kecajkerugo: it is interesting to see that Fuji X-T1 has less noise the Sony, even at ISO6400, unfortunately it is impossible to gather higher iso on the Fuji with the RAWs.
Despite slightly overstating its ISO's and lots of RAW smoothing, the X-T1 is still quite a bit noisier in low light. See here:http://i59.tinypic.com/2db12k5.png
Up to 2 stops difference in shadows.
Boerseuntjie: Wow A7s is amazing at ISO 25600 JPEG's
Unless you shot those RAW files side by side in a controlled environment with the exact same physical exposure, which I am sure you didn't, any conclusion would be worthless.
aandeg: Why does the panasonic have so much noise? The A7s raw at iso 3200 looks like the Gh4 at iso 400.
As you move further up the ISO scale, the difference climbs well past three stops in shadow noise due to much lower read noise and better efficiency per unit sensor area too.
Sdaniella: so, the Sony A7S has video with Full M/M ISO (ExpSim LV) control like FF EOS LV Canons do, but Pany GH4 does not... ok, got it. Oly video is stuck in Auto-ev0/autogain LV (perfect, though for P&S videographers)
wait ... auto-iso isn't available ... hm, ok ESLV is available, but adjustable, but P&S autogain (auto-ev0) mode isn't available ... ok, now, I got it ... mimics most digicam autogain/auto-ev0 video modes, which lack full M/M ISO control ... how ironic.
btw, if you want on-the-go compensation adjustment of iso in M mode with auto-ISO, it's no different from making on-the-go adjustments with M ISO directly. in both cases you are bypassing auto-ev(pick your value) altogether anyway
PS: read out and reset speed are connected through the same clock and in this case it's specifically the reset scanning time that directly influences the usability (and thus availability) of EFC.
Sony has at least 2 patents themselves describing EFC systems to cover their implementations. Second, read out speed has everything to do with electronic curtain (both first and second) and the larger the sensor and the more pixel lines/rows, the more crytical it becomes. Only the third iteration of the 36MP Exmor (Sony design directly based on the 16 MP APS-C sensor first found in the A580/A55) is capable of supplying an EFC. And it's also the one with the fastest read ouf of all 3 generations.
No, the A7R not having EFC has everything to do with read out speed, nothing with Nikon.
peevee1: DPR wrote: "Unlike the majority of cameras, the a7S offers full-sensor readout - its 1080p video is made up from all the pixels within a 16:9 crop of the sensor, then downsized to 1080 resolution."
I am pretty much sure it is not so - it is not 16:9 crop of the whole sensor (which would be 4240 x 2385), but 3840 x 2160 crop of the sensor, which then simply downscaled 4:1.
Also, can you check that no line skipping happens at 1080p30 and 1080p60? At least in the latter case (and MAYBE in 30p), I am pretty sure every second line is skipped.
Precision in the things like that matter - just parroting manufacturers' advertisements is not good enough.
It doesn't skip in 1080P30, as also specified in the official specs.
steelhead3: Not a Nikon 1 user, but I thought the review may be a little harsh...this camera has possibilities if it came with the 20 meg Sony sensor, which Nikon can do easily since Aptina is into automobile products now.
And btw, the NXMini uses the same Sony sensor, but applies horrible RAW NR (smearing) at higher ISO's.
And we can all compare the RAW files ourselves and notice the V3 performs worse, also at higher ISO's with a $900 lens. No amount of reliance on unknown or private sources will change the simple facts staring us in the eyes.
57even: Hmm, can't see many early adopters. Nikon, get a decent sensor, quick!
Better still, put all this excellent technology in an APSC body. You would probably sell millions.
The RAW files are not better than the Dpreview claims which they supported with samples shot in various light conditions (typical deep shadows in typical indoor light included, unlike Imaging Resource). Shot with a $900 lens too, so no excuses possible there either.
Relatively high shadow noise at all ISO's and an early colour shift in shadows at higher ISO's.
HowaboutRAW: Let's hope Google unhides the "limit background/crapware" feature that Google hid with the move from Android 4.1 to 4.2 and higher.
Plenty of free apps out there to tell you what's running in the background and what's loaded at startup. Including control over those apps. The feature you were referring to is for cached programs anyway, which has little relationship to CPU power being used. My cached apps rarely use more than 5% CPU, mostly well below 2%.
If 10 seconds of effort to unhide developer options is too much work....
And for recent phones high on (free) RAM, there are more downsides to limiting the standard amount of cached programs (such as much longer load times of frequently used apps and higher battery drain as a result) than benefits.
Stu 5: Why on earth compare it to the 5D MKIII when everyone knows the 6D is the better Canon camera at high iso of the two.
" So far, the A7s isn't an option to choose on the DPR Noise widget"
A7S in RAW vs the best sensors available, minus the 1DX.
igor_s: Simulation of higher ISOs by brightening in the post decreases the SNR, therefore, the comparison is not fair. However, the 7S has lower level of shadow noise, and therefore should do better at extra high ISOs (where the sensor is poorly illuminated). Perhaps natively by about 1 eV if in "boosted" ISO tests it wins by 2 eV.Unfortunately the above applies only to ISOs above 51200 (if you ever need it and satisfied by the quality). At lower and moderate ISOs the A7S loses to the a7R even in shadows. A specific ultra-low light camera.
Practically every camera does digital boosting of sensitivities at the highest ISO's. For just noise comparisons at extreme ISO's, there's little difference whether done in camera or in post processing.
Stephen123: There’s a pretty big Sony vs. Canon question in what color those arches bait up and left of center look to the human eye. Sony has them white and a bit blown out. Canon has them warm and detailed. If they are cream or yellow to the human eye then Sony is screwing up. If not, Canon is screwing up. It’s not like it’s a small difference.
You do realize that you're actually judging the Adobe calibration for preliminary support of the A7S?
jenbenn: What you forgot to mention is that the 5D has vastly superior dynamic range in all shots. The highlights in the sony raws are all blown out while the canon retains considerably more detail.
BTW I own a 5d III and a Sony A7 (non-r). The canon consistantly delivers less noise and better dynamic range at all isos above 400. The exception being at iso 100 and iso200 where the sony has less shadow noise if you need to lift the shadows dramatically.
To complete this: Any test site (hint: DXO) rating the canon sensor far below the sony sensor, should start using their cameras to take photos instead of performing absurd measurebating tests which do not translate into real life.
People still don't seem to understand that highlight headroom in RAW tells you little about DR, more about metering or how the manufacturer and RAW developers respectively decided to implement metering compared to middle gray plus saturation poit and the default gamma curve.
DR is mostly defined by the shadow range (see also the DXO explanations), highlight range is usually a choice of exposure (relative to the saturation point of the sensor).
onlooker: DXOMark's claimed 1.5 EV dynamic range advantage of A7S over 5DIII does not seem to be reflected in this test. If anything, it appears the other way around.
How did you determine the amount of DR?
SteB: I think that's pretty impressive. Even taking into account down-sizing the other images it's clearly the cleanest from any FF sensor at high ISOs. I think this might have some potential from a macro/close-up photography perspective. Light is always a challenge, and often you have to use flash to freeze motion, or handhold. The big problem with flash in daylight is fall off over distance and black backgrounds. This means the background all has to be close to the subject to prevent it, which restricts how you shoot. I think this type of high ISO performance is maybe getting to the level where with noise reduction software, it might allow a type of photograph previously impossible in field macro photography. We'll have to wait and see if it's there yet. It'd be a lot better if it was A7 price.
The A6000, A580, 5N and now the A7S are all among the highest or highest performing cameras in their respective classes at the time of release, when it comes to high ISO peformances. No talk about incomparable samples or imaginative comparisons will change that.
You still haven't shares apples vs apples comparisons between the D4S/Df and A7S.
Marty4650: So, I guess there really is an advantage to having a few nice big fat pixels, rather than having a whole lot more teeny tiny ones.....
" But I've shot with the D4s into deep shadows and the Nikon vastly out performs the A7s."
Weird claims if you can't back them up and haven't shot the A7S under similar circumstances. In other words, you have nothing to back it up. So far everything available contradicts your claims too. It's even weirder if we take into account that you often claim you don't believe in resizing (normalization) for comparison purposes, which puts both the Df and D4S at an even greater disadvantage.
samfan: Interesting. There's barely any difference between the 7S and the Df at ISO 102 400, especially at the same print size. OK, the Sony is better at ISO 204 800 but I wouldn't want to use 200k and 400k options anyway.
Now imagine saying the above in 2005, that would be fun eh?
In all seriousness, 7S at 400k looks like my old Canon A610 at ISO 400 or Nikon D100 at ISO 6400. So, 1000x more sensitivity than my first serious digicam. Not bad, not bad.
In the low light setup, the shadows from the A7S are cleaner at ISO 102k than those from the Df at ISO 51k. See the blue noise in darker areas, typical for low light.