Nikon J1 versus Nikon D3100 versus Panasonic G3 versus Canon G12 versus Panasonic LX3

I got interested in the J1 when I realized that the new Nikon 1 system will provide some hefty magnification for those situations where the ultimate in low light image quality is not required. Examples include shooting birds or other wildlife and shooting performers on stage ay outdoor concerts. Reach really matters in these instances.

What makes the Nikon 1 system really special is the combination of a small, but not too small sensor and the promise of an F mount adapter for AFS lenses. Wow … that just rocks!

For example, my trusty 70-300VR lens is very sharp in middle at all focal lengths, and with the Nikon 1 crop factor of 2.7, the weakest part of the lens does not come anywhere near the image. The only issue left to be answered will be whether it can handle the pixel density of the Nikon 1 sensor.

So how difficult will this be? Let’s first examine the sensor size differences at play …

You will note that the 4/3” sensor is missing … I added the G3 as an afterthought and forgot to include it here. DPReview’s sensor sizes database shows the size of the 4/3” sensor as 18x13.5mm, so it slots nicely in between the 1” of the J1 and the Nikon APS-C sensor sizes, as expected.

Now, this is not a review of these cameras. And it is not an exhaustive analysis either. In fact, I plan to analyze one set of crops form imaging-resource.com in order to judge how these sensor sizes will do in tough situations.

When shooting birds in the woods, for example, ISO must often be raised to get decent shutter speeds, especially when shooting consumer glass at f/5.6 at maximum zoom. In such situations, bridge cameras and other cameras with so-called pinhead sensors really start to come apart. There just is not enough light gathering ability to maintain details at 800 ISO and above (and when you go into the woods, you will be shooting around there if you want decent shutter speeds.)

The APS-C sensor is a very good compromise these days for wildlife shooters. You get very nice high ISO performance and you gets lots of reach on the same lenses you might have shot years ago on film. This is, of course, for Nikon and Pentax users mainly. Canon cut of their users in the mid-80s with a totally new mount. And the others are all gone from that era.

Anyway, 4/3” is really starting to challenge these days because of improvements in sensor efficiency etc. Of course, all sensors benefit form such improvements, so APS-C and full frame continue to maintain their traditional lead in image quality. But as they all cross certain thresholds, the smaller sensors become applicable to more and more shooting situations.

Before I go on, let’s have a look at the one image that more or less covers the issues for me. This is a 1600 ISO image set from imaging-resource, as mentioned previously. Of course, imaging-resource.com maintains all copyrights for the original images, none of which are shown here. But this is the set of crops … please click on the crops to see them in full size. Otherwise, you will just have to take my word for it.

So, starting from the top. The D3100 has the largest sensor and moderate resolution at 14mp. This is a fine performance for an APS-C sensor of a recent, but not the very latest generation of technologies. It does not, for example, benefit from the dramatic improvements in dynamic range that the D5100, D7000 and K5 all share.

But it still acquits itself very well here. The hair looks very much like hair, with a stranded appearance and minimal (but not zero) clumping apparent. The eye looks detailed with the eyebrow giving a good impression of texture. The sleeve shows good texture in light and shadow, despite depth of field putting the sleeve slightly out of the focus zone. The  numbers on the watch face are all clear and everything has a nice, rounded 3-dimensional feel to it.

The G3 is a real dark horse here. Panasonic have been really improving their jpeg engines lately, with the GH2 being the last one that really impressed me. Here, the Panny shows very nice fine details. Some people will prefer this image over the D3100, and it is hard to disagree. But, of course, I will anyway :-)

What I don’t like about the G3 image is the hair clumping. This is common as borscht as sensor size falls. The noise builds to a point where noise reduction must smear out very fine detail like the lines between the hairs in order to get a decent image. Panny do it as well as anyone, but the laws of physics are not to be denied. The eye, on the other hand, is magnificent. The extra resolution really helps here as does their wonderful sharpening. Nikon are known to favor weak sharpening, which allows for more after the fact processing, but anyone who likes their images out of camera will favor the G3 over the D3100.

The sleeve looks superb. The extra depth of the focus field on the smaller sensor really helps it, as does the higher sharpening. This is a great performance. One could do far worse.

Now … the J1. This is the one I am personally interested in as I might be able to treat my 18-200VR and 70-300VR as the equivalent of a Bigma and an even bigger zoom … so how does this new 1” sensor stack up?

Well, at 1600 ISO, I think it looks a bit more like a compact camera than it does like a dSLR. The G3 looks like a dSLR (excepting the clumping) but this one looks like a better compact camera. Note, though, that I shoot my F550EXR in RAW only and I would shoot the J1 that way too. I will take a crack at the NEFs on the site when Adobe release the next ACR update.

Meanwhile, the hair is a bit mottled looking, which means that they are struggling to handle the chroma noise in this lighting. But there is still pretty good texture. The eyebrow looks good, though as does the eye. The low-ish resolution hurts the eyelash resolution, but it will look good in almost any web or small print output form. The sleeve is very nice. A strong hint of texture on the fabric and the watch face remains quite legible.

Moving on to the two compacts and the clumping is nearly complete at this point. Both show almost no hair texture and both have subtle mottling, although they hold it together very well. The LX5 does a slightly better job of the eye, but both suffer even when compared to the J1. Which is expected as we’ve taken another drop in sensor size.

There is little texture left in the fabric and the LX5 really smears out the threads, which remain quite legible in the J1 and mostly legible in the G12 image. The G12 watch face is only a bit down in legibility from the J1, but the LX5 watch face does not look very good.

I would have to say that, were I to shoot any of these smaller sensors, I would be shooting only RAW. The jpegs are not all that detailed.

So … my bottom line: Clearly, the 4/3 and APS-C sensors kind of stomp the smaller ones. But that should be expected. There is, however, a clear drop in image quality at each step, despite the Panny’s excellent showing. So you must be willing to step up your game if you want to try to obtain the results you can get from your dSLR with one of these smaller sensors.

I find that I can do really well with the F550EXR, a 1/2” sensor long zoom compact, but only if I shoot and process RAW. That is exactly how I would treat the J1, which means that I will have to give one a try one day. That crop factor would be a joy to use under certain circumstances.