different way to decrease downstream noises.
Thank-you to balazer and SteveGJ for this discussion.
Hey Barry, are you seeing this?
I agree, 2-3 stops, maybe closer to 2, but depends where you are in the range.
May I know what’s the filesize (resolution) like when LA-EA4 is mounted with APS-C lenses? Thanks.
12,2/2,25=5,42MP
Thanks, it’d be really “tiny” then… ^_^” If I get this camera for body only, I’ll not have the budget for the lenses. Was thinking to get a LA-EA4 adapter first and pairing with my APS-C lenses for now.
Sell your apsc lenses, dont buy the ea4, and get some vintage manual lenses instead! You will come out with extra cash
Plus one. And they don’t even need to be vintage 🙂
@mediokre:disqus, In fact I was thinking to use a77m2 as my action photography works and the a7s for low light wedding assignment. Manual lenses is not an option. 😉
I’d make that a6000 and a7s, personally!
wow! received such high review from LM.
Now i wonder if the 7s will work better with wide rf lenses. Bigger sensor sites have seem to lead to less edge smearing and purple, as in nex 7 vs nex 5n. Any ideas?
yes, it will
It will have issues. Although, one should better define what ‘issues’ means.
Why are some people so surprised with the A7s performance against the A7r?
The A7s simply has the sensor numbers in its favor for better low light performance. Sensor size and lower photosite density tell most of the story before you ever look at any comparisons.
Seriously,,…This should not surprise anybody.
Quite a few people had been claiming that images from a7r downsized to 12mp would have identical noise to those from a7s. Now we knoq thats not the case at high iso.
No one claimed it would be identical. Find those posts and show… otherwise you’re lying.
If the sensors were from the same generation, the performance would’ve quite probably been significantly more similar.
Because the assumption that bigger photosites are always better at high ISO than camera with higher pixel count and thus smaller photosites is not always true.
Just compare the A7 and the A7r, once resized to 24mpix the A7r has a slight advantage and sharper images than the A7. Even the 16mpix D4S doesn’t show a great advantage agaisnt the D800e, DxOMark measure low light performance by comparing prints at 8mpix once resized the D800e has a nearly identical performance.
It reminds me the NEX 7 vs NEX 5N / NEX 6 so many reviewers compared the 100% crops which is neither fair nor reliable, I owned both NEX 5N and NEX 7 and of course at 100% crops, the 5N was better but when the nex 7 was resized to 16mpix, I couldn’t really tell which was better at 6400 but the NEX 7 was sharper.
So that’s really nice to see the A7s performed so well compared to the A7r.
A7 has on sensor PDAF which eats up some performance. D610 v D800 D610 is a little better (but not that much). It is true, to a point what you are saying though, but the A7S definitely performs better.
How many PDAF pixels does it have? It’s not significant enough to affect A7’s SNR.
It’s not just the pixels themselves but the additional circuitry involved in making them work. Just look at A7R vs D800e, almost identical, then look at A7 and A99 vs D600/610. A7 loses out because of PDAF and A99 loses double due to PDAF and SLT.
There’s no additional circuitry on the sensor for on-sensor PDAF. Have you looked in to how it works?
Differences in a7r and d800e… first, according to who? DxOMark? Same people who say there’s a difference in noise between the D800 and the D800E as well? 😉
That ‘difference’ is well within sample tolerances. Sorry but show me one instance where it hasn’t had an impact in noise performance?
If you understand how it’s implemented you’d know how it doesn’t affect noise. It’s no different to asking to prove how go fast stripes don’t affect the performance of a car 🙂
SONY Pro broadcast guys publish the actual signal to noise measurements of many of their cameras. Example: Sony X camera has 59db signal to noise response.
I’d love to see Sony publish these numbers for Alpha cameras.
Are you sure PDAF eats up some performance ? Because the A6000 has so many Phase Detect focusing points, much more than the A7, and yet it outperformed all other NEXs by about 30% in the iso test of DxO.
And concerning the A7S, yeah of course it does perform great and that’s a very nice thing, but I just said, that bigger photosite = better performance isn’t always true, I did hoped it was true for the A7S, and that it would outperform any other FF camera at high ISO and it seems my hope came true.
How a7 vs a7s noise?
a7 is worse then a7r downsampled, so the scale goes: a7s>a7r>a7
Thanks. Is the 7 noisier than the 7r because of the antialiasing filter? If not why does it violate the bigger sensor site less noise principle?
AA filter has no impact. The A7r just has a better efficiency, in part because of tge gapless microlenses used. Technology used has a greater impact than size of pixels.
PDAF on sensor.
Yes, in part because of that too.
“bigger sensor sites less noise” is something that noobs (e.g. dpreview review team) say. At the end of the day, it’s the area that matters most:
https://www.flickr.com/photos/genotypewriter/3012247230/
DxO also talked about this:
http://www.dxomark.com/Reviews/More-pixels-offset-noise
That was done in their more sensible days…
But you are arguing again the very people that made the A7s. They are very clear about why it works the way it does.
No… you must have read in to it too much or have swallowed marketing hype. Sony didn’t invent physics 😉
A7 has worse noise performance than A7r when compared at same size. Thus the difference between A7 and A7s is even more than that of A7s vs A7r
Of the many people that are going to write up about the A7s, this guy has laid out a very fair and balanced opinion without delving into a supposed analytical take with a biased result.
That’s probably because he is one of those who take photos. Lame.
One of the reason i want a 7s is the possebility to take pictures in the night at iso 12800, and sometimes iso 25600 handheld. So if this article is correct. Then i would love to break the bank for the A7s
I really have to question their testing methodology here. There’s no way a properly down sampled A7R image should be out resolved by the A7s at ISO 100. I’m not seeing a 3 stop advantage over ISO6400 either. 1 stop at best.
One stop is more realistic, 2-3 stops would be absolutely crazy. Even one stop should mean that it will be the first camera to break the ISO 4000 barrier for the low light score in DXOMark
I agree 1 stop is more like it , the fanboys will be reading their attack as I type for such heresy . To be honest I wonder how the Sony 12mp sensor from the D3s would do in a camera with a m,odern processing engine.
The D3S does not have a Sony produced sensor and processing isn’t the real issue with that camera compared to newer cameras (including the D4), it’s the shadow noise (and colour integrity of shadows) at the highest ISO’s that somewhat falls behind.
This sensor is a remake of that sensor, there are some 45000 pixels less in readout on it, the high side of the image is exactly the same, the wide side is a bit less, but not much. The pitch is the same, the pixel count all over matches too. D700/D3 sensors have 12.2 mpix and this one 12.1. Even is some fanboys will now scream once more, this is the same type of sensor and what it does with actual technology and in camera software, is what we see on A7S shots. But, you should know that the D700/3/3S sensor was made by Toshiba, the design was by Sony.
Nonsense, the D3S has a Nikon sensor that is likely fabbed by Renesas with a markedly different layout than any Exmor sensor from Sony, the A7S included. The huge difference in read noise curves will be the real telling tale.
So a sensor with roughly the same pixel count has approximately the same number of pixels in each dimension is the basis for your claim this is just a rework for the D3 sensor? Hardly definitive. It’s pretty clear that Sony designed the sensor to be a good fit with 2K & 4K video. Apply a 16:9 crop add in a few extra pixels to allow for cinematic 4K and the potential for electronic stabilisation (in video) and 12MP comes out as a sensible size. The pixel pitch follows automatically two 12MP sensors of the same format size will automatically have similar characteristics.
In technical capability, the A7S sensor clearly has capabilities that the D3/D4 sensor does not. The ability to do full width readouts at video frequency for one (and the consequent electronic shutter).
The D3/D4 sensor was not designed with video in mind. It’s optimised for high frame rate action photography. The two sensors will, because of some characteristics in common, have some similar properties, but one is clearly not just a rework of the other.
I do agree with you, it’s strange that the A7s seems so much sharper than the 36mpix downsampled to 12mpix… I thought this could be due to the shutter vibration issue of the A7r.
And I don’t see 3 stops of improvement neither 2.5 for me the 6400iso image from the A7r doesn’t looks worse than the 12800iso A7s image… However, he may have took into account, for his 2-3 stop advantage estimation, of the A7r red shift which at 6400iso is more noticeable than the A7s even at 25600iso.
I see about a stop, too. If people don’t require more than 12MP that’s a good camera.
I’m not sure what most people here do with their photos that require such huge blow ups… I’ve gone back and looked at old photo I had from back when I was using a 6MP camera and I know I had shot blown up that were 16 x 20 and perfectly useable… Unless you are print billboards what in gods name does anyone really need the current 36MP cameras for in the first place? If its just to be able to crop then get the right lens and photograph your subject the right way to begin with and you don’t need to crop.
The point is that you’ve lost the ability to do any cropping for a measly 1 stop gain in low light performance, is that really a smart tradeoff? Most would say no, especially since down-sampling a 36MP image provides the same level of noise reduction as the 12MP file at the same exposure. At lower ISO you still have a clean file with 3 times the resolution.
“If its just to be able to crop then get the right lens and photograph
your subject the right way to begin with and you don’t need to crop.”
I’m sure you shoot nothing but JPEG because exposure, framing and color treatment are always perfectly managed by your little robotic brain. Awesome job telling the entire photography community what they’re doing wrong, Mr. Hindsight 20/20.
I don’t really crop, not saying others don’t need to but I don’t usually. Moving costs me nothing, getting a stop faster lens cost me big $$$’s if they are even available.
Sometimes moving can cost you, while cropping will be free.
The usual presumptuous thinking. Who requires blowups; if you need to crop, then your skills are lacking. Go use your 6MP camera if you are so happy with it, others will benefit from the advantages of higher MP camera.
I have to agree, this is only a 1 stop improvement at best… not a very good compromise for such a huge drop in overall resolution.
I didn’t like this test either, on the web it wasn’t very illustrative. But given Mike’s extensive real world experience and reputation for objectivity I wouldn’t be so fast to dismiss it (his opinion/conclusion). These web-tests and analysis are often exercises in futility. The interesting thing was his observations about the A7s file quality, that’s something that goes against most testing methodology but its something that 5D owners noticed when comparing the original 12.8MP images against the newer larger 5D3 files.
please reduce A7s price by 2 stop please 🙂
that will not change anything to the situation, only to the price.
that’s the point, he just wishes it were cheaper…
Look at the book with the title Margaret in ISO 100. It’s all muffled in the a7r and the a7s it’s much more clear. Even the book desktop. Either the down scaling changed something, wasn’t in focus or a setting was changed. Reason I’m bring that up at ISO 100 with no filter shouldn’t the a7r beat it at that?
I do not think that this test was axed on clean focusing but more on noise comparison, and since it was in dim light, the ability to see in the dark. Over 25600 you can then claim that A7S still sees something, but, is something what you expect from a stills camera with 405k ISO ability. At 25600 it is already something. The question is how far we can go with good noise rework software. And the next question is the ability to shoot higher than 25600 ISO that much useful, except the fact you can pretend to have a camera who can do that. For a sports reporter that shoots pictures you see in reduced size in a newspaper or net page, this is a bargain without any doubt. Would that one then not go for a D4, one more question to ask.
I think focusing is important in noise comparison. If one part of the image is out of focus it messes up the noise comparison. Since we look for when the image becomes too degraded. If we start with non-focus image, then we focus on “noise degradation” then the whole setup is messed up. Just feel like it needs to be redone.
The image test confirms exactly what I claimed. If you look at the color tablet, you will see that bot camera produce noise at 6400 ISO. A7S get a stop better in all ranges, and up to 12500 this stop is visible. Now, look at 25600 iso, here we are over the limit what is useable in still shots, the brown edge of the tablet is stitched with all colors of the rainbow, and the visible noise is quiet rough. Now, for video all that is ok, but as I claim since weeks and for what I have been insulted by some pantpissers, for stills, this camera has absolutely no advantage, and to pay 500 bucks more for a stop better is nonsense. If Reichmann finds 3 stops anywhere, he needs serious glasses. This is a video camera and remains one that can do stills, but without any major advantage on A7 or A7S.
Even if it’s a stop difference, that’s also about the difference between a typical FF and APS-C camera from a similar generation. Which puts your “500 bucks” and “nonsense” nicely in perspective.
Dont forget potentially better support with rf lenses. Thats huge for m mount lens users. Imagine a great still and video camera with leica lenses. Only option right now. No need to suffer gh4 sensor crop.
With your track record compared to his, guess what opinion actually counts?
I have tested the a7s alongside with my A7 in Hong Kong, a7s ISO 1600 look just like the a7 ISO 200 photo
Really? Because $500 rarely buys you a stop faster lens, single stops cost a lot believe it or not. And then you have to buy a stop faster lens for each and every lens… now THAT sounds expensive.
Great article! Pity, that noise comparison files are not downloadable.
The 2+ stops difference are pixel per pixel, once the resolution is equalized, until ISO 12800 I see no difference form A7r to A7s.
That’s what i also would expect. The area of the sensor is the same, only the wiring overhead for the higher pixel density will create some light efficiency reduction. But that can not be very big. I think the main advantage is the low-light performance in Video because the A7s is using the full sensor area where the other cameras are doing sub-sampling
I totally agree on video performance. Such a shame it does not record 4K internally though.
Even if the actual area of the photosites is less in the 36MP sensor, the microlenses are there to bring up the fillfactor.
How are they any better than the ones on the A7S?
Not saying better. Larger photosites, even when compared 1 at a time aren’t going to be proportionally better because microlenses will benefit smaller photosites better.
Well you didnt look very well then I guess, A7R was down sampled and still looks quite a bit worse.
@abortabort
Under a magnifying glass perhaps.
Ok. If you prefer the A7R output stick with that, it needn’t be complicated.
Ignore the trolls, do not waste your time for them.
The A7s is not doing so much better, I guess not even half a stop. If it was 2 stops better it would get a score of 10000 and more at DxOMark.
I thought it might score a little better, but after playing with them I think it will score a fair bit better than A7R. However scores are not everything as they measure performance at a particular point, how both roll off after that point can be quite different. DR for example mightn’t be ‘much’ better at max DR, but it will be quite a bit better at higher ISOs.
Perceptive differences are never the same as statistical differences. I believe DXO tests SNR solely by light sensitivity, not the actual quality of the noise. Plus, we don’t have data yet anyway.
so he is not able to see a difference in dynamic range!? thats strange because looking at the first bookshelf shots (iso 100) the shadows in the a7s file do obviously have more detail and tones than the one from the a7r…
Tempting body.
A poorly executed test… the a7r images are softer than the a7s images ROFL. This is the first thing Reichmann should have commented on.
And what rubbish is he saying about the a7s being 2.5 stops better? 1 stop tops… even that’s not conclusive because the first problem (mentioned above) isn’t solved.
Can’t remember the last time I came across a technically sound experiment from Luminous Landscapes…
Where are yours then?
The a7s is not even heard of in Australia. Once it’s launched here (for which there’s no public date as of now btw), I’ll do some tests.
Until then I did some simulations of how 12MP sensor images compare to 12MP images from a 36MP sensor… the difference is shocking.
Im in Australia, I’ve shot with one. Public date is currently July 18. What sort of simulations? Of course your simulation ones aren’t going to be great, just like in the article where the A7R is simulated to 12MP to more closely compare. If you have simulated it you have done more or less the same. Which just proves that higher MP down sampling won’t produce the same results.
Well then it must be a Melbourne thing. I even spoke to the main Sony rep here a week or two ago and he had no clue. Doubt he even knew much about the a7s.
And a simulation is different from what Reichmann has done. My simulations are completely computational. A very high resolution (~66MP true RGB) synthetic image’s capture was simulated using a virtual 36MP Bayer sensor and a virtual 12MP Bayer sensor. It assumes everything being perfect. While there are lots of real-world factors that come in to play with real cameras, such an isolated simulation gives a good idea of fundamental resolution relationships between two such sensors.
Fair enough, seems Mr advanced than what I could be bothered doing. For my uses the A7S is a huge improvement, will be ordering very soon. The other advantages play a big part, it isn’t just noise v resolution.
You can read some of my findings here:
http://www.dpreview.com/forums/thread/3685321
such an isolated simulation gives a good idea of fundamental resolution relationships between two such sensors.
How? I’m not so sure it gives a good idea at all. When you take a picture with a camera usually you are taking a picture of something in the real world, not in a raster based world. I don’t see how a running a simulation of how different sensor sizes capture a raster-based image not even twice as large as the the larger sensor gives any sort useable information. You wasted your time with that simulation. Do you know the geometry of Sony’s gapless micro lens array? pretty sure that would have an effect on your simulation. I’d be interested in the same test but with a vector based sample file.
By the sounds of things you’re talking about a subject that you don’t know. So I’m really curious as to how you think it’s a pointless exercise.
All tests have limitations and that’s fine as long as the limitations are clearly stated. Believe it or not even DxOMark tests have limitations and although they are stated very vaguely, people like you take them as gospel.
Woah! pretty sure the a7s is not using Bayer, according to Amazon, “a new-generation RGB color filter has been adopted”
How has this not been posted on SAR!?!
When the RX1 was launched even some Sony sites said it had stabilisation. Don’t trust everything written by marketing people.
Buy it you will love it, it will be out on this Friday in your area.
+1
Even if it is only 1-2 stops better the other benefits of the A7S are unmistakable. If you need resolution sure, but otherwise A7S is a better rounded product that either of the other two.
You can’t say “if it weren’t for ____ it’s a better product”. Personally, I think the a7r is a better rounded product… it has excellent noise levels, the highest resolution and isn’t ridiculously expensive. People who need to have a slight edge in noise at the expense of resolution will like the a7s. Different strokes and all…
But the a7… now I really don’t know the purpose of that camera. It’s not the best at any functionality…
Well you can’t say the same then for the A7R if you know what I mean 🙂 I have no issue with the resolution, but I know some do, luckily for them there is the A7R. You are right, different strokes for different folks, don’t know why Sony users seem to feel the need to poo-poo others requirements because they don’t fit them. Other the resolution)(which is the major differentiating factor between the two) I stand by my statement that A7S is a better rounded camera.
Yes the a7s has the video thing going for it too… being a geek, I need raw video more than anything so I see the a7s as a very special purpose, very high ISO, stills camera. Personally, with stabilised lenses, I can shoot at ISO100 on the a7r at night… sometimes stopped down. So the a7s won’t let me do anything I can’t do right now, conceptually.
The fully electronic shutter… might help with vibrations when using long lenses on a tripod but a lot of the issues the a7r has will be minimized if you resize them down to 12MP to start with. There are other ways to reduce vibrations too.
These are the advantages as I see them:
Faster AF in medium to poor lighting
Better build
Faster bursts
Longer burst depth
EFCS
Fully electronic shutter
Better video
Option of better video still
Cleaner VF in medium to poor lighting
Better DR across the ISO range
Cleaner output across the ISO range
S-Log2
Comes with 2 batteries
Comes with a USB charger and battery charger
Considering some of the main complaints about the A7 series are all improved on the A7S, then yes, I think it is a better ’rounded’ camera. If resolution is more important than all these things combined then yes, the A7R is a better choice.
Faster AF in bad lighting: It’s not a camera made for AF. Also I MF faster and more accurately than people AF.
Better build: Inconsequential… I regularly use heavy lenses (~3Kg) on my a7r and there are no issues.
Faster and longer bursts: Not relevant to people with good timing. I’ve used cameras and sold them without ever putting them in the continuous mode.
EFCS and silent shutter: See my earlier reply.
Better video and S-log2: For me raw in video is just as important as raw in stills.
Better stills from video: I don’t extract 2MP stills from compressed videos… never did.
Cleaner VF in bad lighting: Same as better build… inconsequential. Not going to affect photos. I regularly shoot in very bad lighting with my a7r.
Better DR: Even if the a7s has 3x the DR over the a7r, 3x resolution is more noticeable.
Cleaner ISOs: With or without comparing to the a7s, the a7r noise levels are excellent. Even at night, I haven’t had the need to max out the a7r’s ISOs… even with mirror lenses!
Batteries and charger: Really? This is a concern for you? 😀
Good salient points from both of you. This is where Sony are very smart, we’ll be buying one of each………….
I can see one of my A7R’s going the way of eBay for an A7s!
Yikes! You got two!
It can focus in extreme lowlight , try it yourself for your own needs in terrible light events you will love it!
Yes I’ll definitely try it. I use stabilised lenses and I hardly go in to high ISOs even in the middle of the night. It seems until you’re in to around ISO12800, the extra resolution of the a7r will show over the a7s. For people printing relatively small it’s no issue. So the a7s will have a place for event and wedding photography if someone’s buying a sony FF mirrorless for the first time. It just doesn’t seem a clear upgrade if you already have an a7r.
I’d like to add that A7s level sensitivity would be a godsend for when I’m shooting concerts and environmental portraits. It’s not that I need cleaner OOC appearance, it’s that I need a lot of headroom for editing. Shooting above ISO 2000 on the A7 simply doesn’t give me the flexibility I need with my files–OOC everything is fine up to around 6400, but noise pops out readily if you’re doing serious and often essential adjustments.
I’m not a fan of the sensor in the 24MP FF sensor that Sony has put in to some cameras like the A7. The A7R images, when properly handled can be just as good, if not better for things like noise and DR. If you’re coming from an a7r, I don’t think the a7s will make new things possible just because you can shoot a stop higher in the ISOs. It is already really good and for me the benefit of being able to shoot at 3X the resolution far outweighs the 1 stop of sensitivity that I see beyond ISO12800 when the images are kept side by side.
With things like events, you always have bright spot lights and they always get blown out if you up the ISOs. So you’ll be shooting at reasonable levels of ISOs and at those times there’s no big difference between the a7r and the a7s in noise but the former has the benefit of having extra resolution on tap if necessary. Of course, if you’re shooting heavily stopped down and you’re above ISO12800, you’ll be better of with an a7s.
Right now the A7S has two things going for it: (1) less noise (worth the resolution compromise) above ISO12800 and (2) the silent shutter. The A7R replacement will most definitely have some option for a quieter operation. Its noise levels won’t also be worse than what the A7R is capable of now. When that happens, what’s the a7s good for if you’re stills photographer? The next A7 might do 4K internally, although at a lower quality than what’s possible with the a7s.
Good for you, you just explained why the A7R is better for ‘you’. Well done, now you don’t need to do anything, you already have everything you need, heck why even look at the review? There was no way it was ever going to be better than the A7R for ‘you’. Why don’t you go take the time you have wasted here defending your purchase and actually go and shoot your perfect camera? 🙂
No. It’s only after this test it became indicative that the a7s isn’t worth the 3x compromise in resolution for inconsequential gains in lower noise.
Also there are people spreading wrong info… I try to point things like that out for the sake of others.
That’s not point out ‘wrong info’ saying it is inconsequential to you does not make it factually incorrect. You didn’t need these tests to ‘prove’ anything, you had already made your mind up.
Couldn’t be more wrong about having made up my mind.
Also I didn’t say everything I say is for fixing errors in others’ comments.
“Faster and longer bursts: Not relevant to people with good timing.” Really?! you can’t be serious. If that was true why would pros use the D4s or the 1Dx over the D800e. Why do those systems exist if burst is nbd? I mean they are pros, if anyone has “good timing” they do. There are lots of shots that only exist because of burst. Some of the best shots in my portfolio are the collaterals taken when I bursted 3-5 shots off.
“I MF faster and more accurately than people AF.” Sorry bro, thats total bullshit, pros use AF because its faster and more accurate than a human can be. I see no way you can focus faster than my 17-40L; it never takes more than a 1/4 sec and has yet to be anything but tack sharp and on subject in 10k shots. A human couldn’t turn the focus ring to the right position as fast even if they didn’t have to think about nailing the focus, and that lens Nails the focus. I’ve seen veteran photogs use MF and they aren’t that fast. Pros use Autofocus. fact. thats why Canon dominates the Pro scene, and why they dominate best photo competitions, because they beat Nikon to fast, reliable autofocus.
This camera has significantly greater than 1.5 stops of ISO (I peg it a bit over 2) and has ~3 more stops of DR at any given ISO over 800 than the a7r. If you looked at and played with the .dng’s abortabort posted on dpreview it’s pretty clear that it’s at least twice the performer of the A7r in low light, if you care about sharpness. If you don’t mind a dark or blurry mess, go shoot concerts with the A7r. Sure you can get similar images with the A7r and similar settings, but the A7s will out resolve it pretty quickly as these tests show, and then the added DR you get with the A7s is just gravy.
You claim that you never push the ISO to 25k even shooting at night, and I’d love to verify that, but you seem to have rinsed your flickr of EXIF data, which really makes me wonder if you are legit at all.
You act like the only reason Sony incorporated Silent Shutter is to reduce vibrations. I guess if you are stuck shooting on a tripod that would make sense, I only ever shoot handheld and as such I am much more interested in being able to shoot streets at night without pissing everyone off and making them think that I’m Paparazzi. Or shoot during a concert recital, or in a Museum, or around children or or or. So many places and subjects get antsy when you start snapping off stills with an SLR. Maybe people down under just don’t give a damn about photographers, but here in California it is certainly something to think about. I’m super stoked to be able to shoot noiselessly, something I’ve been loving about my iPhone for years.
” If that was true why would pros use the D4s or the 1Dx over the D800e”
Because they’re pros. They all have to fit in a steamy small pit to save their lives. It’s actually no different to a clown car. They machine gun repetitive action for hours and get a handful of shots at best. It’s not a matter of skills.
“I mean they are pros, if anyone has “good timing” they do.”
Couldn’t be more wrong. A lot of photography writers are commercial pros… they naturally glorify themselves just like characters of writers are always glorified in books and movies. There’s nothing to admire about anyone with a 24-70 on one shoulder and a 70-200 on the other, machine gunning every thing they see.
“… it never takes more than a 1/4 sec and has yet to be anything but tack sharp and on subject in 10k shots.”
But I don’t take 10K shots. If I was spraying and praying, yes, I’d use AF too. For that you need no skill. What’s a 17-40L anyway? Who uses junk like that? Pros who don’t know how to use a real lens? ROFL
“A human couldn’t turn the focus ring to the right position as fast even if they didn’t have to think about nailing the focus, and that lens Nails the focus.”
If you’ve ever shot in MF a lot and if you’re good at it, you’d know that not all MF tasks are about turning the focus ring from CFD to infinity and back.
” If you looked at and played with the .dng’s”
Anyone can play with files… but hardly anyone knows how to compare them properly. Even Reichmann has screwed things up in this regard but he had the right idea by down-ressing the a7r images. I bet you also believe DxOMark.
“You claim that you never push the ISO to 25k even shooting at night…”
Yes… do these look like high ISO shots?
https://www.flickr.com/search/?tags=canon35mmf2isstabilisationtest
Or have I faked the night? lol
“I’d love to verify that, but you seem to have rinsed your flickr of EXIF data, which really makes me wonder if you are legit at all.”
Stripping the EXIF is for my protection as well. If you want the basic shooting details of any of the digital shots, ask and I can look things up and tell you. With the film shots, I might not remember exactly because I don’t take notes all the time.
“Or shoot during a concert recital, or in a Museum, or around children or or or. So many places and subjects get antsy when you start snapping off stills with an SLR. Maybe people down under just don’t give a damn about photographers, but here in California it is certainly something to think about. I’m super stoked to be able to shoot noiselessly, something I’ve been loving about my iPhone for years.”
Sure… didn’t say the a7s isn’t for you. I was only responding to abortabort’s comments on what the a7r can’t do.
Nah, I only put faith in what I see in my experiences with RAW files and what others produce with equipment.
The point of this thread is that A7s is a better all arounder than the a7r, which I stand by.
“If I was spraying and praying, yes, I’d use AF too.” lolz, burst is not synonymous with spray & pray, if you shot action you’d understand that. You are never going to time a humming bird, you have to burst. lots of fast subjects are just too fast. But you aren’t that type of photographer, and thats fine. What isn’t fine is that you have the opinion that you practice the only true art of photography, which is patently false. You are probably about as far from an artist as a photographer gets; and you proved it with that paragraph. While shooting burst may take no skill, neither does shooting LF technically, or any kind of shooting for that matter. Creating good images is an art that requires a keen eye for composition and if you want to be successful outside of landscape, very quick thinking as well. You shit on the 17-40L, but I’ll be out shooting long after your LF with wide angle canon glass has to be folded up and packed away due to rain. Moreover, that zoom will give me composition headroom that you don’t have on primes. At the end of the day I’ll walk away with a dozen salable images and you’ll be lucky to have one. And I do shoot vintage glass on ilford delta actually, so don’t act like I don’t know anything about manual focus, what I do know is that it’s BS that you can MF faster than a $5k canon prime can AF. That is straight deceit man.
“There’s nothing to admire about anyone with a 24-70 on one shoulder and a 70-200 on the other, machine gunning every thing they see. . . It’s not a matter of skills. (sic)” You’re just jealous that they get to do for a living what you spend all your money on as a hobby. In all honesty, I don’t know how often photographers are in a “hot steamy pit”, but I’ve never seen that. And they don’t usually “machine gun everything they see” Pro get paid to take photos because they are really fucking good. Don’t fool yourself otherwise (you will, you are totally full of yourself) When I see pros they are usually someplace I can’t get that has a much better view of the action. They get the best seats in the house, and Photojournalism takes more skill then you will ever have. A real, “many are called, few are chosen” profession. (unlike being a scientist I might add)
“A lot of photography writers are commercial pros…” Yeah, I don’t read what photographers write about. Could not care less, don’t tell me I’m biased.
I /talk/ to (retired) pros about shooting and experiences and read books on editing technique. All the professional photographers I’ve talked to and shot with save one were retired, and guess what? they were all phenomenal at their art, and didn’t tell me that pros were magic or anything like that. They have all told me about the same thing, focus on composition and on telling a story. They used zooms and AF and burst, and they produced more compelling images than anything I’ve seen on your flickr. Moreover, in retirement they didn’t go to shooting film with a stopwatch and a lens cap, they bought a nice little pocket cam and were happy that it was so much easier now. They didn’t obsess about resolution and catches and gear like you seem to. They focused on the photography, and composition and color and /feeling/. Your metaphor about writers and their characters sucks and doesn’t make sense, btw. Every single photojournalist I’ve talked to has been ashamed of themselves and happy they aren’t in the game anymore, but not because they weren’t having fun most of the time, but because they found it damaging to just be an observer and recorder and not ever lend a hand.
“Stripping the EXIF is for my protection as well.” hmmm? as well as what else? Are you certain you are a scientist? That’s pretty awful english skills you are displaying there, do you have someone else write your papers? and your protection from what? copy cats? ha! or from the real owners who might find you out if you posted their EXIF? or is it that you really just edit the shit out of your stills and the refuse to own up to it. Everyone I know who hides EXIF data from their flickr does it to make themselves look like a more skilled photographer than they are. I don’t believe anything a photographer tells me about a photograph unless the EXIF data backs it up, or I can see the film negative. I’m wayyyye to skilled with photoshop to be naive about what it can do (although I’m starting to get the impression you aren’t), and yes with patience I could create every shot you have their on a digital rebel, kit lens and photoshop cs6. Literally had this assignment in an editing class. 1/4, f2, hand held no retouching? Prove it. Why should I believe you? Serious question for yourself. If you don’t post EXIF data, how is anyone supposed to believe that you aren’t just fabricating/stealing these images? I know that I’ve seen my own work around with EXIF stripped, filter applied, claimed as someone else’s.
“Anyone can play with files… but hardly anyone knows how to compare them properly.” And I suppose you do? The one who can’t even make logical arguments or understand that, “Better video/
Option of better video still” means that you can get a 4k recorder and have better video still, not have better stills from video. You don’t ever identify Reichmans flaws, that’s for damn sure or how he could downsample better. Moreover you won’t even check the raw files.
Jesus fuck.
Great review!!!
Why not get both ? They both combined price is cheaper than the d4s alone. So get both r and s.
I just can’t agree with this. I was lured by the A7S and it’s overall appeal at first. However, it’s easy to forget just how good both the A7 and A7R do in terms of high ISO. There still some of the highest rated cameras in this regard. I think the ISO advantage on this thing is being overhyped for stills. Video capabilities- if you have the money- no question at all, it’s great. Factor is the cost though of the A7S and (better all around) just doesn’t hold up very well. For stills- sorry, it’s a pass at least for me.
Most of people buying this seven s already have a seven or r in their hands. I have both plus a99 and gh 4.
And what’s your thoughts on the A99 verses the A7R. I guess what I’m saying is- which one do you enjoy shooting more? I’m on the fence (own an A900).
Can we get a reputable source on that other than anecdotal evidence from gear enthusiasts?
Actually a D610 is better than a A7 or A7r in low light, high iso and a 6D is even better still.
Canikon trolls can’t take it more! And I don’t even blame them!!
With the triple attack of A7/R/S Sony will be the #1 camera company before the end of 2014. Those three cameras will make the competition tremble with fear. Add to that the A6000, the A77II, the RX series, and whatever Sony has in store for Photokina, and we have a massacre that will kill all other manufacturers, making Sony the only game in town. These are great times!
If I’m not mistaken some of those cameras are not full frame. They aren’t gonna kill anything. Sony must kill their crap sensor cameras before they can kill the competition.
Those crap sensor or as it is known to most of us as APSc sensor cameras, are what pulls most people into buying more expensive and better quality cameras. But then smaller sensor cameras also have their advantages that FF can’t possibly achieve, thats the camera and lens size and weight.
So while I understand your push for larger sensors, and I myself feel like its time to upgrade my A65 to a camera in FF A-mount, unlike you, i also see all the advantages that a smaller sensor may bring.
If sony kills all their small sensor cameras, and will not have anything to replace them with in the same price range, Sony are as good as dead.
Exactly! For macro and wildlife, a good apsc camera has very clear advantages. Not to mention APSC is the majority of their marked and the gateway to serious photography/penury. Fair enough, FF bodies are getting cheaper, but it’s nothing like APSC prices.
Yeah right everybody with his head in his shoulders figure that out
Troll really what are you talking about ?go back to Nikon canon leica fan club where you absolutely belong to.
I love sony but that won’t be happening anytime soon
#2 within 10 years is a reasonable goal. Survival and 20% share would be admirable.
+1
They’ll have to work hard on their E-Mount native lenses as well as fill in some gaps to A-Mount (85/1.8 comes to mind). They will also have to convert all their A-Mount lenses to use SSM or SAM. They would have to update optics in SAL35F14G to 21st century standards.
They would have to introduce the cheap’n’good lenses into their E-Mount line, similar to SAL35f18.
And most of all – They would have to fire their entire marketing department and hire some competent people, that can actually do their job properly.
Fire the Sony.com content management team as well while you’re at it. Those countless misstypes and incorrect feature lists on quite a few of those cameras scare away potential customers, when bad reviews show up due to a missing/incorrect feature.
Brand satisfaction index is very important, and they have to work hard to raise it.
All of the above, basically. Sony has a fantastic capacity to innovate, but it’s let down by lens selection/updating, marketing (seriously – WPA is a wasted opportunity every year), accessories and support. If they’re serious about being #2, they need to have a package that Canon/Nikon pros can’t say no to.
We can even use Nikon or canon lenses on the a seven bodies so no need Nikon canon fat ugly cameras that usually rejected by social meetings or formal events
No need native af lenses we just use some zeiss and leicas
Sony is already no two, just surpassed Nikon last year.
They will not even make a jelly tremble with fear with these three cameras. The A7 and A7r together are not selling enough compared to the main brands. The A7s will not make enough of a difference. If it sells as well as either of the other two A7’s it will be doing well.
Please elaborate why it is great and why being the only game in town will give benefits to the customers. I always thought competition is good. This is a typical fanboy comment without substance.
Having the option of buying an inferior product is not good. If the best product is the only choice, then everyone will be happy with their choice. And Sony makes the best cameras.
That is a bold claim. I disagree. They certainly make the best FF mirror-less and affordable cameras. But still there is a D800e,D610,… which are very good, too, even perform better according to many reviews, albeit at larger weight and bulk. The SLTs are a mixed bag, too. I would, however, agree, that Sony is maybe the best sensor manufacturer.
What I haven’t seen yet is what the ACTUAL linear resolution penalty of the A7s is vs. A7r with a high quality lens. You’re not really getting 40% more usable linear resolution with the A7r… but is it close? Is it really just a 20% penalty? Are the corners better in the A7s due to the larger photosites? THAT is the important set of questions for me.
More than 20 percent for sure,but less than 38percent inbetween that.
The biggest DISAPPOINTING BUAHAAAAAA BUAHAAAAA
`Based on these tests as well as my day-to-day shooting I do not see any
obvious difference in dynamic range between the two camera. Many people
had speculated that the larger pixels of the A7s would produce wider
dynamic range capability. Not that I can detect.`
The subject of the test wasn’t really optimised to test DR, but probably the best think to look at is how much detail you can make out in very dark backing of the bookshelf where there is the impression of wood grain. What will happen is that at higher ISOs there will be less detail visible in those very dark areas as noise appears and/or noise reduction wipes it out.
At lower ISOs there won’t be any visible difference in DR as the subject simply doesn’t have enough of it to show (especially the way the files will have been converted), but at higher ISOs it is there, but nothing like the 3 stops claimed – more like 1.5.
A better subject would have been one with a much more obvious native dynamic range than a bookcase, but that wasn’t really what was being shown.
And it seems that all of this system is gonna be ready end 2015 ,not now
16-35mm f/4 Zeiss OSS (2014) (announced May 1, 2014)
85mm f/1.8 Zeiss OSS (2014)
100mm f/2.8 Macro G (2014)
35mm f/1.4 Zeiss (2014)
24mm f/2 Zeiss (2014)
28-135mm f/4 G OSS (2015) (announced May 1, 2014)
135mm f/2 Zeiss (2015)
21mm f/2.8 Zeiss (2015)
35mm f/2 G OSS (2015)
50mm f/1.2 Zeiss (2015)
Troll go away go back to your micro land.
Comparing the a7s to medium format is stretching things a lot.
No matter how good each pixel is this is still only a 12mp image.
To be more precise it is only 6mp green, 3mp red and 3mp blue.
It is then interpolated from 4 different theoretical 3mp images to create a 12mp image…demosaicing.
None of the 3 color pixels created are the actual pixels seen by the sensor.
You can upres it or you can always stitch it. So the a 7s file upres might possibly be better than the seven R file,
Clearly you dont do either of those..as I often do or you would not have made that statement.
Takeing a set of multiple shots suitable for stitching on anything other than a stationary subject is extremely difficult an the percentage of keepers is very low compared to single shots.
Actually doing the stitching is a lot of time consuming work often requiring multiple attemps with none comming out right.
One recent photo I stitched took 3 days of work to stitch 13 exposures together, resulting in a 9′ 200mp print.
Imagine stiching 30 or 40 12 mp pix together instead of 13 24 mp pix.
Upscaling does not add detail to an image.
I often res up 9x and then use unsharp masking to simulate a high res picture but the detail is not in any way comparable to actually having the resolution.
When the resolution is actually needed those methods are no sustitute.
Lastly mf is far to difficult, heavy ,and expensive to use for applications such as wildlife photography..although I would imagine someone somewhere actually does it.
Because stitching, etc. are ONLY done for telephoto moving wildlife photography, right?
Barry STFU
animalsbybarry, the hero of his own life.
So since you dissagree with my claim that stitching a 12 mp image is no substitute for a higher res or mf camera…tell us about how you have done it or how you plan to do it or about your friends that are doing it.
If I am so blatently wrong I would rather know the right way to do it rather than the obviouse fact that I am just wrong.
What Mike is referring to has nothing to do with “the numbers” we’re back to an aesthetic observation once again. You’ve shot MF B&W and 35mm film and made RC whole-plate prints before? Well they kinda tiny prints (that both formats easily handle) and if we were to extrapolate even further where both film stocks exhibited zero grain and depth of focus was identical — well, that’s what Mike’s referring to. I’ve noticed it in files produced on my old 5D compared to the newer 2 & 3.
if you pushed the gain on the old 5D (12.8MP) up to around 400 ISO and graded judiciously the resulting files “could” look sublime. You have to be able to see it though (literally and figuratively)!
Reichmann is a detailed oriented guy. When he says he manual focused on both cameras, and repeated the tests, it’s unlikely that he would miss the fact that one camera is out of focus during a critical test shoot. Still, it’s impossible to believe that a downsized A7R image would be that much softer than the A7S using the exact same lens. Whatever the cause it’s not likely to be due to the 36MP vs 12 MP sensor. If the issue is not focusing errors then there must be another cause. Perhaps it’s shutter shock, more focus shift on the A7R, sharpening applied to A7S due to incorrect RAW software support, etc. I highly doubt it’s due to the sensor.
Regarding the headline grabbing “medium format look” some people use that to describe the Foveon sensor or some Leica lenses due to color rendering. However, the A7S and A7R should render colors identically. Unless Reichmann can tell us exactly what attribute he believes is “medium format” looking he’s probably experiencing some placebo effect.
And it may be something to do with some extra internal raw NR of the seven R.
Indeed. Sony loves to cook their RAWs. Cooking the RAW would explain both the sharpness and the less than expected noise discrepancy. Once the A7s is shipping I’m sure someone will run both RAW files through an agnostic RAW developer to see what’s up.
Sony has gone on record to say that there’s no change to the sensels other than making them bigger. No magic fairy dust was used. There shouldn’t be any magical increase in sharpness or “medium format look” that just pops up. It’s a simple potential increase of S/N which may theoretically lead to lower noise and better colors in the higher ISO’s relative to a downsized 36MP image.
There could be a whole number of things at play in his “MF” observations. We used to get into these debates back in the old film days when dealing with different “looks” achieved from different dilutions of the very same developer. Most of the time it didn’t amount to a hill of beans but sometimes it could be that magic bean that made all the difference in the world! It’s all good — and as a couple of guys on this very thread proved in a very articulate debate, both the cameras are the absolute best at what they do and we definitely need both (as Sony planned it:).
This is a great camera for adapting a-mount lenses via the SLT adapter. The sensitivity of the sensor renders the light loss nil. The high ISO abilities can compensate for the lack of IBIS.
Can’t wait to try it on one of my long-term projects where I often find myself in the most interesting company sitting around a campfire at midnight. Those two extra stops of sensitivity would make a huge difference there. I’m finding my A7R is maxed-out at 16000 ISO even for my grainy B&W conversions. Can’t believe that though 16000 ISO — remember shooting TMZ?
Sensitivity is amazing. It’s amazing when users complain that ISO 6400+ has too much noise. Such images could not exist a short time ago. I rarely shot film speed above 800, the max was certainly 1600 for me.
according of Luminous Landspace-
“While the ability exists to use an almost unlimited array of legacy lenses on A7 series cameras, the number of native FE mount Full-Frame lenses is still rather sparse. The following, according to Sansmirror.com is Sony’s lens roadmap from now (mid-2014) through end 2015.
16-35mm f/4 Zeiss OSS (2014) (announced May 1, 2014)
85mm f/1.8 Zeiss OSS (2014)
100mm f/2.8 Macro G (2014)
35mm f/1.4 Zeiss (2014)
24mm f/2 Zeiss (2014)
28-135mm f/4 G OSS (2015) (announced May 1, 2014)
135mm f/2 Zeiss (2015)
21mm f/2.8 Zeiss (2015)
35mm f/2 G OSS (2015)
50mm f/1.2 Zeiss (2015)
Add to this the already shipping…
Vario-Tessar T* FE 24-70mm F4 ZA OSS
Sonnar T* FE 55mm F1.8 ZA
FE 28-70mm F3.5-5.6 OSS (kit lens with the A7)
Sonnar T* FE 35mm F2.8 ZA
FE 70-200mm F4 G OSS
…and the lens future for the FE line looks very good”.
what a bright future of FE mount 😀 really can’t wait 😀
What’s the max fps in burst drive mode in the a7S compared to the a7R and a7? Still holding myself until Photokina.
as a side note, I’m waiting what Sony will offer for A-mount in Aug before I get myself again buying in E-mount territory (sold my 3 year old NEX-5 and a bunch of lenses a few months ago) or jump in RX-series.