peevee1: High ISO looks worse than Nikon P7700, but better than Fuji X10, which is commendable.Look at the globe for clear difference in readability.
RX100 is of course much better.
@Dimitri:"Now you see, that RX100 is loosing in low light completely unless you resort to 28mm and soft corners with RX100"
Seems like you missed the comparison above that shows the RX100 sensor has a benefit of about a stop and a half to begin with. So "losing completely" is typical Dimitri talk. Rather quite similar, with a clear benefit at the wide end for the RX100.
And the IS claim is just that, a claim.
Tom Goodman: Sample postings are useless. Only real-world experience on your computer and in print matters.
No problem, one can only take so much criticism when it's unwarranted... I'd think.
Simon97: The Nikon P7700 looks better with low ISO jpegs, but Canon looks better at 1600 (not sure why DPR uses 3200 to show default high ISO with these smaller sensor cameras). Raws are more even. I'd say look at the camera's other features and decide that way.
The RX100 doesn't really distance itself from these smaller sensor cameras here. It does have the resolution advantage but it makes softer images (especially off axis) and noise performance isn't much better. Of course resizing the RX100 image to 12mp will show its clear advantage but you pay the price in the wallet and zoom range.
@ bobbarber:That's a bit too simplistic. This is a close up scene where field curvature etc. hamper the RX100 lens the most. For more distant subjects, the lens performs much better, as also noted by Dpreview.
So towards the tele end, you gain up to half a stop in noise with the G15, but the less you zoom, the more you gain with the RX100, up to a stop and a half at the wide end. That's about the difference between a FF and APS-C camera from the same generation in terms of noise.
And let's not forget we're basically comparing coat pocket and jeans pocket cameras here, i.e., different size (and weight) class.
Instead of guessing, just resample the RX100 to 12 MP (or upsample the Canon to 20MP) and compare the output at ISO 3200 from the RX100 to ISO 1600 from the G15.Here's the result from ACR with NR turned off:http://i49.tinypic.com/1z3w6f5.jpghttp://i50.tinypic.com/4tpnhj.jpg
Pretty clear that despite the stop difference (and despite the Tinypic artifacting...) the RX100 shows quite a bit less noise still.
Rachotilko: Guys, stop the RX100 BS, please !
With its f/4.9 at tele end, it has to resort to ISO1600 when these small-sensor marvels can use ISO400 to keep the same shutter time for the same scene.
This ISO1600 vs IS0400 handicap can not be offset by the advantage provided by its larger sensor.
So: enjoy your "invention", and let the rest of us enjoy the fast lens.
Not everyone shoots at the long end all day long. At the shorter end you have close to a stop and a half benefit over the smaller sensors in terms of noise, probably the same in terms of DR. And that's in a much smaller package and the G15 weighs almost 50% more too.
"As for the RX100, it's less than a stop better."
Make that well over a stop when normalized to the same output.In terms of DR probably also well over a stop (ignoring the usual Canon CMOS banding, which in practise would make that even more).
They give a good idea what to expect in the circumstances displayed.
peevee1: Time properly judged the greatness of the leap - a really compact pocketable P&S matching or beating most DSLRs and mirrorless cameras with their kit zooms (as most customers using them), not to mention beating all the direct compact pocketable competitors by 2 stops or more (which usually translates to 4 years or more of technological development). DPR did not quite get the point, giving the camera only Silver award, just like to the bottom-of-the-heap Rebel and the rest of them innovation-deprived junk.Saying "we were very impressed" now will not save your face.
@Marike, yes it does. Take a brand new T4i and kitlens. Even at the long end of that kitlens, the difference in F stops is 2/3 EV in favour of the RX100. The Canon takes back a stop due to its larger sensor in low light, which leaves a mere 1/3 of a stop benefit. Back to the wider end the tables turn and the RX100 leads by 1 EV after taking both the above factors into consideration. It beats the T4i with DR and is close in terms of resolution.
audiobomber: I expect the RX100 would have easily earned a Gold award if it had fewer pixels, a hot shoe and a faster lens at the long end.
It easily outresolves the Nikon 1 series, as anyone can see. And it also outresolves both the LX7 and XZ1 as seen in the review: 2600 vs 2000 LPH and FAR less moire than either.
Pixelbinning serves little to no purpose when read noise is extremely liw as seen from this Exmor sensor. Adding it would in fact be a marketing ploy.
limlh: A game changer that symbolizes the development of digital photography. While other companies are pandering to conservatism and imitating old film cameras, Sony is not afraid to thread new ground and make full use of available technology. That is why RX100 represents the invention that is digital photography, and not Sigma that tried to squeeze a big sensor into an outdated body. Nor Fujifilm although it is also very innovative with sensor tech.
@Marike, my comment about low light speed of the Nikon was specifically about AF speed. That's where the Nikon is slower than even most recent CDAF systems, including the RX100. Only the Canon M seems to take exception to that comment.
And where I did mention the low light and DOF capabilities of the RX100, I compared it to APS-C DSLR cameras and their slow kitlenses. Which means that the stop or so you lose due to sensor size, is mostly made up for through lens speed. Sure there are exceptions but it even has more DR than the best Canon FF you can buy.
And I clearly talked about zoomlenses (specifically the standard kitlens range), so that rules out a single Nikon prime. Easy to take my comments out of their context and turn it into something I didn't claim. My context focused on what the majority of DSLR users shoots with.
And the obvious fact remains that all other cameras you and I mentioned are in a fully different size class. That's what seals the deal really.
Hopeless past ISO 800?
ISO 3200 concert:http://g1.img-dpreview.com/8E24E4A145784006991D4FA4D21FB592.jpg
That's more of a DXOmark assumption. If you scale the RX100 to FF, you have about 150 MP and a low light score around 3000. What would you have suggested in terms of MP and how much improvement based on that? If a factor 10 makes no difference worth mentioning, be prepared to go extreme. Say less than 1 MP? ;-)
Actually it does refute that claim in practise. Because if your theory was right, the latest high end sensors with pixels a factor 5 to 10 larger should show much much better performance per unit area yet. They don't. All just over 50% efficiency, which leaves less than a stop to the theoretical maximum. Which they won't touch anyway.If this was marketing over quality, what on earth went wrong with those D4, D3S, D800 and 1DX sensors? ;-)
Marike, neither the Nikon, nor the Canon would fit any jeans pocket. The Nikon has speed in good light but trails most cameras in not so good light and lacks a fast aperture zoomlens (at the wide end at least). The Canon is seriously compromised in speed in almost every way. The RX100 could be considered the first real digital pocket rocket.
In terms of DOF control, low light capabilities, DR, resolution and buffer clearing it's similar to a decent APS-C DSLR plus kitlens. Add good video qualities at1080P60 to that. All in a package the size of a Canon S100.....
The old misconception again. SNR is not compromised by the "extra" pixels in practise, proven by the fact that it's no worse than the Nikon 1 series sensor carrying half as many pixels. In fact, it has considerably less read noise at low ISO to boot and in terms of efficiency per area practically ties with the best DSLR sensors availble. Despite having pixels up to a factor 10 (or more in the case of the D3S) smaller.
zoranT: why isn t global shutter implemented in all cameras? too expensive tech? what s the disadvantage?
"why aren't these camera's CCD to begin with?"
They used to be. But the current Sony CMOS sensors offer:- much faster read out speeds (see high fps at high res)- much more lattitude (14 stops)- lower noise, especially read noise (the above is also related to this)- better efficiency (less power consumption, ceteris paribus)
Just to name a few things.
Great to see the lens tests return. What was good is still there and the additions are appreciated too.
ManuelVilardeMacedo: Why should anyone buy this instead of an entry-level DSLR?
@ Jorginho:If there's one thing that should be clear by now, is that there is no real correlation between pixel density and high ISO performance in practise. So suggesting the new sensor is going to be better than the RX100 sensor, is quite an assumption, which I very much doubt to be true (based on best efficiency currently available in any sensor, which the RX100 already hits).
iudex: First of all I must say I do not mind the look of V2; I do not buy cameras because they look nice, but because they take good pictures.But...The V2 is as big as mirrorless competition with APS-C.The V2 is as expensive as mirrorless competition with APS-C.I don´t get such good pictures from 1" sensor that I get from the APS-C.So why the he.l should I buy a V2 instead of an APS-C camera?
@Bavarian Raw Shooter:Only the mid telephoto range, simply because the NEX series doesn't have a collapsing design here yet.On the other hand, the existing Nikon 10-100 superzoom is actually heavier and arguably larger (much larger diameter) than the black Sony 18-200.