14-megapixel Sigma DP2 camera now officially available

A few retailers went ahead and put the Sigma DP2 up for pre-order back in March, but it looks like the camera is now finally officially available to own in the UK -- and, judging from some of those aforementioned retailers, at a few places in the US as well. Those in the UK can expect to pay £599.99 for the camera, which updates the DP1 with a new 41mm f/2.8 lens and Sigma's improved TRUE II processing engine, while folks in the US will apparently have to pay $699 retail (although at least some places still seem to be listing the $649 pre-order price). Interestingly, Sigma will also apparently be offering the older DP1 alongside the DP2 for the time being, although it's not clear how long it'll continue to do so.























Looks nice,
and as everyone knows, the more megapixels, the more better your photoes!
Smiley Faece :)
It has a sensor much larger than other 10+ MP point and shoot...
yeah i know, i was being serious dumbass.
actually Oli, the bigger the megapixel doesn't mean the better the photos, it means the bigger the picture can be printed.
Better photos determined by lens quality, not megapixel. a 5MP DSLR Canon takes much better photos than a 12MP Sony camera.
Oh wow, thanks toy,
perhaps you misunderstood.
Double sarcasm is thrice the fun.
The quality of a photo is determined by the person behind the camera. Lenses are more about clarity and sharpness.
it's APS-C so the sensor is like 6-8x larger than normal POS ahem...i mean P&S cameras, the quality of this thing is supposed to the same as a DSLR of the same price range.
How the hell did you guys miss his obvious sarcasm? Are we going to have to start using the /sarcasm tag again?
Camera enthusiasts have no time for sarcasm
Your public is pretty dense... don't forget, this is still an internet forum...
mega-pixels giga-schmixels...
Keep your faeces off this blog please, smiling or not..
In this day and age a single focal length where the aperture size isn't really all that impressive (f/2.8) isn't all that viable a product, especially at $699!
It's the only way to cram such a large sensor into a little camera body.
Im with you..I'd get even a G9 or something else
If it were fullframe for $700 then yes, but (n)
It is if you aren't a cripple. I zoom with my feet very frequently.
It looks like old style cameras, reminded me an analog camera I had years ago.
I really like this camera, altough it is a little bit pricy.
What's the sensitivity on the sensor like? Are the Foveons less sensitive by design?
THEORETICALLY
Foveon can gather more light, because it collects all colors throughout the sensor area...
Bayer on the other hand, only gathers one color per unit area. In other words, where there's green sensor, it discards all other colors (so basically discards 2/3 of the spectrum).
REALISTICALLY
Doesn't seem to be the case, but hard to compare a first/second generation Foveon with 23rd generation Bayer. They seem to be comparable at this point. It's also unclear as to how ACTUALLY the foveon sensor is constructed to see what percentage of the light is actually gathered.
I think the DP1 is still a viable, separate product as it has a different focal length lens. Hopefully they will update the processor in the DP1.
i'll pass
So much for 12 megapixels being the upper limit for cameras.
hey they said it was 12 kilopixels in 1997
Foveon MP != CMOS MP. Both are actually fake.
CMOS pixels capture only one color. Foveon pixels capture all three colors. IOW, 10MP CMOS sensor has effectively 2.5M color pixels (5M green, 2.5M red and 2.5M blue), rest is interpolated. Foveon 10MP would have 3.3M real three-color pixels. (See http://en.wikipedia.org/wiki/Sigma_DP1 )
Biggest problem of Foveon right now I gather is that Sigma doesn't have much money (compared to e.g. Canon) to invest into its development.
http://en.wikipedia.org/wiki/Foveon_X3_sensor
The problem I have is that my LCD screen has RGB pixels. If my screen had pixels that could display any color (and various shades as well), then it might be something to look at.
But, it's the quality of the pixels (and number of them) that matter to me. Their theory might be right and better, but they have a lot of competition still using the current standard CMOS/CCD sensors that produce better images today.
@rcappo
Now that you mention the screens...
The screens are MUCH CLOSER in methodology to the Foveon processor. They have a square pixels that contain 3 colors in equal quantities (no double-green like Bayer pattern), the sure as hell don't extrapolate color information onto the neighboring pixels either (again, like Bayer does).
You got your HATE mixed up, buddy.
This is really a 4.5 MPx sensor, but each site gets "full color." Bayer sensors give better spatial resolution and interpolate color. Historically speaking, the Foveon sensors have also lacked the low-pass (anti-aliasing) filter necessary to avoid aliasing (e.g. Moire). I don't know if the DP2's sensor also has this technical drawback. Also, the Foveon sensors have had trouble getting their color rendition right, and have relatively poor high ISO/noise performance, despite their relatively large pixels.
Note that the human eye also interpolates color to some extent and has greater spatial resolution than color resolution.
While some people enjoy their Sigma cameras, there seem to be several technical flaws. I would avoid them and wait for other P&S-style cameras with big sensors like Micro4/3 from Panasonic, Olympus (forthcoming) or Samsung (also forthcoming).
Erm.. not having an AA filter is a GOOD THING. It's not a drawback - it's the whole point.
Normal RGB sensors need the fuzzy filter to remove moire (and reduce effective resolution), this one doesn't.
Yem,
You need to go back and study your Shannon/Nyquist. In any digital sampling system, if you omit an anti-aliasing filter, you will get aliasing when there is energy in the signal above the Nyquist frequency. If you omit the AA filter, you get pseudo sharpness, false detail, and expose yourself to high frequency aliasing and additional Moire. The artifacts in a Foveon-style sensor without an AA filter look different from those you would see in a Bayer sensor, (e.g. you don't get COLOR Moire), but you still get aliasing and luminance Moire.
If you're an engineer who makes an A-D converter (image/audio/whatever) and omit an anti-aliasing low-pass filter of some kind, you are making a technical mistake.
-a
Of course, with pixels (true RGB pixels) you'll always have extra "detail", because the edges of pixels are theoretically pure straight lines as opposed to real detail (which is rarely straight, and never exactly straight where pixel meet)
I don't see a problem with this. The low-pass filter you're talking about (only if that's your forte) would be to quadruple (or increase any-fold) the resolution and then diffuse the pixel information into the nearing pixels (or something a little more complicated). Either way, the frequency domain you're talking about is the actual pixel count; there's NO WAY to insert a low-pass filter into the optical path, so it must be done post-digitizing stage. And I'd argue that this is a made up problem, and a simple rhetorical fundamentalism applied to a problem that never existed.
The reason Bayer does "not" have those problems is because that is what it naturally has to do in order to re-mosaic the image (again this happens, as processing after digitizing the pixel data).
@Pretol:
On the contrary: the low pass filter _has_ to go in the optical path. That's the whole point. you have to get rid of the too-high-frequency component of the signal before you do the sampling. If you don't, you'll get artifacts. You have to put a blurring filter in front of the sensor to lower the effective resolution to just slightly below the effective resolution of the sensor.
With a Bayer, you might need to do a stronger AA filter than Foveon to get proper color rendition, but even if you've got a completely black and white sensor, you still need one.
Just run AA in Photoshop, if you feel it's necessary.
Jeebus,
Once you've gone from the analog to the digital domain, it's likely too late. Post sampling, you can't necessarily tell the difference between signal that was orignally above the Nyquist freqency and now aliased below it, and signal that started out below the nyquist frequency. Ask any competant engineer with knowledge of A-D conversion or sampling.
Yes, there are techniques that can reduce the effects of a missing analog domain AA filter, but they are not nearly as effective as simply avoiding the problem in the first place, and can cause you to lose real detail in the process.
As of this writing, the wikipedia entries on Aliasing and Anti-Aliasing have good example images and explanations. To be sure, you can get aliasing, e.g. Moire, during either the A-D (i.e. taking the picture) or the D-A processes (displaying the image). You need AA in both places to avoid it.
Read up on your Nyquist-Shannon sampling theorem. You need a bandwidth-limited input signal for A-D.
I'm not sure if this link will work, but it's a link to igital imaging for photographers By Adrian Davies, Phil Fennessy p 31 in Google Books
http://books.google.com/books?vid=ISBN0240515900&id=wsxk03-gceUC&pg=PA30&lpg=PA30&ots=BlxgR6ZXmz&dq=anti-aliasing+lithium-niobate&sig=YUsQZMtxjtq70oOYWmU1K2yfoNY#PPA30,M1
Great we in the UK get to pay $900, thanks Sigma.
That was I was thinking. Shouldn't someone buy them in the states and then ship them to the UK as grey market?
i thought that looks like the back of a new and improved... well, centered, part of a LG Viewty-esque phone.
Fyi; Sigm plans to keep selling the DP1 alongside the DP2 because the DP2 does not replace the DP1. The DP2 is largely the same as the DP1 aside from the lens. The DP1 has a telephoto 41mm prime while the DP2 has a wide angle 24 mm prime. Although I have no idea who in their right mind would pay for both of these cameras. Might as well pick up the SD15 which is an SLR and therefore you can change lenses. http://www.engadget.com/2008/09/23/sigmas-sd15-dslr-an-sd14-with-improved-image-processing-and-li/
A dSLR however would be a much larger camera. The point to the DP series is that it's more or less pocketable.
Does any one know where to find a comparison between this and Panasonic DMC-LX3/Leica D-Lux 4?
Yes and no. You can compare the DP1 (should be similar to the DP2) output to the LX3 output at Imaging-Resource.com. http://www.imaging-resource.com/IMCOMP/COMPS01.HTM
Check the Sigma forum on DPreview.
About time! :)
The DP series has its drawbacks, but image quality is not one of them. Awesome sensor, awesome lens, awesome pictures. It really is like a quite decent DSLR in your pocket.
Oh, and as already pointed out, it's complimentary to the DP1, it's not a replacement. I'd love to have both.
Amazon definitely had some and shipped them in the US. I pre ordered one and received it early this week.
@toy, sony makes slrs, actually, they make better slrs than canon (excluding 1d)
Don't say stupid things. Sony's DSLR's arent better or worse than Canon's. They've got different drawbacks and advantages. That's all.
Actually, Sony makes shitty SLR's compared to Canon. Canon has professional line of lenses and also the broadest lenses. They have been used and trusted by pros worldwide, and they weren't disappointed by it's quality. I have a Canon and I love it.
@iKurt 5D
Have you actually used any of the Sony/Minolta "pro" lenses? They're easily as good as, if not better, than L lenses - both in build and in image quality. It's just a shame the bodies don't quite match up so well.
@iKurt5D
It's fun criticizing a product that you don't have right ?
For that amount of money I could buy an XSi with Canon's award winning array of pro lenses that are trusted by pros worldwide. I'd recommend you skip Sigma and buy Canon XSi instead, since the XSi supports interchangeable lenses and the kit includes an IS lens. This Sigma is not a proven technology and they're not the leader in the industry.
You're missing the point...
Canon XSi (with 18-55mm kit lens): 12 x 14 x 9 inches, ~4 pounds
Sigma DP2: 4.5 x 2.3 x 2.2 inches, 9.3 ounces