Henrik Herranen: "It's also important to remember that not only is it substantially faster than any other zoom for APS-C (or indeed full frame)" [...]
You know quite well that f/1.8 on APS-C is the equivalent to f/2.8 on FF. Same DoF for equivalent focal lengths, same images. Saying the lens is "faster than any zoom on FF" doesn't take into account that a lens never stands alone, but is always a part of a system. (Remember all: ISO is just a number that specifically does NOT take into account the total amount of light the sensor gathers.)
Example: The Canon PowerShot G2 I bought in 2002 had an f/2.0-f/2.5 lens. Was the SYSTEM "faster" than a 24-70 f/2.8 mounted on a FF camera? You pick your guess.
Don't get me wrong: the Sigma seems truly to be an exceptional lens. It has good distortion, low chromatic aberration, and vignetting figures no fast FF zoom has been even close to. So it's a gamechanger in the APS-C world. Still, your speed claim as opposed to FF is misleading. A lens doesn't stand alone.
"Sorry, but f1.8 gives more light to a sensor than f2.8, ANY SENSOR => f1.8 has faster shutter speed than f2.8, ALWAYS."
No, you're assuming the same ISO and ISO does not mean the same amplification and thus same amount of noise.
You said that the f/1.8 lens would allow more light and that's simply false, because we're not looking at just a lens, we're looking at a system. A larger image circle is being illuminated on the FF sensor, which means more light, less noise, all else being equal. So that cancels out the difference in aperture and explains why people use aperture equivalents too. Not just because it tells us more about DOF differences.
Don't stop at shutterspeeds, think noise. f/1.8 on APS-C, f/2.8 on FF, same shutterspeed will result in the same noise for the same visual exposure. Those (sh.speed. and size of entrance pupil really) are the only 2 physical parts that determine the amount of photons and thus the amount of noise. The rest is amplification after the fact.
"n those terms yes, BUT: f1.8 => gives more light to APS-C sensor than f2.8 gives to FF sensor => faster shutter speed => faster lens"
No, if transmission losses were equal, both the FF and APS-C sensors would receive about an equal amount of photons using the same shutterspeed. Hence why noise will be about equal too for a given >visual< exposure.
QuarryCat: 810g weight, 12 cm long, 72 mm filter and then only 18-50 mm half format = boaring heavy.
For me it's useless - what shall I do with a 2,9/29-80 on a Eos 70D? I would always take a 2,0/24 mm; 1,4/50 mm and 1,8/85 mm or even Tamron 2,8/24-70 mm VC or a 5D3 or 6D.
The Tamron isn't nearly as sharp wide open and loses about half a stop of light on the FF cameras compared to the Sigma on APS–C.
Accounting for TStops, the statement from Dpreview is actually correct, even including FF lenses and using equivalents.
This plus IBIS = yum.
Transmission scores are also unusually close to indicated F stops, which means that in practise, noise/exposure benefits over smaller aperture lenses can be even greater than the differences in apertures and/or exposure times suggest.
JackM: I guess I don't see the point of this lens anymore, now that ISO performance is at least 1.33 stops better than it was when I first bought my 17-55/2.8. I'll keep the extra 20mm (30mm) of reach, thanks.
And don't say DOF control. The DOF of a 35mm lens at f/1.8 is pretty deep unless you are right on top of your subject. Certainly not enough shallower than f/2.8 to warrant giving up the range.
Look at the Tstops, it's about 2 stops difference in reality. Obviously, the difference in shutterspeeds will not reflect that, the noise for a given visual exposure will (cameras also boost under the hood to compensate for Tstops).
dpLarry: 13mp on a tiny sensor in a phone. Huge grainy file. I'd prefer 5mp with over 2 1/2 light gathering ability.
@Dplarry:You're confusing the sales success of a phone as a whole with the qualities of the camera.In good to medium light, the camera of the One lags behind most competitors by some margin, in low light, it's pretty similar to some of the others, including the S4. See for example this blind test at the same output size done amongst hundreds of people (Google Translate will help out):http://mobile-review.com/articles/2013/blind-comparison-result.shtmlAnd some context:http://www.mobile-review.com/review/phototest2013.shtml
TrojMacReady: You can add a camera shortcut to the lockscreen (without the need of 3rd party apps or rooting, it's just a simple settings change). Here are a few ways to do it:http://www.androidcentral.com/three-ways-add-camera-shortcuts-galaxy-s4-lock-screen
Wow, you guys are fast. I see it's already removed from the list of "The Bad" on the conclusion page. Kudos!
I assume page 3 adjustment will follow.
You can add a camera shortcut to the lockscreen (without the need of 3rd party apps or rooting, it's just a simple settings change). Here are a few ways to do it:http://www.androidcentral.com/three-ways-add-camera-shortcuts-galaxy-s4-lock-screen
ntsan: Judging from all the photos released the 808 still have advantage in terms of image quality. The noise on the 1020 is a bit too much, and bokeh is not as great too. (Toshiba sensor > Sony sensor)
But with OIS the video should be much improved though
The 808PV sensor is quite a bit larger, so despite the BSI design of the 1020 sensor, it shouldn't be too much of a surprise. That being said, the new sensor does seem to handle wide DR scenes much better (the 808 was notorious for clipping).
Photato: DPReview comments about the Nikon D800E (36mp) full frame DSLR:"Pushing these cameras to achieve their maximum level of detail requires an investment of both time (methodical preparation) and money (the very best lenses Nikon makes). "41mp is a marketing gimmick...
@ Draleks:That's a DOF/focus difference in part due to the different FL's used, not really relevant to the discussion of resolution.
I'm also not sure why "of course it should lose squarely" against the Canon S110.
Gadgety: "Nokia has managed to take some very novel and impressive imaging technology and fit it all eleganty in a device that is effectively no bulkier than most high-end smartphones. "
So true. It's very, very impressive. I guess the desire to lessen the bulk has also diminished the sensor size from the 808's 1/1.2 to 1/1.5. Never mind, this is the #1 camera phone.
Yet with a lens that is a hair faster.
Photato: You'll need a tripod to get 41mp worth of resolution, good light and when subject is still. Or when seeing resolution charts in review sites.In the real world however, when handheld, 41mp is a blurry mess.Bigger pixels is the way to go. Just like DSLRs
I doubt they were using a tripod on some of the shots taken from the helicopter... ;)And while not 33MP of true resolution (in the aspect ratio used) and limited by soft corners, it still beats all shots I have seen from other phone cameras in terms of resolution, minus the Pureview. And scaled to the same output size, there's much less halos and other sharpening artifacts, even compared to some older ILC's. Bigger pixels wouldn't have helped in that regard. At all.
qianp2k: When DXOMark will test this 41mp cellphone camera or older 41mp 808 cellphone camera? I want to see their P-MPix, guess no more than 5 mpix :-)
It (the PureView) outresolves the older 12MP µ4/3 cameras with a decent lens. So I doubt that.
(unknown member): You are better off with a Sigma DP1 or 2 Merrill series camera for the low ISO range and something else for high ISO shooting and you'll have plenty of cash left over. For the low ISO shots you'd have by far much better image quality with the Sigmas. It's stunning and shameful how DPPreview ignores the newest Sigmas.
Except they (the Sigmas) have high read noise and thus far less shadow range.
rhlpetrus: Where's the advantage of this camera over the D600 in IQ? The samples show better detail in favor of D600. Maybe the fixed lens is not so good. I think a fixed lens camera for this price should have just the best lens in the world.
Uh, the fixed lens is about as good as it gets at 35mm. At F2 it even beats the legendary Sigma for corner sharpness.
You do realize there's a DOF and curvature difference because the RX1 was shot much closer to the scene (35mm vs 85mm)? Have a look at the Queen card.
Falcon31: Nice tech, but what problem does it solve?
The whole point of a small sensor system is that it is small. Now with the speedbooster we can use our old BIG HEAVY lenses again. Now that is what I really want.
A) There are movie cameras with a NEX mount (which aren't necessarily small)B) There are NEX owners with a large set of Nikon glass, basically turning their NEX into an (emergency?) backup bodyC) Not all Nikon glass is large
Mssimo: That is real pizza..not the junk we have here in the US.
@ John Koch, the "world" was long eating what they considered and still is considered pizza before American Italians existed. And not every pizza has a tomato sauce so I'm not sure what the Aztecs have to do with this or the US.
Yanko Kitanov: At higher ISO BIS is clearly better - but is high ISO what you buy a pocket cam for? What do you use more often - low ISO or high ISO? If you have your answers, please note that at low ISO the older non BIS sensor is BETTER, the BIS architecture has some clear drawbacks at low ISO, fact.
Facts without RAW comparisons are opinions or guesses.
tolleknolle: The claimed 1/2 stop seems to be in JPEG processing. RAW data puts the original RX100 well ahead of Mk2 at lower ISOs: http://www.techradar.com/reviews/cameras-and-camcorders/cameras/compact-cameras/sony-rx100-ii-1160503/review/5#articleContent
Techradar comparisons have always been a joke. Non normalized and non controlled tests, resulting in spikes and misleading results all over the place. Hence why their results rarely make sense and are usually contradicted by more reputable hand repeatable tests (DXO et al.).