Alphoid

Because people shoot different things, and need different focal ranges. There's no magic about an 18-50mm. 18-50mm is more useful for landscapes, but isn't a good option for portraits. On crop sensor, 24-75mm starts to be quite decent for portraits. If you were using out-of-camera shots, 33-100mm would be ideal, but 75mm crops decently to 100mm.

Good comparison, although "roughly equivalent" is a bit of a stretch. The NEX is roughly half a stop ahead in equivalent aperture, which sets DOF. The Nikon 1 is a niche product for places where you need the focus system. The NEX is well ahead for general-purpose photography (especially when you factor in the 50mm f/1.8 is stabilized).

@magneto shot

Multiple small sensors can give narrow depth-of-field.

Just so folks are clear, this is very different from Lytro. A camera array like this can give dSLR-style depth-of-field and low-light performance in a cell phone form factor (at a dSLR-level price). Splitting the lens across many sensors gives more sensor, but without needing long focal length, and you can get a large effective aperture in a small form factor.

In addition, if gives lightfield information, so you can adjust focus and depth-of-field. You can do HDR by setting individual sensors to different exposures. You can shoot half of the sensors during flash, and half with motion blur, and combine. You have distance information, so perfect magic-wand-object-select tools.

What remains to be seen is whether Nokia can partner this (or PureView) with a modern display (e.g. 1080p) and OS (e.g. Android). If so, I'll buy one.

Nokia should dual-boot Windows Phone and Android, and let the consumer choose. This seems like a win-win for Nokia, Microsoft, and the consumer.

The copyright at the bottom says "All Rights Reserved" (approximately). I'm not sure if dpreview would go for me modifying their review...

Good first pass at a review. Things that are clear missing:
* Proper specification table. At the very least, this should include sensor size, min/max ISO, modes available (PASM), and RAW support.
* Action shot. The core reason to use a telephone and not a camera is because it is with you. Do I get blur?
* Comparison at identical and reasonable sizes (e.g. 4 real megapixels -- the resolution of a 27-30" monitor so the most you'll practically use, or even lower). Sinc scale all images to 2-4MP, and allow us to view with that as an option.
* Low-light shot where you're limited by handshake.
* Low-light shot where the 250x188 crop is effected by low-light (as you'd post on the web).
* Some semblance of a test of the optical image stabilization. Without IS included in the test, the Samsung clearly wins. With IS, HTC might win.

In an ideal case, I could also view the images double-blind. You scale to ~3MP. I rank by quality. I find out which came from where later.

sunnycal: NetMage is correct. But just do an experiment. Shoot a photo. Crop out a portion of the photo, and resample both to the same resolution. Compare noise levels. You will find the cropped version will have higher noise.

DStudio: If you shoot maximum ISO and crop, you'll get insane noise. If you're willing to live with insane noise, you can also underexpose while shooting RAW, and boost exposure in post-processing.

Considering Sigma's 200-500mm f/2.8 is $26,000, Sigma sometimes gets away with a lot. Still, if Sigma can give this a reasonable price -- competitive with the f/2.8 zooms -- they can probably dominate the market.

sunnycal: If you have sensors of the same technological sophistication, and you shoot APS, 1/8s, ISO1250, f/1.8, the depth-of-field, image-level noise, exposure will be identical to shooting full frame 1/8s, ISO2000, f/2.7. Full frame will have somewhat better resolution and/or dynamic range. All else being equal -- which with lenses it almost never is -- the f/2.7 lens will also have lower aberration.

sunnycal: You are incorrect. handrian is right. The level of noise of the overall image will be identical if you shoot f/2.7 full-frame or f/1.8 cropped (assuming same exposure, and appropriately scaled ISO). Only real difference is that you lose a few MP going to DX-size, and likely have worse lens performance (we have to wait for samples, but it's hard to build an f/1.8 zoom with the same IQ as an f/2.8).

Nope. It's not the first constant f/1.8 zoom. There are plenty of f/1.8 zooms (and even faster) for video cameras, machine vision applications, etc.

I love the Sony simulated photos. At least this time they included a disclaimer. Previously, I'd seen photos of with/without photos for their steadyshot image stabilization showing sports photos with motion blur added/removed.

The Canon Digital Rebel was the first sub-$1000 dSLR. Adjusted for inflation, the D7100 is $962 in 2003 dollars. To the best of my knowledge, this is the first camera breaking the same ground for professional dSLRs as the original Rebel did for dSLRs.

Exciting times.

If the lens were flat and not in contact with the object, you would be correct. In the case of microscopy, the object can be immersed in the optics (e.g. oil immersion microscopy) or otherwise surrounded by optics from all directions. You're not quite at f/0, but close enough it doesn't matter anymore (performance is nearly identical to f/0).

The last bit of this article is incorrect. f/1.0 is not a full stop faster than f/1.4, so the last two shots did not have identical exposure. As you reach f/0 (which is about where many microscopes operate), the light does not go to infinity. f/1.4 is about the last stop where dividing by 1.414 to go up a stop is a fair approximation (this relies on small angle approximations such as sin(x) being approximately equal to x).

For a more detailed explanation, see:
http://en.wikipedia.org/wiki/Numerical_aperture#Numerical_aperture_versus_f-number

NA sets exposure. f-stop and m-stop are derived from NA. As a sidenote, I'd love to see dpreview operate on m-stop and mISO.

Key question is whether they're any good. My existing Tamron lenses have been disappointing for both IQ and especially build quality. One developed horrible zoom cheap after a few years. The second developed a rattling sound. Warranty appears useless since I did not keep receipt, although the lens was introduced recently enough that they know it is under warranty... They were well treated and lightly used.

It's very simple:

1) I want three knobs. One is ISO, one is aperture, and one is shutter speed. I can set any of the three to auto. The whole concept of two knobs/PASM is based on having to swap out film to change ISO. We've been digital for over a decade, and we still haven't caught up.

2) I want a mode where, when the camera powers up, everything is set into full auto. I can still change everything, but I don't have to worry that a week ago, I was shooting in a dark chapel at ISO6400, with yellowish white balance. Late Minolta/early Sony were a lot like this.

I don't really care what I'll pay for it -- unless there's something else wrong with it, I'll buy it -- but in practice, I'd like to see this on a bottom-of-the-line learner camera. Lots of students could benefit from additional controls without paying $1500 for fancy focus tracking, bleeding edge sensor, etc.

Wonderful! Desperately needed.

A very useful test bench would involve a compact sensor positioned behind a universal lens mount with MF so you could compare Canon vs. Nikon vs. Sony. u4/3 or NEX would work in abstract, but smaller, higher-resolution sensor would be able to tell you a lot more.

Thank you for the thoughtful response. Personally, I'd buy one as soon as it has a real OS (not Symbian). I assume you're working on this.